RIVERS AND CANALS
VERNON-HARCOUR T
VOL. I.
Bonbon
HENRY FROWDE
Oxforo University Press Warehouse
Amen Corner, E.C.
Qlew 2)orR
macmillan & co., 66 fifth avenue
RIVERS and CANALS
THE
FLOW, CONTROL, AND IMPROVEMENT OF RIVERS
AND THE
DESIGN, CONSTRUCTION, AND DEVELOPMENT OF CANALS
BOTH FOR NAVIGATION AND IRRIGATION
WITH STATISTICS OF THE TRAFFIC
ON INLAND WATERWAYS
BY
LEVESON FRANCIS VERNON-HARCOURT, M.A.
MEMBER OF THE INSTITUTION OF CIVIL ENGINEERS
AUTHOR OF 'HARBOURS AND DOCKS,' AND 'ACHIEVEMENTS IN ENGINEERING'
IN TWO VOLUMES.—VOL. I
RIVERS
SECOND EDITION, RE-WRITTEN AND ENLARGED
AT THE
Oxford
CLARENDON
1896
PRESS
Ö^forb
PRINTED AT THE CLARENDON PRESS
BY HORACE HART, PRINTER TO THE UNIVERS!! Y
I )t (o I g cl(->
LIBRARY ^ •'
BUREAU OF RAILWAY ECONOMIC';.
C. E. StECHERT & CO.,
IM WrST 25TH ST.,
' CITY.
OCT 0 1910
PREFACE
When, early in 1892, my publishers informed me
that a second edition of ' Rivers and Canals ' would be
soon required, two alternative courses appeared to
be open to me. Motives of expediency indicated the
advisability of merely revising the book, so as to bring
it up to date ; for in the first edition of a book, with
costly illustrations, and of which more than half the
copies have been sold in America and foreign countries,
at the usual reduced rate, the author's share of the
profits little more than suffices to defray the expenses
incurred in the preparation of the drawings for the
plates and woodcuts. On the other hand, the experi¬
ence gained, since the book was first written, from
attending six international navigation congresses, in
visiting numerous rivers and canals on the continent,
by inspecting professionally a number of rivers in the
United Kingdom, in writing ten papers on river and
canal works for the Institution of Civil Engineers,
international congresses, and other societies, and in
the preparation of full reports upon the condition and
improvement of several rivers, could not have been
embodied in a mere revise of the book. Under these
vi
Preface.
circumstances, it appeared incumbent on me to dismiss
all considerations of personal advantage, and to en¬
deavour to the best of my ability, and with such
opportunities as my scanty leisure time afforded, to
utilize widened experience, increased knowledge, and
the additional published matter available, in the pre¬
paration of as complete a book as practicable, within
reasonable limits, on river and canal engineering.
Accordingly, with the exception of some paragraphs
from the first edition, incorporated here and there in
the earlier part of the book, and especially in the first
two chapters, the book has been practically re-written.
Rivers and canals have been dealt with in two separate
parts, forming respectively the two volumes, as distinct
from one another as the intimate connection of the
two subjects will allow, instead of being mixed up
together as in the previous book ; and the various
branches of each subject have been classified, as far
as possible, under separate chapters. Although several
of the plans and sections illustrating the previous
edition have necessarily been reproduced, but with
additions and to a reduced scale, the plates have been
entirely remodelled ; and a considerable majority of
the drawings did not appear at all in the former set
of plates. Some of the least important of the previous
illustrations have been omitted ; but the plans have
been so reduced and arranged, that a considerably
larger number of illustrations have been brought
together within a smaller space than formerly ; similar
scales have been adopted, as far as possible, for the
plans and sections in the same plate ; and the colouring
Preface.
vii
introduced into the drawings, at the suggestion of the
lithographer, adds greatly to their clearness.
Besides re-writing the descriptions of river and
canal works given in the first edition, under the light
of more recent experience and fuller information
numerous other works have been described, such as
for instance the training works of the Loire, the
Weser, the Nervion, and the Ribble, irrigation canals,
the inland navigation systems of various countries
with their traffic, and the Manchester, Baltic, Corinth,
and Nicaragua canals. The principal publications
from which information has been derived are given
in the footnotes, to render reference to them easy ;
and plans and sections, kindly furnished me by
engineers, have been acknowledged in a similar manner.
I also desire to express my sense of the assistance
I have derived, in the preparation of this book, from
the excellent technical library of the Institution of
Civil Engineers, the published Proceedings of that
Institution, and the Papers read at the international
inland navigation congresses held within the last ten
years.
My chief aims in this book have been to endeavour,
by the investigation of the various physical conditions
affecting rivers, as illustrated by observations, the
results of works, and experimental inquiries, to place
the principles of river engineering upon a more
scientific basis, and to give, for the benefit of engineers,
concise accounts of the chief works carried out for
the control and improvement of rivers, and the con¬
struction of canals. Elementary considerations and
viii
Preface.
explanations have been retained in the earlier chapters
of the book, with the object of rendering it useful
to engineering students ; whilst it is hoped that the
descriptions of important works of general interest,
with the aid of the illustrations, will render the book
acceptable to the public, and that statistics relating
to traffic on inland waterways and ship-canals will
prove of interest to traders.
In conclusion, I would fain venture to hope that
my earnest endeavours to render this book worthy
of the important branch of engineering with which it
deals, a means of promoting the progress of river
engineering on scientific lines, and valuable generally
to engineers, will meet with the approval of my
professional brethren at home and abroad ; and if this
book should succeed in gaining their approbation,
I shall regard it as my highest and most enduring
reward.
L. F. VERNON-HARCOURT.
6 Queen Anne's Gate, Westminster.
December 9, 1895.
P.S. The final revision of a portion of the text,
and the completion of the index, have had to be
accomplished, under somewhat difficult conditions,
during a voyage to India, taken for the purpose of
inspecting the River Hooghly, and reporting thereon
to the Commissioners of the Calcutta Port Trust.
CONTENTS OF VOL. I
CHAPTER I.
PHYSICAL CHARACTERISTICS.
i
Introduction. River Basins. Differences of Rainfall. Variations in Rainfall.
Evaporation and Percolation. Influences of Forests and Vegetation. Effect
of Period of Rainfall. Impermeable and Permeable Strata. Available
Rainfall. Forms of River Valleys. Fall of Rivers. Transportation of
Material. Influences of Lakes. Divergence of Current. Tidal, and Tideless
River Outlets. Importance of Scientific Study of Rivers ....
CHAPTER II.
MEASUREMENT AND FORMULAE OF DISCHARGE.
Importance of Discharge Measurements. Definition of Terms:—Discharge;
Slope; Sectional Area; Wetted Perimeter; Hydraulic Mean Depth; Mean
Velocity. Measurement of Cross-Sections. Velocity Observations :—Floats ;
Current-Meters; Gauge-Tubes; Hydrodynamometer. Comparative Values
of Methods for measuring Velocities. Distribution of Velocities in a
Channel. Mean Velocity deduced from Surface Velocity. Formulae
of Discharge:—Humphreys and Abbot's Formula; Darcy and Bazin's
Formula; Ganguillet and Kutter's Formula. Concluding Remarks.
CHAPTER III.
REGULATION AND CANALIZATION OF RIVERS.
Imperfect natural Condition of Rivers. Variable Flow of Rivers. Progression
of Detritus. Highest and Lowest Water-levels to be observed. Regulation
of Rivers:—Removal of Shoals and Obstructions ; Contraction of Channel;
Cross Jetties ; Longitudinal Training Banks ; Protecting and Easing Bends ;
Straight Cuts; Submerged Cross Dykes; Influence of Descent of Detritus;
Remarks on Regulation of Rivers ; Instances of Regulation Works. Canaliza¬
tion of Rivers :—Primitive Navigation; Stanches for Flashing; Locks,
Weirs, and Level Reaches ; Instances of Canalized Rivers ; Remarks on
Canalization of Rivers ..........
x
Contents.
CHAPTER IV.
dredging and excavating.
Definition of Terms. Dredgers:—Bag and Spoon; Bucket-Ladder Dredger,
Stationary and Hopper, details, cost of machines, and cost of Dredging ;
Dipper Bucket Dredger ; Grab Bucket Dredger, varieties, and capabilities ;
Sand-Pump Dredger, instances, capacity for work and cost of Dredging;
Eroding Machines ; Remarks on Dredgers ; Removal of Rock under Water.
Discharge of Dredged Materials—through a Hopper ; by a Chain of Buckets ;
direct into Wagons ; through Long Shoot; by Pump and Floating Tubes.
Excavators:—Bucket-Ladder Excavator, method of work, cost, capabilities ;
Steam Navvy, varieties, cost, capacity for work ; Grab Excavator ; Remarks
on Excavators
CHAPTER V.
locks; and fixed, and draw-door weirs.
Locks :—Description of Parts ; Construction, varieties ; Lock-Gates, single and
double, wooden and iron, strains, support, movement. Weirs:—object,
classification. Fixed Weirs:—Description, disadvantage; Oblique Weirs,
object ; Angular and Horse-shoe Weirs, advantages, forms ; Construction ;
Anicuts, methods of construction, instances, defects. Draw-door Weirs:—
Description, object ; Teddington Weir; Indian Dams; Draw-doors sliding
on Free Rollers, instances, effects, for regulating tidal flow on Thames
and Weaver; Segmental Gate Weir, description, object, instances. Con¬
cluding Remarks
CHAPTER VI.
movable weirs.
Definition; Classification. Frame Weirs:—Needle Weir, description, method
of working, instances, improvements, advantages, limit of height ; Sliding
Panels, object, advantages, instances; Hinged Curtain, description and
method of working, compared with panels ; Suspended Frames, object,
description of Poses weir, advantages, cost compared with other forms.
Shutter Weirs:—Bear Trap, description, method of working; Thenard's
Shutters, description, working, modifications adopted in India; Hydraulic
Shutters, description and cost; Chanoine's Shutters, modifications and
working, introduction of foot-bridge, cost, adoption in America, system of
double grooves, advantages ; Tilting Shutter, description. Drum Weirs:—
Description, at Joinville, on the Main, at Charlottenburg, merits. General
Remarks, summary of respective types, and comparison of systems
Contents.
xi
CHAPTER VII.
PREDICTION OF FLOODS ; AND PROTECTION FROM INUNDATIONS.
PAGE
Causes of Floods. Prediction of Floods :—by Rainfall Returns ; various
circumstances producing Flood Rise ; Observations of Rise of Tributaries ;
accuracy attained. Protection from Inundations :—necessity for, in certain
cases; Methods adopted ; Extension of Vegetation and Forests ; Regulation
of Torrents, and Arrest of Detritus; Catch-water Drains; Reservoirs for
impounding Floods; advantages, objections; Removal of Obstructions;
Utility of Movable Weirs; Enlargement of Channel; Straight Cuts;
Embankments, low banks, high banks, respective advantages, conditions
necessary for high banks, high banks overtopped by rise of flood-level ;
Instances of Embanking Rivers, Mississippi, Po, Loire, and Theiss ; Rising
of River-Bed, in Japan, 011 Yellow River, causes; Pumping; Protection
of Upper River Valleys. Ports of Refuge. Concluding Remarks . . 148
CHAPTER VIII.
DELTAS OF TIDELESS RIVERS ; AND IMPROVEMENT OF THEIR
OUTLETS.
Contrast between Deltas of Tideless Rivers, and Tidal Estuaries. Deltas :—
Formation, Form, Bars; Advance; Rhone Delta; Nile Delta; Danube
Delta; Volga Delta; Mississippi Delta; Comparison of these Deltas;
Methods of Improvement. Dredging and Harrowing on Bar :—Objects ;
at Outlets of Danube and Mississippi; at Volga Mouths, results, proposals.
Parallel Jetties at Outlet:—To scour the Bar; Conditions of success, and
permanence ; Rhone Jetties, previous conditions of Mouths, works carried
out, their effects, causes of failure ; Danube Jetties, Sulina Mouth selected,
form of works, results, causes of success, compared with Rhone works,
gradual shoaling in front, improvement of Sulina branch ; Mississippi
Jetties, South Pass Outlet selected, works described, results, works at Head
of Pass, shoaling in front of outlet, dredging, prolongation of jetties imminent ;
Mississippi and Danube works compared. Ship-Canal instead of a Delta
Channel:—Object, at Ostia, San Carlos Canal, Mahmoudieh Canal,
St. Louis Canal, proposed for Danube and Mississippi. Conclusions. . 172
CHAPTER IX.
JETTIES AND BREAKWATERS AT THE MOUTHS OF RIVERS.
Bars: fluvial, marine; origin of; favourable and unfavourable conditions.
Jetties at Outlets of Tideless Rivers free from Silt : Swinemünde, Vindau,
Dwina, Pernow, Chicago, Buffalo, and Oswego. Converging Breakwaters at
Mouths of Tidal Rivers: objects; Liffey, Merrimac, Tees, Tyne, Wear,
Nervion; Remarks. Training Jetties at Mouths of Tidal Rivers: objects
and instances ; Yare, Adur, Adour ; combined with training works on Maas
up to Rotterdam, and on Nervion up to Bilbao. Remarks on works at
Outlets; deepening aided by dredging; materials for jetties; alteration
of outlet ............. -°3
xii
Contents.
chapter x.
TIDAL FLOW IN RIVERS; AND FORMS OF ESTUARIES.
PAGE
Tidal Flow in a Creek; its effects. Tidal Flow and Fresh-water Discharge
in a River. Alluvium in a Tidal River ; its origin ; instances. Influence
of Fresh-water Discharge; advantages of large discharge; examples. Influ¬
ence of Tidal Flow ; importance of tidal volume ; contrast of state of tidal
and tideless rivers. Influences of Flood and Ebb Tides ; periods of greatest
velocity; divergence indirections. Conditions affecting Tidal Flow ; large
tidal rise ; small fall of river ; examples. High-water Line ; general uni¬
formity ; instances of rise ; irregularities. Lowering of Low-water Line ;
advantages ; indication of improvement. Progress of Tide up a River ;
indicated by simultaneous tidal curves ; acceleration of flood tide. Bore,
cause, instances. Obstructions to Tidal Flow, effects, disadvantages.
Effect of Tidal Flow on Outlet. Forms of Channels and Estuaries;
best theoretical form ; varieties in forms of estuaries. General Principles
relating to Tidal Rivers . 229
chapter xi.
DREDGING IN TIDAL RIVERS.
Advantages of Removal of Hard Shoals, and Dredging. River Clyde :—Original
Condition ; Removal of Hard Shoals ; Jetties and Training Walls ; Deepen¬
ing by Dredging, and Blasting Reef ; Extent of Improvements effected, and
Increase of Trade of Port; Need of Maintenance of Depth; Cessation
of FToods. River Tyne :—Original Condition ; Jetties and Training Walls ;
State in i860; Piers at Mouth ; Dredging; Regulation Works ; Improved
State ; Influence on Traffic ; Remarks. River Tees :—Condition ; Cuts and
Jetties ; Training Walls ; Dredging; Blasting Reef ; Breakwaters at Mouth ;
Improvements in Condition ; Remarks. River Usk :—Description ; Defects
in Channel up to Newport; Improvements proposed. River Mersey:—
Description of Estuary ; Influence of Inner Estuary on Outlet ; Objections
to Training Works in Inner Estuary; Dredging on Bar. Concluding
Remarks 250
chapter xii.
TRAINING WORKS IN ESTUARIES.
Simplicity of Regulation Works compared with Training Works in Estuaries.
Influences of Training Works in Estuaries; Improvement of Trained
Channel ; Concentration of Scour ; Accretion promoted outside Channel.
Training Works in the Wash:—Condition of Rivers draining Fens;
River Witham, Training Works, New Outlet Channel, Effects of Works ;
River Welland, Training Works, Necessity for Prolongation ; River Nene,
Straight Cut, its Prolongation, Training Works, Need of Extension ; River
Ouse, Eau Brink Cut, Straight Cut below King's Lynn, Training Works
in Estuary ; Remarks on Improvement of Outfall of these Rivers. River
Dee Training Works :—Original Condition of Estuary; Training Works in
Estuary, Reclamation the Main Object, Unfavourable Line adopted, Effects
Contents.
xiii
of Works ; Remarks on Works, their Injurious Effect on Estuary. Ribble
Estuary Training Works:—Former Condition, Changes in Channels and
Outlet ; Training Works, their Extension, and Improvements in Channel ;
Works in Progress and Proposed ; Effects of Works on Condition of Estuary ;
Remarks. Weser Training Works :—Description of River and Estuary ;
Training Works in Tidal River and Estuary ; Results ; Remarks on Improve¬
ments and Prospects 281
CHAPTER XIII.
training works in estuaries (continued).
Physical Conditions of the Loire and the Seine contrasted. Loire Estuary
Works:—Description of Estuary; Training Works below Nantes, their
extension and results; Ship-Canal formed alongside part of Estuary to
provide a better navigable depth ; Dredging in Estuary, and in Trained
Channel ; Improvement of Navigable Depth up to Nantes ; Remarks on
Loire Training Works, difficulties involved. Seine Estuary Works:—
Original state of Tidal Seine ; Training Works in Estuary, their extent and
consolidation ; Effect of Training Works on Navigable Channel ; Accretions
in Seine Estuary, their origin, extent, and volume; Alterations in Lines
of Depths at Outlet; Tancarville Canal, description and object of Canal ;
Schemes for Completion of Seine Training Works, difficulties, variety of
proposals, their aims and merits ; Remarks on Seine Training Works, value
for Rouen, cause of anxiety for Havre ; Importance of Completion of
Works, in trumpet-shaped form aided by dredging. Remarks on Training
Works :—General Conclusions deduced from Experience ; Enlargement of
Trained Channel, indicated by instances. Concluding Remarks . . 304
CHAPTER XIV.
experimental investigations on training works in
estuaries.
Origin of Investigations. Advantages of experiments with working Models of
Estuaries. Material employed for Bed of Estuary. Arrangement of Models;
Method of producing Tidal Flow and Fresh-water Discharge, and Formation
of Training Walls. Suitability of Seine Estuary for instituting Investigations
on Training Walls. Experiments with Model of Tidal Seine :—Description
of Model ; Resemblance of Model to actual Estuary ; Changes produced on
inserting Training Works ; Effects produced on inserting various Schemes for
prolonging Training Works, namely, contracted outlet, sinuous channel, and
trumpet-shaped channel ; Comparison of Results of Schemes. Large Rouen
Model described. Experiments with Model of Mersey Estuary :—Contrast
to Seine Estuary ; Description of Model ; Resemblance of Channels in
Model to those in Estuary ; Transformation produced by inserting Training
Walls in Inner Estuary ; Improvement effected by Outer Training Works.
Concluding Remarks, reliability of method, its great value for preliminary
investigations 327
LIST OF ILLUSTRATIONS IN VOL. I
RIVERS.
PLATE I.
HYDROLOGY OF RIVERS.
River Seine Basin. Distribution of Rainfall.
River Seine Basin. Tributary Basins, and Nature of Strata.
Diagram showing Daily Height of Seine at Poissy. 1878-1881.
Diagrams showing Daily Height of Seine at Bezons. 1881-1890.
Flood Diagrams. November, December, and January, 1882-1883.—Nogent-sur-
Seine.—Brenne at Montbard.—Yonne at Sens.—Seine at Montereau.—
Marne at St. Dizier.—Marne at Chalifert.—Seine at Paris.—Aisne at
Pontavert.—Oise at Venette.—Seine at Mantes.
PLATE II.
DREDGING AND EXCAVATING MACHINES.
Bucket-Ladder Hopper Dredger. Longitudinal Section.
Grab Bucket Hopper Dredger. Longitudinal Section and Cross-Sections.
Semi-Cylindrical Grab Bucket.
Three-Bladed Grab Bucket.
Grab Bucket Hydraulic Dredger.
Dredger with Long Shoot.
Sand-Pump Dredger.
Steam Navvy.
Rock Breaker. Cross-Section.
PLATE III.
REGULATION WORKS : FIXED AND DRAW-DOOR WEIRS.
REGULATION WORKS.
Variation in Depths of Winding Biver.
Biver Bhine. Regulation Works near Schierstein.
Biver Eibe. Regulation Works near Grunewalde.
Biver Niemen. Regulation Works near the River Jura.
Biver Blione. Regulation Works near St. Alban, and St. Estève.
xvi
List of Illustrations.
FIXED WEIRS.
Weir ou Hiver Severn. Plan and Section.
La Temple Weir. River Lot, France. Plan and Section.
Fossat Weir. River Lot, France. Plan and Section.
Dehree Anicut. River Sone, India. Section.
Godavery Anicut. River Godavery, India. Section.
DRAW-DOOR WEIRS.
Teddington Weir. River Thames. Section and Elevation.
Sluiee-Gates with Free Hollers. Plan and Section.
La Monnaie Weir. River Seine, Paris. Section.
PLATE IV.
I.OCKS AND MOVABLE WEIRS.
Upper Seine. Paris to Montereau. Longitudinal Section.
Lower Seine. Rouen to Paris. Longitudinal Section.
Port-à-L'Anglais Lock and Weir. River Seine. Plan.
Poses Locks and Weir. River Seine. Plan.
Port of Frankfort. River Main. Plan.
Hiver Main. Mainz to Frankfort. Longitudinal Section.
Martot Weir. Needle Weir, River Seine. Section.
Suresnes Weir. Navigable Pass, Frame with Panels, River Seine. Section.
Port Villez Weir. Navigable Pass, Frame with Curtains, River Seine. Section.
Port-à-L'Anglais Weir. Shutter Weir, River Seine. Section.
Poses Weir. Navigable Pass, River Seine. Section.
Hydraulic Shutter Weir. River Yonne, France. Section.
Sone Shutter Weir. Sone Canal, India. Section.
Throstle-Nest Weir. River Irwell. Section.
Bear-Trap Weir. La Neuville, River Marne, France. Section.
Charlottenburg Drum Weir. River Spree, Germany. Section.
PLATE V.
OUTLETS OF RIVERS FORMING DELTAS.
Mouth of the Hhone. Plan.
Delta of the Danube. Plan.
Sulina Mouth of Danube. Plan.—North Pier. Section.
Delta of the Mississippi. Plan.
Mississippi South Pass Jetties. Plan and Section.
Mississippi South Pass Outlet Channel. Longitudinal Sections.
Hhone Bar. Sections.
Sulina Mouth of Danube Bar. Sections.
List of Illustrations.
xvii
PLATE VI.
WORKS AT MOUTHS OF TIDAL RIVERS.
River Liffey. O'Connell Bridge to Dublin Bay. Plan and Longitudinal
Sections.—North Breakwater. Section.
Kiver Yare. Breydon Water to Mouth. Plan and Longitudinal Section.
River Wear. Sunderland Bridge to the North Sea. Plan and Longitudinal
Sections.—Roker Pier and New South Pier. Sections.
River Adour Outlet Channel. Plan and Longitudinal Sections.
River Nervion. Confluence of River Asua to Bay of Biscay. Plan and
Longitudinal Sections.—Jetty. Section.—Breakwater. Section.
River Maas Outlet Channel. Plan and Longitudinal Sections.—North Jetty.
Section.
River Maas. Lower River, and Old Outlet. Plan.
PLATE VII.
TIDAL DIAGRAMS, AND ESTUARIES.
TIDAL DIAGRAMS.
Simultaneous Tidal Curves, River Seine. Spring Tides.
Tidal Diagram, River Seine. Spring Tides.
Simultaneous Tidal Curves, River Weser. Average Tides.
Tidal Diagram, River Weser. Average Tides.
Simultaneous Tidal Curves, River Mersey. Equinoctial Spring Tides.
Tidal Diagram, River Mersey. Equinoctial Spring Tides.
ESTUARIES.
River Thames. Gravesend to the Nore. Plan and Longitudinal Section.
River Humber. Hull to Spurn Light-Ship. Plan and Longitudinal Sections
River Mersey. Runcorn to the Bar. Plan and Longitudinal Sections.
PLATE VIII.
TIDAL RIVER IMPROVEMENTS.
River Tyne. Scotswood Railway Bridge to North Sea. Plan and Longitudinal
Sections.—North Pier. Section.
River Clyde. Glasgow Bridge to Greenock. Plan and Longitudinal Sections.
River Tees. Stockton Bridge to the Fairway Buoy. Plan and Longitudinal
Sections.—South Gare Breakwater. Section.
River Usk. Newport Bridge to Mouth. Plan and Longitudinal Section.
PLATE IX.
TRAINING WORKS IN ESTUARIES.
River Seine. Quillebeuf to Havre. Plan and Longitudinal Sections.—
Training Wall as reconstructed. Section.
River Loire. Nantes to the Bay of Biscay. Plan and Longitudinal Sections.
—Training Wall. Section.
VOL. I. b
xviii
List of Illustrations.
River Weser. Bremen to Mouth. Plan and Longitudinal Sections.—Training
Wall. Section.
River Ribble. Preston Dock to the Irish Sea. Plan and Longitudinal Sections.
River Dee. Chester to the Irish Sea. Plan and Longitudinal Sections.
PLATE X.
EXPERIMENTS ON TRAINING WORKS.
Seine Estuary. H. L. Partiot's Scheme, 1859.—MM. De Coene and Le
Brnn's Scheme, 1888.—E. Lavoinne's Scheme, 1884.—L. L. Vauthier's Scheme,
1888.—Administration Scheme, 1887.—L. F. Vernon-Harcourt's Scheme,
1888.
Mersey Estuary. Model Chart of Mersey.—Manchester Ship-Canal Scheme,
1884.—Training Works in Outer Estuary.
LIST OF WOODCUTS IN VOL. I
HO. PAG h
j. Section of the Valley of a River on Impermeable Strata . . . . 14
2. Section of the Valley of a River on Permeable Strata . . . • J4
3. Fall of Rivers ........ . . • • 15
4. Sharp Bend of a River. Plan and Section . . . . . .18
5. Section of Channel .22
6. Double Float ............ -6
7. Screw Current-Meter .......... 28
8. Vane Current-Meter. Elevation and enlarged Section . . . 29
9. Gauge-Tubes 31
10. Bag and Spoon 72
11. Lock. Sectional Elevation and Plan . . . . . . .101
12. Lock-Gates. Plan 105
13. Jetties at Swine Mouth of River Oder. Plan ...... 206
14. Jetties at Mouth of River Pernow. Plan ....... 207
ERRATUM.
Page 128, line 16 from top,for Moskowa read Moskva.
RIVERS AND CANALS.
PAßT I. ßlVEßS.
CHAPTER I.
PHYSICAL CHARACTERISTICS.
Introduction. River Basins. Differences of Rainfall. Variations in Rainfall.
Evaporation and Percolation. Influences of Forests and Vegetation. Effect of
Period of Rainfall. Impermeable and Permeable Strata. Available Rainfall.
Forms of River Valleys. Fall of Rivers. Transportation of Material. Influences
of Lakes. Divergence of Current. Tidal, and Tideless River Outlets. Importance
of Scientific Study of Rivers.
RIVERS form a natural and easy means of communication
between the sea and the interior of a country, and afford safe
and convenient roadsteads for vessels. They also furnish the
chief sources of water-supply ; and the most fertile districts
are situated along their banks. Consequently most of the
important cities of the world have been built on the banks
of rivers. Rivers, however, are not always suitable foi-
navigation, in their natural condition, even in the lower portions
of their course ; and, owing to the continual changes taking
place in their channels and at their outlets, they are liable
to deteriorate if left to themselves. Moreover, rivers, whilst
serving as the main arteries for the drainage of a country, and
proving most valuable for irrigating lands in hot countries
during the dry season, and as a source of water-power, are
liable to devastate their valleys by extensive floods during
periods of excessive rainfall. According^, the regulation,
improvement, and control of rivers constitute one of the most
B
2 Influences of Physical Conditions. [chap.
important, and at the same time one of the most difficult
branches of civil engineering.
The natural state of any river depends upon a variety of
physical conditions, such as the area of its basin ; the amount,
distribution, and period of the rainfall over this drainage area ;
the nature of the superficial strata of its basin, and the extent
to which the surface is covered by vegetation ; the slopes of
the valleys of the main river and its tributaries ; the inclination
of the river-bed, and the material of which it is composed ;
and the amount of detritus brought down by the current.
On approaching the sea, a river is, in addition, affected by
the tidal rise in the sea at its outlet, by the size and form
of its estuary, by the sediment introduced by the flood tide,
by the direction of the prevailing winds, the travel of material
along the coast and the heaping-up action of the waves,
and by the tendency of the material brought down by the
river to deposit at its mouth in tideless seas.
River Basins. The basin of a river is the whole area drained
by the river, or the area over which the rainfall tends to flow
into the river or its tributaries. The boundary line of a
river basin is the ridge along the summit of the valleys of
the river and its tributaries, on each side of which the rainfall
flows in opposite directions into two different river basins,
just as the rain runs into different gutters on either side of
the ridge of a roof. This boundary line between adjacent
basins is termed the water-parting, from the separation of
the flow on each side of it.
The size of a river, beyond the reach of tidal influences,
is naturally in some measure proportionate to the area which
it drains ; whilst the area of a river basin is dependent on the
extent of the continent, and its general configuration. Large
river basins are only found in extensive continents, where
plains stretch for long distances inland from the coast. The
insular position, and small extent of Great Britain preclude
it from possessing large rivers ; and even the river basins of
i.] Areas of Basins and Lengths of Rivers. 3
the European continent appear somewhat insignificant when
compared with those of the vast continents of Asia, Africa,
and America.
Amongst English rivers, the Thames comes first with
a basin of 5,244 square miles. Next in importance is the
Severn, with a basin of 4,350 square miles, which is followed
by the Yorkshire Ouse, with a basin of 4,133 square miles;
whereas the Dee, the Mersey, and the Ribble, though
possessing large tidal estuaries, drain areas of only 813, 1,724,
and 815 square miles respectively. The Tyne, the Tees,
and the Clyde, also, though affording deep-water access
to flourishing seaports, owing to the improvements effected
in their tidal channels, have basins of merely 1,130, 735, and
945 square miles.
Turning to the principal rivers of Europe, draining far
more extensive tracts of lands, the Seine has a basin of 30,370
square miles, about six times the drainage area of the Thames
(Plate 1, Fig. 1) ; the Rhone has a basin of about 38,000 square
miles ; and the Loire, a basin of 45,000 square miles ; whilst
the lengths of the main stream of these rivers exceed 400
miles. Proceeding further into the interior of Europe, con¬
siderably larger river basins are met with ; thus, for instance,
the Elbe has a drainage area of 55,000 square miles, and
a length of 700 miles ; the Rhine has a basin of 75,000 square
miles, and a length of 800 miles ; the Don and the Dneiper
have basins of 170,000, and 245,000 square miles respectively,
with lengths of about 1,300 miles; the Danube has a basin
of 312,000 square miles, with a length of 1,700 miles; and
the Volga has a basin of 563,000 square miles, with a length
of about 2,000 miles.
Considerable uncertainty exists as to the exact drainage
areas of the rivers in the larger, and less explored continents ;
but there are several which undoubtedly exceed the largest
river basins of Europe. Thus in Asia, whilst the Ganges,
with a basin of about 470,000- square miles and a length
B 2
4 Basins and Lengths of Largest Rivers, [chap.
of about 1,500 miles, and the Yang-tse-kiang, with a basin
of about 620,000 square miles and a length of 3,200 miles,
as well as the Hoang-ho or Yellow River and the Amur,
equal or exceed the largest rivers of Europe, they are surpassed
by the Yenisei with a basin estimated at 880,000 square
miles and a length of 3,700 miles, and the Lena draining
central Siberia, and having a basin of 940,000 square miles.
Moreover the largest river of Asia, the Obi, flowing through
western Siberia, and falling into the Arctic Ocean after a course
of 3,200 miles, has a drainage area of about 1.300,000 square
miles, more than double the extent of the largest European
river basin.
In Africa, the Nile is generally reputed to have the largest
basin, with an area estimated at 1,500,000 square miles, the
length of the river being supposed to exceed 4,000 miles.
The next largest river basins are the Congo, with an area
of 1,350,000 square miles, the Niger, 1,150,000 square miles,
the Zambesi, 850,000 square miles, and the Orange River,
400,000 square miles, the last of which alone comes within
the limits of the largest river basin of Europe.
South America derives its large river basins from its
peculiar configuration, with the lofty chain of the Andes
rising close along its western coast, and vast flat plains
stretching eastwards ; so that its rivers which flow into the
Atlantic Ocean nearly traverse the continent, having their
sources in the perennial snows of the western mountain range.
Accordingly, the Amazon, flowing across the broadest part
of the continent, draining an area of about 2,250,000 square
miles, and having a length of nearly 4,000 miles, possesses
the largest basin in the world ; whilst the La Plata basin,
with an estimated area of 1,600,000 square miles, exceeds
in extent any river basin in the eastern hemisphere. The
Mississippi is by far the largest river of North America,
having a basin of 1,244,000 square miles, four times that
of the Danube, and a length of 4,200 miles ; whilst the
i.] Extreme Differences of Rainfall. 5
St. Lawrence has a basin about equal in extent to the Danube
basin.
From the foregoing statistics, it will be seen that drainage
areas may vary from the small gullies of streams rising close
to the sea-coast—like Blackgang and Shanklin chines, in the
Isle of Wight—up to the vast basins of mighty rivers, extend¬
ing over great portions of the largest continents of the globe.
Differences of Rainfall. As rivers originate from rainfall,
and form the channels by which the surplus rainfall flowing
off the land is conveyed to the sea, observations concerning
the rainfall over the basin of a river are very valuable in
studying the régime of any river. Generally the rainfall is
greatest in mountainous districts, especially where the hills
rise near the coast, and where the prevalent winds blow from
the ocean. These warm winds arriving fully laden with
the moisture collected in their passage over the ocean, are
arrested by the high land, and having their temperature
reduced in rising up the mountain slopes, deposit their
moisture as rain in passing over the mountain range. On
the contrary, rainless districts are found in the interior of
large continents, cut off from the sea by intervening mountain
chains which deprive the air of its moisture before it reaches
these regions. Accordingly, very great differences in rainfall
are experienced in various parts of the world, ranging from
zero over the deserts of the centre of Asia and Africa, up
to an average annual rainfall of 474 inches at Cherrapunji
in Assam, on the Khasi Hills, 4,455 feet above sea-level1,
the station having by far the heaviest recorded rainfall in
the world.
Sometimes the amount of rainfall varies greatly in different
parts of the same country. Thus whilst India possesses
in its north-eastern corner the rainiest station known, and
Uttray Mullay in Southern India has an average annual
rainfall of 263 inches, the hill station of Leh to the north
1 ' The Climates and Weather of India, Ceylon, and Burmah.' H. F. Blanford.
6 Instances of Differences in Rainfall, [chap.
of the Himalayas has an average rainfall of only 2-7 inches,
and Jacobabad in Central India an average fall of only 4-4
inches in the year. Mexico and Peru also have very rainy,
and rainless, districts only moderate distances apart. Even
in England, the average annual rainfall ranges between 22
inches at Ely, and 175 inches at The Stye in Cumberland.
Considerable differences in the rainfall may also be met with
within the limits of a single moderate-sized river basin,
varying with the elevation and with the proximity to the sea.
Thus the rainfall over the Seine basin, which is a fairly level
basin of moderate extent, ranges from under 24 inches up
to over 48 inches in a year (Plate 1, Fig. 1) ; whilst in the
Severn basin, which extends into the mountainous districts
of North Wales, the average annual rainfall ranges from
under 30 inches up to over 80 inches. Accordingly, the
various portions of a river basin may be subject to very
different meteorological influences, affecting the flow of the
tributaries draining them ; and the flow of the main river
consists of a combination of these various influences, which
are successively imparted to it at the confluence of its several
tributaries. A plan, therefore, of a river basin, showing the
main river and its tributaries, and also indicating the different
average amounts of rainfall over this area, enables some
idea to be formed of the comparative influence of the rainfall
on the various tributaries, and. indicates the places at which
the flow of the main river is thereby affected (Plate 1, Fig. j).
Variations in Rainfall. The amount of rainfall in the
same locality varies from year to year ; and a reliable average
can only be obtained from observations extending over several
years. In some years, the rainfall is much below the average,
leading to a deficiency of water for navigation in the dry
season ; in other years, the rainfall is unusually large, causing
the rivers to overflow their banks and inundate large tracts
of land. Thus, in the exceptionally dry year 1887, the
rainfall over the British Isles was only two-thirds of the
i.] Yearly Fluctuations of Rainfall. 7
average; whilst in the very wet year 1872, the rainfall ex¬
ceeded the average by 36 per cent., and in 1877 by 25 per
cent1 The rainfall of very wet years exhibits generally a
greater deviation from the average than in very dry years ;
and the proportionate variation of the rainfall is greater in
dry districts than in wet ones. Taking the driest and wettest
stations in England where records are kept, the rainfall was
14-63 inches at Ely, and 130-90 inches at The Stye in 1887,
and 27-20 inches, and 243-98 inches respectively in 1872.
The maxima and minima, however, do not occur at every
station in the wettest and driest years, for the rainfall at Ely
was 29-03 in 1877, or nearly 2 inches more than in 1872;
whilst some stations experienced more rainfall in 1887 than
in 1884, the excess amounting to nearly 2 inches at the
very dry station of Hunstanton.
The yearly fluctuations of rainfall are greater in tropical
countries than in the temperate zones, especially in places
where the rainfall is moderate ; for whereas the rainfall
of 1872 only attained about three times the rainfall of 1887
in two or three places in England, the maximum rainfall in
India and Burmah amounts in many parts to three or four
times the minimum rainfall, and even considerably more at
a few places.
The rainfall, moreover, varies at different seasons of the
year ; and the periods during which the maxima and minima
occur depend on the locality, for the rainfall is due to the
prevalence of the special winds blowing from the ocean, to
which the locality may be most exposed. There are
generally a wet and a dry season, extending over well-defined
periods, within the tropics, and in places visited by period¬
ical winds such as the monsoons ; but in temperate climates
where the winds are more variable, there is not a similar
regularity in the rainy periods. Nevertheless, extended
observations indicate that in temperate regions, districts
1 ' British Rainfall, 1887.' J- Symons.
8
Evaporation and Percolation. [chap.
possess, on the average, a special period of maximum rain¬
fall, differing according to the period at which the wettest
winds generally prevail, and the position of the district in
relation to the ocean. Even in the British Isles, the ordinary
period of maximum rainfall varies according to the situa¬
tion ; for it occurs in December and January on the extreme
western coasts, it is in August in the eastern midlands of
England, whilst over the rest of Great Britain, October is
the wettest month. There are similar differences also in
the periods of minimum rainfall.
Evaporation and Percolation. The rain falling on a river
basin does not all find its way into the main river which
drains the basin and discharges into the sea. Some of the
rainfall is drawn up again into the air by the evaporation
produced by the sun and wind ; a portion sinks into the
soil and affords moisture to trees and plants ; and a portion
fills subterranean cavities which form the sources of springs.
Evaporation varies with the locality, the season of the
year, the state of the weather and the wind, and the nature
of the soil on which the rain falls : it is greater from a
surface of water than from the ground in summer, and from
the ground in winter. The effects of evaporation over
extensive areas of water are illustrated most forcibly by the
state of the Caspian Sea, which, though it receives the
waters of the Volga and other large rivers, and has no out¬
let, has had its water-level gradually reduced by the excess
of evaporation over the combined discharges of these rivers,
so that it is now about 85 feet below the level of the Black
Sea. The waters of the Jordan also fail to raise the level
of the Dead Sea, which lies in a great depression with its
water-level 1,300 feet below the Mediterranean ; whilst the
discharge from the 750,000 square miles draining into Lake
Chad is entirely absorbed by evaporation.
Evaporation attains its maximum during hot, dry weather,
immediately succeeding rain on flat impervious ground, or
i.] Influences of Forests on Rivers.
9
from a sheet of water; and it is least in cold, damp
weather.
Percolation depends on the nature of the soil, and the
weather ; it is small in amount through earth, but large
through sand ; and no percolation takes place in warm sum¬
mer weather. The greatest amount of percolation occurs
with melting snow, and especially with small falls of snow
followed by a thaw. Water penetrating into the ground
to a depth of two or three feet appears to be removed from
the influence of evaporation.
Influences of Forests and. Vegetation. A covering of trees
and grass is very beneficial in increasing the average dis¬
charge of a river by reducing evaporation, especially in the
summer months when water is most needed, and in equalizing
its flow during the rainy season, especially in the steeper
parts of the basin, by retarding the arrival of the rain to
the watercourses.
The clearing of forests tends to reduce the rainfall of a
district; but the main cause why forests and grass increase
the average flow of rivers is their arrest of evaporation, in
some instances the evaporation in the open country having
been found five times as great as the evaporation under
trees. The shelter afforded by the foliage prevents some
of the rain from reaching the ground, but this is much more
than compensated for by the smallness of the evaporation ;
so that the effective rainfall in a forest is considerably
greater than in the open country, independently of the
larger rainfall over a forest, the proportion of which would
be difficult to determine. Forests and vegetation are
specially valuable on impermeable strata and on moun¬
tain sides, where, by regulating the downward flow of the
rain, and by binding together the loose soil, they diminish
the rapid rise of torrential streams, arrest the washing away
of the soil, and prevent the formation of gullets which
facilitate the descent of the water. Extensive clearings of
io Effects of Summer and Winter Rains, [chap.
forests, and close cropping of pastures have led to such
injurious denudation and floods in mountainous regions,
that it has been found necessary to enforce the replanting
of trees and the promotion of the growth of vegetation,
which have produced beneficial results.
Effect of Period of Rainfall. The total amount of rain
falling over the area of a river basin, in any given period,
is not necessarily any measure of the discharge of the river.
The rainfall over the Seine basin is greater in summer
than in winter ; but a glance at the diagrams showing the
daily heights of the Seine from 1878 to 1890 (Plate 1, Figs.
3 to 6), indicates that large floods never occur in the river
between May and October, whilst they are very frequent in
the other half of the year.
As evaporation is very active in the summer, most of the
rain which falls during the summer months is returned to
the atmosphere, and only a small proportion finds its way
into the rivers, unless the slope on which it falls is very
steep, the strata of the basin impermeable, and the rain¬
fall very great in a short period. Winter rains, on the
contrary, find their way almost entirely into the river which
drains the basin on which they descend, as they fall upon
ground saturated by the wet of autumn, and when evaporation
is nearly suspended. A continuous, or very heavy rainfall
produces a greater effect on the discharge of a river than
the same amount of rain distributed over a longer period.
In many cases the greatest river floods are produced when
a heavy rainfall takes place on melting snow.
As summer rains, in temperate regions, produce much
less effect on the flow of rivers than rains in the winter
and spring, a drought in the early autumn may often
be predicted at the commencement of the summer, if
the rainfall has been small in the preceding winter and
spring.
In considering the flow of rivers within the temperate
i.] Influences of Seasons and Strata on Rivers, n
zones, the year may be conveniently divided into the warm
and cold seasons, extending, in the northern hemisphere,
from May to October, and from November to April re¬
spectively, floods being generally only prevalent during the
cold season. (Plate i, Figs. 3 to 6.) An exception, however,
in the period of the occurrence of floods, must be noted in
the case of rivers originating from glaciers and snow-clad
mountain heights ; for the melting of the ice and snow by
the summer heat produces floods in these rivers, whereas
the cold of winter reduces their flow to a minimum. The
natural divisions of the year in respect of rivers in tropical
countries are the dry and rainy seasons ; for the rivers are
nearly dried up towards the close of the first period, and
are immoderately swollen by the periodical rains.
Impermeable and Permeable Strata. The nature of the
strata forming the basin of a river and its tributaries exerts
a very marked influence on its flow.
Where the strata are impermeable, the rain falling over
the basin is rapidly discharged by numerous little rivulets
into the main stream, and the river is torrential in its char¬
acter. It rises very rapidly in rainy weather, and its fall
is also rapid ; and in continued dry weather it becomes more
or less dried up. As, however, no strata are absolutely im¬
permeable, some of the rain sinks into the fissures in the
ground, and is delayed in its flow to the river. Accordingly,
the rise of a torrential river is always more rapid than its
fall, which is delayed by the later arrival of the rain that
has penetrated below the surface. These characteristics of
torrential rivers are illustrated by the flood diagrams of the
Brenne, the Yonne, and the Marne (Plate 1, Figs. 8, 9, and
11), tributaries of the Seine in the steeper, impermeable,
and rainiest portion of its basin.
Where the strata are permeable, the rain sinks into the
ground, and only finds its way gradually into the main
stream. Consequently the rise of a river ti'aversing per-
i2 Mixed Character of Floods of Rivers, [chap.
meable strata is more regular than that of a torrential river.
The floods, however, of rivers flowing over permeable strata
are more injurious to the adjacent lands, owing to their longer
continuance, than the more rapidly rising floods of torrential
streams which quickly subside. The flood diagrams of the
Little Seine and the Aisne (Plate i, Figs. 7 and 14), exhibit
those properties of gently-flowing rivers draining permeable
strata, and present a marked contrast to torrential floods.
The basins of large rivers generally comprise several
kinds of strata, possessing different degrees of permeability,
so that some of the tributaries receive chiefly the drainage
of impermeable strata and others of permeable soils, of
which the tributaries of the Seine just referred to are in¬
stances (Plate i, Fig. 2). The principal river possesses,
in such cases, mixed characteristics, which depend on the
proportion of permeable and impermeable strata over the
basin, and may vary in different parts of its course, and
even at different periods according as the rainfall predom¬
inates over one or other portion of the basin. This result is
exemplified by comparing the rise of the same flood in the
Seine above and below the confluence of the Yonne, and of the
Marne, where the influence of the floods of these rivers on
the floods of the Seine is clearly manifested (Plate 1, Figs. 7,
10, and 13). An advantage of a diversity in the strata of the
tributary basins consists in the ordinary absence of coincidence
in the arrival of the floods from the several tributaries at
any point of the main river, owing to the rapid propagation
and subsidence of torrential floods, which often have passed
off before the floods from the gently-flowing tributaries have
reached their maximum. Accordingly, an exceptional flood
is more likely to result from a casual coincidence in the
arrival of the floods from the various tributaries, than merely
from a heavy rainfall over a basin comprising various
strata.
Available Rainfall. If a relation could be established
i.] Proportions of Available Rainfall. 13
between the volume of water which finds its way into
a river, and the rainfall over the river basin, for strata of
different permeability, it would be only necessary to ascertain
the area of the basin, and the average nature of the super¬
ficial strata,, in order to deduce the discharge of the river
for any given rainfall. This proportion is specially affected
by the degree of permeability of the strata forming the
basin ; for whereas the rain falling on an impermeable
stratum finds its way rapidly into the river, when falling
on a permeable stratum it sinks into the soil, and besides
affording a considerable amount of moisture to vegetation,
it supplies springs which either do not flow into the river at
all, or in many cases join it at lower parts of its course.
The discharge of the River Po, draining an impermeable
basin, has been found to amount to 75 per cent, of the
rainfall ; whilst the discharge of the Garonne above Mar-
mande, and of streams draining clay beds, has been estimated
at 65 per cent, of the rainfall. In the case of the Saône,
with a less impervious basin, the discharge is equivalent to
half of the rainfall. In the Upper Seine basin, however,
where the permeable strata predominate (Plate 1, Fig. 2),
the discharge does not quite amount to a quarter of the
rainfall ; whereas in the almost wholly permeable basin of
the Eure, the discharge is only 15! per cent, of the rainfall.
Accordingly, it appears from these observations, that the
available rainfall, or the proportion of the rainfall which
actually reaches a river, may vary from 75 per cent, on
impermeable strata, down to 15 per cent, on very permeable
soils. These proportions, however, are merely yearly averages,
■ and would be much less in the warm season, and greater
in the cold season.
Forms of River Valleys. As a river forms its own
bed in the bottom of the valley through which it flows,
the nature of the bed depends on the geological strata
which lie on the surface of the valley, or which the river
14
Forms of Valleys of Rivers, [chap.
has laid bare in the course of ages. The general form of
the bottom of a river valley varies according as the strata
are impermeable or permeable. The water flowing down the
slopes of valleys composed of impermeable strata gradually
wears away the projections, and fills up the hollows with
soil from above, so that the slopes become curved as indi¬
cated in the sketch (Fig. i ). The rain, however, falling on
Section of the Valley of a River on Impermeable Strata.
Fig. i.
a permeable river basin, as it sinks into the soil, does not
materially modify the form of the slopes. Accordingly the
valley, in this case, generally retains its original form, being
nearly level at the base, or sometimes slightly raised along
the banks of the river in consequence of the successive
deposits produced by the river in flood-time, as shown on
the accompanying sketch (Fig. 2).
Section of the Valley of a River on Permeable Strata.
Fig. 2.
Fall of Rivers. The fall of the bed of a river corresponds
approximately to the general fall of the land over which it
flows, and varies considerably in different localities. Usually
the fall is greatest near the source of the river, gradually
diminishing as it approaches the sea (Fig. 3) ; but this
reduction in fall is not always continuous, for sometimes
the fall increases again after the river has traversed a flat
Fall of River-Beds.
table-land, or passed through a lake, as indicated by the
diagram. If a river valley is terminated by a high mountain
range, the river commences as a torrent, with a more or
less rapid fall according to the declivity of the mountain
side. If the mountains rise near the sea into which the
river flows, the average fall of the river is considerable, as
on the western side of America. When, on the contrary,
the mountain ranges are separated from the ocean by vast
plains, as on the eastern side of the Rocky Mountains and
the Andes, the general fall of the river is gentle on emerging
from the mountains.
Transportation of Material. Rivers, besides conveying
rainfall to the sea, are also the transporters of alluvium.
Gradually, though slowly, all rivers are bringing detritus
down their valleys ; and the materials of the mountains are
being carried to the sea. Glaciers, in their slow but irresistible
descent, grind the rocks of the mountain slopes down which
they imperceptibly glide, and when melting in the valleys
carry a large quantity of matter along with them. The
slopes of the hills are also disintegrated by frost and rain,
Fall of Rivers.
Fig- 3-
i6 Transport of Alluvium by Rivers. [chap.
and the débris are brought down by the watercourses in
flood-time. The velocity of the flow of rivers in the earlier
portion of their course enables them to hold this matter in
suspension, or to roll it along their bed ; but as the fall
of their bed, and consequently their velocity decreases, the
heavier particles fall to the bottom, or come to rest, and
tend to fill up the river-bed. The larger detritus is gradually
reduced to gravel, sand, and silt by long-continued attrition,
and is carried down the river in successive stages by floods, or
is strewn over the valley. This deposit of material specially
occurs when the velocity of the stream is abruptly checked
by a sudden reduction in fall, as for instance where the
River Po, on leaving the slopes of the Alps, descends into
the plains of Lombardy, or when the flow of a river is
arrested on entering a tideless sea.
The power of a current to transport material varies with
the velocity of the current, and depends upon the specific
gravity of the material, and the size of the particles. The
results of a variety of experiments indicate that a current
with a velocity of 3 to 6 inches per second will transport
silt and clay, of 8 inches per second will transport sand,
and of about 1 to feet per second will carry along gravel ;
whilst a current with a velocity of 6 feet per second will
roll stones along. As the current slackens, the materials
are sorted, the heaviest particles being deposited first.
Influences of Lakes. Lakes exercise a beneficial influence
on rivers which flow through them, by arresting the alluvium
brought down by the rivers, and by regulating their flow. The
current of a river is checked on encountering the inert waters
of the lake, which leads to the gradual deposit in the bottom
of the lake of the sediment with which the river is charged.
Thus the Rhone, on entering the lake of Geneva, deposits
the whole mass of detritus, which it has brought down
from the Alps, near the head of the lake, and issues from
the lake at Geneva as pure and clear as the waters of the
i.] Influences of Lakes on Rivers.
*7
lake ; and it presents a remarkable contrast to the turbid
Arve, which comes straight down from the Alps, and with
which its waters are mingled about a mile only below Geneva.
A river in flood flowing into a lake produces only a very slight
rise over the whole area of the lake, and is thus prevented
from rapidly swelling the discharge of the river below, which
has accordingly a much more uniform flow than it had
above the lake. Thus the Rhone and the Rhine are more
regular in their flow on issuing from the lakes of Geneva
and Constance, than in the upper portions of their course ;
and the river St. Lawrence, being fed almost entirely by
the lakes Superior, Huron, Michigan, Erie, and Ontario, is
so uniform in its flow that its level remains nearly always
constant.
Divergence of Current. The instability of a river channel
is not confined to the bottom of its bed. A river rarely flows,
especially across a flat, alluvial plain, in a straight and
uniform channel. A very slight impediment, such as a fallen
tree or a hard projection at one bank, or any irregularity
in the bed, will direct the main current against the opposite
bank, which, if composed of soft materials, is gradually
eroded ; and deposit collects near the other bank, so that
the course of the river is by degrees modified. After im¬
pinging against the concave part of one bank, the current
is turned again at the next bend to the other bank, thus
gradually producing and intensifying the serpentine course
so noticeable in many rivers when flowing through plains
(Fig. 4, p. 18). The greatest current runs close along the
concave bank, which, besides promoting the erosion of the
bank, produces also a scour at the bottom. Accordingly,
in winding rivers, the channel is deepest close to the concave
bank, and gradually shoals towards the convex bank, as
shown in the section ; so that the deepest part of the channel
shifts across the river between two successive bends. There
is always deep water along the concave bank, as the current,
C
i8 Vaviations of Depth in Winding Rivers, [chap.
in being turned round the bend, necessarily keeps close
to this bank ; whereas the current in tending across the river
at the change in curvature of the bank, is less confined,
and more variable in direction, so that the channel is shallower
between the bends (Plate 3, Fig. 1).
Fig. 4.
River Outlets. The mouths of rivers exhibit considerable
diversities, depending on the form, size, and general physical
conditions of the estuary which connects them with the sea,
and the rise of the tide in the sea into which they flow.
A broad distinction must be drawn between rivers up which
the tide flows and ebbs for some distance, and the size of
whose outlets depends mainly on the tidal ebb and flow, and
rivers discharging into tideless seas, whose outlets, like their
channels above, are wholly due to their fresh-water discharge.
Tidal rivers generally emerge into large estuaries or bays
before reaching the sea, affording large spaces for the influx
of the tide, as for instance the Thames, the Mersey, the
Humber (Plate 7, Figs. 7, 9, and 11), the Severn, the Seine
(Plate 9, Fig. 1), the Garonne, the Scheldt, the Elbe, the
Weser (Plate 9, Fig. 8), and the St. Lawrence. Tideless rivers,
on the contrary, form continually advancing deltas by the
i.] Tidal and Tideless Outlets of Rivers. 19
deposit of their alluvium in front of their outlets, on entering
the denser inert sea ; and they meander through the delta
they have heaped up in the sea, in several encumbered and
constantly lengthening channels, of which class of rivers the
Rhone, the Po, the Danube, the Volga, the Nile, and the
Mississippi are notable examples (Plate 5, Figs. 1, 2, and 5).
In tidal rivers, the sedimentary matter, which is partly carried
down from inland, and partly brought in by the flood tide
from the erosion of neighbouring coasts and sandbanks, is
kept in almost constant motion under ordinary conditions,
and is carried out to sea by the ebb tide ; whereas the
material accumulating at the mouths of tideless rivers is
entirely composed of detritus carried down by the river, and
is only partially removed by littoral currents and wave dis¬
turbance. The bar at the mouth of a tidal river is formed
by the heaping-up action of the sea on the beach, and the
travel of sand and shingle along the coast under the influence
of the prevailing winds, tending to close up the mouth of the
river ; but this barrier is to some extent lowered by the ebb
and flow of the tide, aided by the fresh-water discharge. The
bar in front of the outlet of a tideless river is due to the
deposit of alluvium, resulting from the slackening of the river
current on emerging from its channel into the sea.
Intermediate between these two distinct classes are some
rivers which, owing to the flatness of the alluvial land they
traverse on approaching the sea, and the small tidal rise at
their mouth, have formed several outlet channels, of which
the mouths of the Rhine and the Maas are instances ; and
there are others which, in consequence of their large flood
discharge, and the enormous quantities of sediment they bring
down, overpower the tidal flow, and form a delta, such as the
Ganges, the Irrawaddi, and the Orinoco.
Scientific Study of Rivers. The diversified physical con¬
ditions of rivers, as indicated by the foregoing considerations,
have frequently led engineers to express the opinion that each
c 2
2o Value of Scientific Study of Rivers.
river must be considered quite independently, without reference
to the experience gained on other rivers. Physicians might
with equal reason have declared that, as men are so different
in constitution, the treatment of each person should be con¬
sidered independently, in which case little progress would
have been effected in medical science. In investigating any
particular river, it is undoubtedly an engineer's duty to ex¬
amine, as far as practicable, the various physical characteristics
of the river ; but an intelligent investigation of any river
assists in determining the special influences of the several
physical conditions, and thus every fresh river studied affords
further insight into the bearing and effect of each special
condition. The relative importance of the several physical
conditions may, indeed, vary in every river ; but the natural
laws which regulate the influence of each condition remain
unaltered ; and the effects produced are the combined results
of all the influences, in their several proportions, in each special
case
Probably no river basin has been studied so carefully as
that of the Seine ; and the records furnished by Mr. Belgrand
and his successors 1 have enabled me to draw up a fairly
complete summary of the physical characteristics and hydro¬
logy of this river (Plate x), which may form a model of the
information desirable with respect to other river basins. Such
details, together with a longitudinal section and cross-sections
of the river, charts of the outlet at different periods, and
tidal diagrams, furnish the engineer with valuable material
in designing improvement works. The results of these works,
moreover, should be carefully followed, so that they may all
form onward steps in the science of river engineering.
1 ' La Seine, Études Hydrologiques,' E. Belgrand ; and ' Service Hydrométrique
du Bassin de la Seine.'
CHAPTER II.
MEASUREMENT AND FORMULAE OF DISCHARGE.
Importance of Discharge Measurements. Definition of Terms :—Discharge ;
Slope ; Sectional Area ; Wetted Perimeter ; Hydraulic Mean Depth ; Mean
Velocity. Measurement of Cross-Sections. Velocity Observations :—Floats ;
Current-Meters ; Gauge-Tubes; Hydrodynamometer. Comparative Values of
Methods for measuring Velocities. Distribution of Velocities in a Channel.
Mean Velocity deduced from Surface Velocity. Formulae of Discharge :—
Humphreys and Abbot's Formula; Darcy and Bazin's Formula; Ganguillet
and Kutter's Formula. Concluding Remarks.
The measurement of the discharge of a river is often
necessary previously to the execution of works of improve¬
ment, and is also essential to a complete knowledge of the
physical characteristics of a river. The flood discharge is
required in designing works for the mitigation of floods,
and the discharge at various periods is important for
arranging irrigation works ; whilst the low-water discharge
is also needed for ascertaining the navigable capabilities of
a river.
Measurement of discharge, moreover, forms the funda¬
mental step towards the establishment of any formula of
discharge, for calculating beforehand the discharge through
a deepened river or new cut, and along an irrigation or
drainage channel, so as to ascertain the effects of such works
before they are undertaken, and in order to regulate the size
22 Terms used in Discharge Measurements, [chap.
and slope of the channel so that it may pass the desired
volume of water.
Definition of Terms. The discharge of a river is the
volume of water which it passes into, or towards the sea
in a given unit of time. It is usually reckoned in cubic
feet per minute or second.
The slope or fall of a river is the inclination of its water
surface ; and it may be conveniently expressed either in feet
and inches per mile, or in terms of the actual ratio of the
fall to the length.
The sectional area of a river is the area of the cross-
section of its channel taken at right angles to the current,
and is expressed in square feet. When the actual, or possible
discharge of a river is to be measured, only that portion of the
cross-section is taken into account which is below the actual,
or assumed water-line.
The wetted perimeter is the portion of the boundary line
of the cross-section of the channel below the water-line.
The hydraulic mean depth, or, as it is often termed, the
hydraulic radius, is obtained by dividing the sectional area
of the stream by the wetted perimeter of the channel. For
instance, if a channel (Fig. 5) has a base of 20 feet,
with slopes on each side of if to 1, a depth of water
of 6 feet, and consequently a width at the surface of 38 feet,
the sectional area of the stream is 29 feet x 6 feet =174
square feet, and its hydraulic mean depth, or R,
-20:0" •*
Fig- 5
sectional area _ 174
= 4-18 feet.
wetted perimeter 41-6
h.] Methods of measuring Cross-Sections. 23
In fact, R is the height of the rectangle whose base is the
length of the wetted perimeter, and whose area equals the
sectional area of the stream.
The mean velocity is the average velocity of all the ele¬
ments of the current.
The discharge is obtained by multiplying the sectional
area of the stream by the mean velocity.
Measurement of Cross-Sections. To ascertain the dis¬
charge of a river, it is necessary, in the first place, to find its
sectional area. In gauging a river, as straight and uniform
a reach as possible should be chosen, and where the flow
appears to be regular, as this indicates that the section is
fairly uniform. To ensure accuracy, several cross-sections
should be taken along the part of the river selected. The
cross-sections are obtained by taking soundings, at known
intervals, in a straight line across the stream.
When the river is not very broad, a strong thin rope,
previously wetted and pulled out, tightly stretched at right
angles to the stream, from bank to bank, in the line of
the proposed cross-section, serves conveniently as the base¬
line for the soundings. Where great accuracy is required,
a wire should be employed in place of a rope. The positions
of the soundings are measured along the rope, starting from
the edge of the water on the right bank1. When the depth
does not exceed 15 feet, the soundings are most conveniently
taken by means of a round, slightly tapering, wooden pole,
on which the lengths are marked from the bottom upwards.
The pole must be terminated at the bottom by a broad flat
piece, so that it may not sink into mud ; and if the stream
is strong, a little lead may be inserted at the bottom of
the pole with advantage. The soundings are taken from
a boat kept with its head up stream and just under the
rope, so that the man taking the soundings may be able
1 The right and left banks of a river are considered to be to the right and left,
respectively, of a person when looking down stream.
24 A rrangements for measuring Cross-Sections, [chap.
to lower the pole in the line of the rope. When the river
is too deep, or the current too strong, for using a pole for
the soundings, a measuring chain with a weight at the
end is employed.
It is advisable, if possible, to select a time for taking the
cross-sections when the river is neither in flood nor very low.
In flood-time there is frequently considerable difficulty in
fixing the cord, in keeping the boat in position, and in taking
the soundings. When the river is very low, there is less
width of channel available for sounding, which is the easiest
and quickest way of taking the measurements in the bed
of the river, and there is liable to be too little water for
floating the boat over the shallows.
The level of the water at the various places chosen for
the cross-sections, and any alterations in level whilst the
soundings are being made, are obtained, either by means
of a spirit-level, or by reading off the height of the water
on marked gauge-boards set in the water to a known datum.
The measurements for the cross-sections are extended from
the edge of the water over the banks on each side, for
a certain distance, by levelling, so that the cross-sections may
be known for varying heights of the river.
When a river is too broad for stretching a rope or
wire across it, the line of the section is marked by
erecting a pole on each bank ; and a man is stationed at
one of the poles. Directly the boat conveying the
sounding party, slowly dropping down stream, crosses the
line, a signal is made by the man on the bank, and the
sounding is immediately taken, everything having been got
in readiness beforehand with the sounding chain lowered
nearly to the bottom. At the same instant, the exact
distance of the sounding from the banks is noted, either by
two observers, stationed with theodolites at the extremities
of a measured base-line on one of the banks, who take
the angles between the boat and the base-line, or by a person
h.] Measurement of Velocity of a Stream. 25
with a sextant in the boat taking the angles subtended by
known objects on shore. Any required number of soundings
can be thus taken along a particular line across the river,
at different distances from the banks, from which the cross-
section can be plotted.
Velocity Observations. Having obtained a sufficient
number of cross-sections, and deduced from them the mean
sectional area of the river or channel under examination, it
is necessary to ascertain the mean velocity of the current in
order to estimate the discharge of the river. There are three
general methods by which this velocity may be obtained :
i. Floats, 2. Ctirrent-Meters. 3. Gauge-Tabes.
i. Floats. The simple expedient of throwing a floating
substance into a stream, and observing the time it takes to
traverse a measured distance, for ascertaining the velocity
of the current, presents many difficulties in practical appli¬
cation. A floating body merely indicates the velocity of
the portion of the current in which it happens to be situated ;
whereas the velocity of a stream varies, not only at different
distances from the banks, but also at different depths in
the same vertical line.
For surface velocities, an orange forms an excellent float,
as its specific gravity is very little less than that of water,
and it therefore floats almost wholly immersed ; but pieces
of wood, cork, or other easily procurable floating substances
are often employed. Surface-floats, however, are useless in
windy weather, as wind has a great effect on the top layers
of water, so as even sometimes to make the surface water
move in an opposite direction to a feeble current below, as
observed by me during some float experiments in the Firth
of Clyde1.
1 Minutes of Proceedings Institution C.E., vol. lxxi, p. 50.
26 Velocity measured by Double Floats, [chap.
The double float is employed for measuring sub-surface
velocities at various points in a stream, so as to obtain the
mean velocity of the current. It consists of a small surface-
float connected by a cord to a large float carefully weighted,
so as to sink in the water and keep the connecting cord
stretched, but not enough to pull the surface-float under
water (Fig. 6). The length of the connecting cord is varied
according to the depth at which the velocity is to be measured ;
and whilst the current imparts its motion
to the lower float, this motion is ren¬
dered apparent by the corresponding
movement of the upper float.
The extensive series of experiments
on the flow of the Mississippi, by
Messrs. Humphreys and Abbot1, were
made with double floats. The lower
floats were kegs, from 9 to 12 inches
high, and 6 to 8 inches in diameter,
without top or bottom, and ballasted
with lead so as to sink and assume an
upright position. They were retained
at the desired depth by surface-floats,
to which they were attached by cords
of suitable length.
Double floats were also used in measuring the flow of
the Connecticut River2, as well as two types of current-
meters, by General T. G. Ellis.
The double float employed by Mr. Robert Gordon in
gauging the flow of the Irrawaddi, consisted of two wooden
cylinders, 6 inches in diameter : the lower float, 1 foot
long, was weighted with lead, so that the surface-float,
1 ' Report on the Physics and Hydraulics of the Mississippi River.' Captain
A. A. Humphreys and Lieutenant H. L. Abbot, 1861.
2 ' Report of the Surveys and Examinations of the Connecticut River.'
General T. G. Ellis. 'Report of the Chief of Engineers, U.S.A.,' 1875, ^art 2>
p. 305, and Plate 1.
Double Float.
Fig. 6.
it.] Tube-Floats, and Velocity-Rods. 27
i inch thick, to which the lower float was attached by a thin
cord, floated about three-fourths immersed 1.
In the Roorkee hydraulic experiments, conducted by Lieut.-
Col. Allan Cunningham on the Ganges Canal -, two patterns
of double floats were adopted for measuring sub-surface
velocities. A sphere of wood, 3 inches in diameter and
weighted with lead, was connected by a thin brass wire to
a pine disk, 3 inches square and I inch thick ; and a copper
spherical shell, if inches in diameter, weighted with lead,
was connected by a very fine silk thread to a cork disc,
1 inch square and | inch thick.
Tin tubes, 2 inches in diameter, loaded with lead so as to
float vertically, nearly reaching the bottom, were employed
by Mr. J. B. Francis in the measuring flumes at Lowell,
Massachusetts, and other places, for determining the amount
of water drawn from the canals3. The flumes were con¬
structed in tolerably straight and uniform parts of the canals,
and rectangular in section, by lining the bottom and sides
with smooth planking, so that the current through them
might be as regular as possible ; and the tin tubes indicated
the mean velocity of the vertical section of the stream in
which they floated. By means of these tube-floats, placed
at different distances from the banks, the mean velocity
across the whole cross-section of the stream can be
ascertained with considerable accuracy.
The velocity-rods used for determining the mean rate of
flow on the Ganges Canal were very similar : they consisted
of loaded wooden poles, 1 inch in diameter, and also x-inch
tin tubes weighted at the bottom with a short piece of iron
rod, and closed at each end, the latter of the two proving
far the best4.
1 ' On the Theory of the Flow of Water in Open Channels.' Robert Gordon.
a ' Roorkee Hydraulic Experiments.' Captain Allan Cunningham.
3 ' Lowell Hydraulic Experiments.' J. B. Francis.
* ' Recent Hydraulic Experiments.' Major Allan Cunningham. Minutes of
Proceedings Institution C. E., 1883, vol. lxxi, p. 20.
28 Description of Current-Meters. [chap.
2. Current-Meters. Several current-meters have been
designed for measuring the velocity of a current by the
number of revolutions performed, in a given time, by a screw
or vane, fixed on a horizontal axis in the front part of
the instrument (Figs. 7 and 8). Woltrnann's turbine current-
meter, with improvements by Mr. Baumgarten, is the form
of instrument generally used in France1; and it was also
used in gauging the Connecticut River, as well as a novel
type of meter with four conical-ended vanes, designed to
move with little friction and to avoid catching weeds2.
The revolutions are recorded by toothed wheels, connected
with the axis of the screw, which turn an indicator revolving
round a graduated dial. The distance between each mark
on the dial is arranged to correspond with the space travelled
by the indicator when the current has run a certain distance,
usually taken at one foot. One or more wings at the tail
of the meter keep it, when immersed, with the axis of the
screw in a line with the current. A brake is fitted to the
instrument, so that the screw can be set free to revolve,
or can be stopped, by merely pulling tight or slackening
a string, above water, connected with the brake.
1 Annales des Ponts et Chaussées, 1847 (2), p. 326; 1858 (1), p. 121; i860
(I), p. 215 ; and 1883 (1), p. 219.
2 'Report of the Chief of Engineers, U.S.A.,' 1875, Part 2, pp. 306-309,
and Plates 2 and 3.
screw cu rrent-m ete i
Fig. 7.
Ii.] Methods of using Current-Meters. 29
To measure the velocity of a current, after noting the
reading of the index, the current-meter is immersed at the
desired position in the river, and lowered to the required
depth from a boat, being held in position by one or more
strings or rods. The brake is then raised, and the time
noted. At the close of the period of observation the brake
is applied, the meter is lifted out of the water, and the
distance traversed by the index is ascertained. If the index
registers in feet, the number of divisions traversed by the
Fig. 8.
index in the period, divided by the time, gives the velocity
of the current at the point where the meter was held. If
the mean velocity in a definite vertical plane is required,
it is only necessary to lower the current-meter gradually,
at a given spot, to the bottom, and then raise it again at
a uniform rate.
Though the manufacturers profess to make the index
register the exact distance run by the current, this is only
3o Improved Current-Meters : Gauge-Tubes. [chap.
approximately realized ; and when very exact results are
required the current-meter should be tested, either by placing
it in a stream with a uniform known rate of flow, or by
drawing it along at a given velocity in still water, in order
to ascertain what each division of the index, or a definite
number of revolutions of the screw, actually represent, and
what allowance should be made for frictional resistance at
different velocities.
In order to reduce the complication of toothed wheels
which check the rotation, to detect any temporary stoppage
of the rotation, and to avoid the necessity of lifting the
current-meter out of water after each observation, an electrical
contrivance has been fitted to the meter, which rings a bell
above water when a certain number of revolutions have
taken place. The same object has also been attained by
making the axis of the screw move a small hammer in
a definite number of revolutions. The hammer strikes
a tightly stretched wire which passes up out of the water
to a resonance box, so that each stroke of the hammer is
distinctly audible 1.
3. Gauge-Tubes. In 1732, Pitot proposed to gauge the
flow of water with a glass tube, having a short bend at
one extremity at right angles to the tube, and open at both
ends. The tube being immersed vertically in a current, with
the bend horizontal and its end facing the current, the water
in the tube is raised by the pressure of the current ; and
the height of the column of water above the surface is
proportionate to the velocity of the current, which is repre¬
sented by the formula,
V-jx V igh,
where h is the height of the column, and ¡x is a coefficient
which can be determined by drawing the tube through still
water at known rates. In this simple apparatus, the height
of the column was difficult to determine, owing to the ripples
1 'Deutsche Bauzeitung,' 1880, p. 229.
Ii.] Description and working of Gauge-Tubes. 31
on the surface of the stream and the oscillations inside the
tube, and was, moreover, not quite an exact measure of the
velocity.
To obviate these defects, Mr. Darcy employed two
tubes (Fig. 9), I inch in dia¬
meter, connected at the top
by à copper tube in which a
stopcock is fastened, which
enables communication to be
made at pleasure with the
outer air 1. Copper tubes are
fastened at the bottom of each
of the glass tubes, and bent at
right angles to the glass tubes,
and to each other, for a length
of 6 inches, and terminated by
mouthpieces having orifices of
only TV inch diameter. The
lower copper tubes have a
double-acting stopcock which
opens or closes them simul¬
taneously. The instrument
is lowered into the water with
both stopcocks open, unless it
has to be put right under
water, in which case the upper
stopcock is closed before the
instrument is submerged ; and
the tubes are immersed to the
desired depth, with one mouth¬
piece pointing to the current
and the other at right angles
to it. The water flowing into the tubes rises highest in the
Gauge-Tubes.
Fig. 9.
1 Annales des Ponts et Chaussées, 1858 (1), p. 354; and 'Recherches
Hydrauliques/ Darcy and Bazin.
32 Gauge-Tubes, and Hydrodynamometer. [chap.
tube pointing to the current ; the lower stopcock is then
closed, the instrument is raised, and the difference in level of
the water in the two tubes is measured at leisure.
It is important to see that the mouthpieces are clear,
that the tubes are kept perfectly vertical, and to wait for
taking the observation till the oscillations of the column
of water have ceased. It is also advisable to repeat the
operation three or four times, and take the mean of the
results.
This apparatus was used by Messrs. Darcy and Bazin
in their experiments on the canals of Chazilly and Grosbois,
for measuring the discharge through various channels, with
the object of investigating the laws of the flow of water
in open channels, and obtaining a general formula of dis¬
charge. Improvements have been recently introduced in
the apparatus by Mr. C. Ritter1, and the instrument has been
frequently used in France.
Hydrodynamometer. An ingenious contrivance, designed
by Mr. de Perrodil, measures the velocity of a current by
the torsion produced on a wire by the pressure of the water
against a disc2. The disc is fixed at the end of a horizontal
arm projecting from the extremity of a brass wire placed
vertically in the stream. The velocity of the current is
obtained by the equation V=cVa, where a is the angle
of torsion ; and the value of c can be determined, either
approximately from the coefficient of torsion of the metal
employed, or actually by experiment. The instrument used
by Mr. de Perrodil indicated a velocity as low as | inch
per second ; and it could be constructed so as to be still
more sensitive.
Comparison of Methods for measuring Velocities. Ex¬
perimenters have expressed considerable differences of
opinion as to the correctness of the methods employed for
1 Annales des Ponts et Chaussées, 1885 (1), p. 1058.
2 Ibid. 1877 (1), p. 467, and 1880 fi), p. 11.
Ii.] Comparison of Methods for gauging Flow. 33
measuring the velocity of a current. Messrs. Humphreys
and Abbot discarded the current-meter, as unsuitable for
the deep, turbulent, and irregular current of the Mississippi,
and relied exclusively on double floats; and on the Connecticut
River, double floats were found the most reliable method
for measuring sub-surface velocities in a uniform channel.
Lieut.-Col. Allan Cunningham tried three patterns of current-
meters on the Ganges Canal, but abandoned them in favour
of velocity-rods, on account of the insurmountable difficulties
he found in their use. On the other hand, double floats
were considered worthless by Mr. Révy, who trusted entirely
to current-meters, of the type shown in Fig. 7, for gauging
the discharge of the Paraná and La Plata rivers1 ; whilst
Mr. Darcy was adverse to double floats, and resorted to
gauge-tubes ; and Professor Un win has expressed his opinion
in favour of current-meters, as more correct and serviceable
than floats2. Mr. Bazin also, after comparing the results of
numerous experiments with double floats and current-meters
in large streams, came to the conclusion that, in many cases,
the use of double floats led to serious errors3.
The main objections urged against double floats are, that
the motion of the lower float is not accurately indicated
by that of the float at the surface, and that, moreover, the
upper float and connecting cord, being acted upon by a
different portion of the current, modify the motion of the
lower float. Also, that it is impossible to ascertain whether
the lower float remains at the desired depth, and is not lifted
by eddies into a higher part of the current, and constantly
shifting its position. Mr. Gordon observed, on the large river
Irrawaddi, that the motion of the top float was too uniform
to allow of the existence of the erratic movements of the
1 'Hydraulics of Great Rivers: The Paraná, the Uruguay, and the La Plata
Estuary.' J. J. Révy.
2 Minutes of Proceedings Institution C. E., vol. lxxi, p. 40.
3 Annales des Ponts et Chaussées, 1884 (1), p. 554.
D
34 Advantages of Floats for measuring Flow. [CHAP.
lower float attributed to it by the opponents of the method ;
and my observations of the motion of the lower float in small
clear streams led me to a similar conclusion.
The tube-float is free from these objections ; and no divergence
from the perpendicular can occur without being manifested
by the inclination of the portion of tube above the water.
The tube-float also measures the mean velocity of the
current in the plane in which it floats, and thus is more
rapid, as well as more accurate in its measurements than
the double float. The tube-float, however, is not suitable
for very deep rivers, or for channels where the depth varies
considerably, or where weeds grow in the bed ; and all floats
are liable to deviate from a straight course. Mr. Francis has
stated that the tube-floats in the Lowell experiments travelled
a little faster than the current ; whereas Major Cunningham
has calculated that such floats move somewhat slower than
the water in which they are immersed. Except for the
opposition offered to the air by the portion of the tube
out of water, there seems to be no reason why the average
velocity of the float and the current should not be identical ;
though with a change of velocity in the current, the momentum
of the float would prevent its assuming at once the same
change in rate.
Besides their applicability to large turbulent rivers like
the Mississippi, floats are valuable for measuring the flow
of sluggish streams in which the ordinary current-meter would
not act ; and, with the exception of the hydrodynamometer,
which has hitherto been only tried experimentally, floats
furnish the only available method in such cases.
Current-meters are merely inferential recorders of velocity,
and depend for their correctness on the accuracy of their
mechanism. Silt and floating weeds are liable to derange
them ; and the increased influence of friction at low speeds
renders them unsuitable for measuring feeble currents. Their
registration, however, can be checked by using two instru-
Ii.] Merits of Current-Meters, Gauge-Tubes, etc. 35
ments, and the continuity of the revolutions proved by an
electrical recorder1. Current-meters possess the advantages,
of not necessitating a long, uniform, unimpeded channel
for the observations, of restricting the measurements to a
single well-defined cross-section of the river, of being easily
placed and kept at the precise positions in the current at
which the velocities are to be noted, and of enabling the
observations to be taken much more rapidly than with floats.
Moreover, they can measure velocities closer to the bottom
than floats.
Darcy's gauge-tube gives good results in small channels,
with depths not exceeding 6 feet, and when the stream has
a fair velocity ; but it would be inapplicable to large rivers,
and very low rates of flow.
The sensitiveness of the hydrodynamometer, rendering it
suitable for measuring very feeble currents, appears to furnish
its chief claim to more extended adoption. Current-meters
and tube-floats are, in general, the best instruments for
measuring the velocity of flow ; whilst double floats and
gauge-tubes may occasionally be employed with advantage.
The choice of method, in any particular case, must depend
upon the size and configuration of the channel, the nature
of the flow, and the general condition of the stream, and
should be made in accordance with the indications given
above. The checking of the results by the occasional use.
where practicable, of a second method of observation, is
always valuable, both in securing a greater degree of accuracy
and more certainty as to the correctness of the experiments,
and also in extending experience as to the relative scope and
reliability of each method.
As the discharge of a river undergoes frequent changes, it
is generally sufficient, in estimating its flow at different
periods or heights, to obtain by experiment an approximate
1 'Die Messungen in der Elbe und Donau, und die hydrometrischen Apparate
und Methoden des Verfassers.' Professor von Harlacher.
D 2
36 Position of Maximum Velocity in a River, [chap.
measurement of its velocity ; but when the amount of water
supplied for any purpose has to be calculated, or still more
when the observations of velocity are to be used as a basis
for a general formula of discharge, the utmost care and
exactness are required in measuring the flow.
Distribution of Velocities in a Channel. The bottom and
sides of a channel necessarily retard the flow of the water
close to them, in proportion to their roughness ; and this
retardation, in irregular channels, is more due to the im¬
peding of the flow by eddies than by friction alone. The air
also in contact with the surface of the stream has a similar,
though naturally slighter influence, unless the wind happens
to be blowing in an opposite direction to the current.
Accordingly, the maximum velocity of flow in a river, along
a straight reach, is found in the central portion of the stream,
and generally somewhat below the surface, the actual position
depending on the size and condition of the channel and
the velocity of flow. The velocity increases at first from
the surface downwards for a short distance, and then de¬
creases down to the bottom where it reaches its minimum.
Messrs. Humphreys and Abbot found that the variation
in velocity of the flow of the Mississippi, in the place where
their experiments were conducted, at different depths on the
same vertical, might be represented by a parabola, having
its axis parallel to the surface, and at the depth below the
surface of the position of maximum velocity, the abscissae
representing the velocities at the different depths, and the
ordinates the vertical distances of these depths from the
line of maximum velocity. In the Mississippi experiments,
the maximum velocity was, on the average, nearly one-third
of the whole depth below the surface, varying with the direction
of the wind. Mr. Bazin had previously shown that the
velocities of the current, in small channels, could be repre¬
sented by the abscissae of a parabola with a horizontal axis ;
and he subsequently indicated the approximation of the
h.] Variations of Velocity in a River. 37
velocities on various rivers, at different depths, to the parabolic
form \ the curve varying with different channels, and with
the position of maximum velocity. Professor von Wagner,
in investigating the vertical curves of velocities of large and
small rivers, also found that they corresponded approximately
with parabolas having their horizontal axis at the depth of
the maximum velocity, but that they deviated from the
parabola towards the bottom, and near the surface, and that
the depth of the maximum velocity varied from a little below
the surface to a little over a fourth of the full depth 2.
The horizontal curves of velocities do not exhibit an
approximation to any definite curve ; for the velocities vary
very little in the main stream, diminishing very little from
the centre towards the sides, till on approaching the banks
they are rapidly reduced.
The retardation of the surface velocity has been attributed,
by some investigators, to the rising of the lower water to the
surface, being checked in its flow by striking against the
rough bottom and sides of the channel. Water, however,
rising from the bottom would have to traverse the layers
of maximum velocity before reaching the surface, and would
check the lower layers rather than the surface layer. The
checking of the flow against the sides might, indeed, tend
to raise the water-level near the banks, resulting in a trans¬
verse flow towards the centre, which would retard the surface
layer, and to which Mr. F. P. Stearns has attributed the
reduction of the surface velocity3. The existence, however,
of this transverse surface flow in rivers would have to be
clearly established before it could be considered as aiding
the action of the air and wind in checking the surface flow.
Mean Velocity deduced from Surface Velocity. Gauging
1 Annales des Ponts et Chaussées, 1875 (2), p. 309, and Plate 26.
2 'Hydrologische Untersuchungen an der Weser, Elbe, dem Rhein, und
kleineren Flüssen.' Professor von Wagner.
3 Transactions of the American Society of Civil Engineers, vol. xii, 1883, p. 331»
38 Ratios of Mean to Surface Velocity, [chap.
operations would be greatly simplified and shortened if the
mean velocity could be arrived at by merely measuring the
maximum surface velocity.
With the object of determining the ratio between the
maximum surface velocity and the mean velocity, De Prony
made some experiments in wooden troughs, and Messrs.
Baldwin, Whistler, and Storrow in channels lined with
planks. The coefficient obtained for converting the observed
surface velocity into mean velocity, were -816 by De Prony,
and from -Bio to -847 by the latter observers in different
channels. Subsequent experiments indicate that the co¬
efficient is generally comprised within the limits of o-8 and
0-9, depending upon the size of the channel and the nature
of its bed. In the Mississippi experiments, the coefficient ex¬
ceeded 0-9 ; but it is probable that the influence of the long
connecting cord and surface-float caused too large values to
be recorded for the velocities towards the bottom, and thus
gave too high a value to the mean velocity ; for Mr. Robert
Gordon, in checking his experiments on the Irrawaddi
with a current-meter, obtained considerable reductions in
the velocities on approaching the bottom, compared with
those recorded by double floats1. The coefficient would be
greatest for large deep rivers with smooth, uniform channels,
and least for small shallow streams with rough beds.
Messrs. Darcy and Bazin derived from the results of their
experiments the following formula, giving the relation be¬
tween the maximum velocity and mean velocity :
U — V= 25-36 VRS ■
where U is the maximum velocity in feet per second, V the
mean velocity, R the hydraulic radius in feet, and ^ the
slope. Assuming an average value for V of 90 VRS, we
' { Notes on Subjects connected with Works in the Irrawaddy Circle, British
Burma, with Records of Experiments on the Double Float and Woltmann Meter
Current Measurements,' 1883. Robert Gordon.
il.] Uses and Nature of Formulée of Discharge. 39
obtain U= 115-36 VRS, and therefore Ux -78= F, correctly
giving a somewhat smaller coefficient for converting the
maximum velocity into the mean velocity than that given
above for the maximum surface velocity.
Formulae of Discharge. One of the great aims of all
hydraulicians who have investigated the flow of water in
channels, has been to obtain a general formula, applicable
to any kind of channel and rate of flow, from which the
discharge can be readily calculated, provided the section of
the channel and the slope are known. Such a formula is
specially required when the channel of a river has to be
enlarged, or a straight cut made, and in designing drainage
and irrigation canals, so that they may discharge the
desired volume of water within the given time, and at a rate
not liable to be injurious to the bed and banks of the
channel. In existing watercourses, the direct measurement
of the discharge is much to be preferred to the results of
any formula.
Formulae of discharge are always given in terms of the
mean velocity, for when once the velocity for a given section
and slope has been ascertained, it is only necessary to
multiply the velocity by the sectional area of the channel, to
obtain the discharge, D = VA. Owing to the variable flow
of water in ordinary channels, a velocity formula cannot
be arrived at by any strictly mathematical process ; and
therefore all the formulae obtained are empirical, being
derived from the results of experiments, and their only
claim for acceptance being based on their accordance with
numerous experimental results arrived at with the utmost
precision. As the reliability of a formula of discharge
depends on the accuracy of the experiments on which it
has been founded, only those experiments should be taken
into consideration, in preparing a general formula, the
correctness of which has been fully established. With this
reservation, a formula of discharge will be suited for general
4o General Formula; for calculating Velocities, [chap.
adoption in proportion as it agrees with the measured dis¬
charges of very different channels under different conditions
of flow.
It is evident that the velocity of a current increases with
any increase of its slope, or of its hydraulic radius ; and
therefore 5 and R must enter into any formula expressing
the value of V, the mean velocity. Coulomb had concluded
from his investigations that the resistance of the wetted
perimeter to the flow comprised two functions, one varying
with the velocity, and the other with the square of the
velocity, from which Prony derived the formula,
« V+b V2 = RS,
in which a and b were friction coefficients to be obtained
experimentally. Subsequently Chézy proposed to omit the
first term as insignificant for a very small channel with a fair
velocity of flow, and established the simple formula,
V= cVRS,
which was accepted by many hydraulicians, who merely intro¬
duced modifications in the value of the assumed constant c.
Downing adapted it to British measures by making í=ioo;
so that it took the form V= too VRS, where V is the mean
velocity in feet per second, R the hydraulic radius in feet, and
5 the slope. Some of those who adopted this formula gave
different values to c according to the size of the stream, the
smallest value assigned to it being 68, for small streams ; but
none of them assigned it a larger value than Downing's1, which
is applicable to a discharge of about 17,000 cubic feet per
second. Most of the more recently proposed formulae retain
the general form of Chézy's formula, but introduce modifica¬
tions in the coefficient c, which has been made variable ;
whereas in the earlier formulae, no modification of the co-
1 'Hydraulic Manual.' L. D'A. Jackson, p. 252.
ix.] Formula from Mississippi Observations. 41
efficient was introduced for variations in slope and hydraulic
radius, or for the degree of roughness of bed.
Humphreys and Abbot's Formula. Though the object
of Messrs. Humphreys and Abbot's investigations was to as¬
certain the flow of the Mississippi by direct measurement, with
a view to the regulation of the river, they saw the importance
of endeavouring to deduce a general formula of discharge
from the results of their observations on one of the largest
rivers of the world, to serve as a guide in determining the
flow in very large channels. Having found that the maximum
velocity was below the surface, they deemed it necessary to
add the surface line to the wetted perimeter, as equally
retarding the flow ; so that they replaced the hydraulic
radius, 7? = —, which appears in all the other formulae, by
R'~ ——— , or the area of he cross-section divided by the
P+W
entire perimeter of the section, of which W is the width at the
surface of the water, and P the wetted perimeter. They also
introduced a coefficient into their formula, containing the
hydraulic radius R, and therefore changing with variations
1-69 ....
in R, namely m= .— ; and the formula in its hnal shape
VR + 1-5
became,
F= j (0-0081 m+ y225 R's/S)^— o-ogVm | .
In this complicated form, it is difficult to compare this
formula with others ; but omitting the small first and third
terms, and assuming W equal to P, and substituting a func¬
tion f varying with R, in the remaining middle term in
place of the omitted terms, we obtain a simplified equation,
1
F=io-6 f s/RS*, in which it is evident that the main
difference between this and the earlier formula, V=c VRS,
is that the velocity varies with the fourth root of .S instead
42 Linnted Value of Mississippi Formula, [chap.
of the square root, and to some extent with changes in R.
The modification introduced with regard to £ makes the
new formula only applicable to rivers with very slight slopes ;
and the absence of any variation in the coefficient according
to the nature of the bed, though of comparatively little
importance in a large river like the Mississippi, renders it
unsuitable for smaller streams, where the nature of the bed
in a restricted cross-section greatly affects the flow. More¬
over, the important alteration in the value of the hydraulic
radius, by the introduction of W, does not appear to be
warranted, either by the general moderate depth below the
surface of the position of maximum velocity, or by the
friction of the surface water against the air compared with
the friction occasioned by the bed ; and if the reduction in
the surface velocity is attributable to the roughness of the
sides, the coefficient should vary according to the amount of
this roughness.
It is, accordingly, evident that the Mississippi formula,
whilst valuable as throwing light on the conditions of flow
in large rivers with gentle slopes, is unsuitable for general
adoption. Nevertheless, the Mississippi experiments have
the special value of furnishing a test of the applicability of
other general formulae of discharge to the case of rivers like
the Mississippi.
Darey and Bazin's Formula. The experiments inaugu¬
rated by Mr. Darcy, and continued after his death by Mr.
Bazin, were undertaken with the object of ascertaining the
influence which the condition of the lining of a channel
has upon the discharge. The results of these experiments
in small, regular channels were embodied in a formula,
R
in which a and ß are given different values according to the
h.] Formula drawn up by Darcy and Bazin. 43
degree of roughness of the bottom and sides of the channel1.
Thus a varies from -000046 for cement or planed wood, up to
•00085 for earth ; and ß varies from -0000045 up to -00035
for the same differences of lining.
This formula, accordingly, whilst resembling Chézy's formula
in its general shape, introduces variations for the roughness of
the bed of the channel, as well as for changes in R allowed
for in the Mississippi formula, the constant c being replaced
by the variable coefficient / The formula naturally
V.+£
presents a considerable contrast to the Mississippi formula,
as the two sets of experiments on which these formulae were
based were conducted under diametrically opposite conditions,
representing the two extremes of such measurements. In the
Mississippi, changes of slope had great influence on the
discharge ; whilst in Mr. Bazin's observations, the roughness
of the channel exercised a very notable effect.
Comparatively recently, Mr. Bazin has revised his original
formula by the aid of further experiments in small earthen
channels and canals, and on the Seine and the Saône2,
arriving at the expression,
7_ \ )i
1 ~ I -00008534 (i+Ç)j VRS-
As the roughness of the bed, composed of earth, is supposed
to be similar in these cases of ordinary rivers and streams of
moderate dimensions and having a steady flow, the coefficient
simply varies with R ; and when R = $ feet, F=8o-2VRS;
wheni?=io feet, V=çi-i VRS-, and when R = $o feet, V=
IOI-5 VRS.
1 ' Recherches Hydrauliques,' H. Darcy and H. Bazin, p. 125 ; and Annales des
Ponts et Chaussées, 1871 ,1), p. 22.
2 Annales des Ponts et Chaussées, 1884 (il, p. 587.
44 Bazin s Formula for Torrential Rivers. [CHAP.
For torrential rivers carrying along shingle, the value of the
coefficient has to be modified, so that the formula becomes,
i J* _
•000122 (14-
and under these conditions, when R = $ feet, V=6v8 VRS;
when 7? = 10 feet, V=72-2 VRS ; and when R = 30 feet, V=
82-9 VRS. Observations have shown that the value of the
coefficient is 65-2 for the torrential tributaries of the Upper
Loire, and 68-8 for the Rhine at Basle, in spite of the fair size
of the river, the bed being covered with large shingle.
The values of the coefficient given above for certain values
of R are very nearly comprised within the range of values
assigned to c, in the formula V=cTRS, by the earlier
hydraulicians. Instead, however, of the coefficient being con¬
sidered constant, and having a special value assigned to it by
each experimenter according to the streams observed, the above
coefficient varies with R for channels having ordinary, regular,
earthen beds, and is changed when the nature of the bed is
different.
Ganguillet and Kutter's Formula. The formula of Messrs.
Darcy and Bazin, though very valuable for estimating the
discharge of small channels, and based upon correct scientific
principles, was not applicable to large irregular rivers with
very slight slopes, like the Mississippi. Moreover, Messrs.
Humphreys and Abbot expressed a desire that the suitable¬
ness of their formula for rivers with considerable slopes
should be tested by experiments. These circumstances led
Messrs. Ganguillet and Kutter to gauge the flow along moun¬
tain channels built of rubble masonry, with a semicircular
cross-section and a considerable slope, serving to discharge
flood waters ; and finding that the Mississippi formula was
inapplicable to these conditions, they proceeded to construct
a formula capable of embracing the results of the Mississippi
no Formula adopted by Gangnillet and Kidter. 45
gaugings, as well as their own, Mr. Bazin's, and other obser¬
vations.
Starting from Messrs. Darcy and Bazin's formula as a basis,
they eventually found it expedient to modify the form of the
y
coefficient, so that the formula became, V — d RS
1 + Vr
where y and x vary with the slope1. This introduction of
the slope into the coefficient was rendered necessary by the
anomalous fact that, whilst c decreased with a decrease of the
slope in small channels where the bed was not very rough, it
increased with a decrease of the slope in the Mississippi results.
This change was found to occur when R passed the value of
about one metre, or 3-281 feet, though the actual value of R
at which the change took place varied with the roughness of
the bed. This anomaly has been attributed to the intensifica¬
tion of the conflicting currents and eddies by an increased
slope in large rivers with very irregular beds, and in small
streams with very rough channels.
In pipes and small channels with steep slopes, where the
influence of a variation of the slope on the coefficient may be
neglected, making y— — + a and x — an, where I— VR' for the
special value ^'=3-281, and therefore /= 1-8x1, a=41-6, and
11 the coefficient of roughness varies between -009 and -04, the
equation becomes,
l i-811
— +a 1-41-6
V= dRS = — __ d~RS.
an a.1'071
i+~VR 1 + dR
For the general formula, applicable to any sized channel
and any slope met with in practice^ a factor containing the
1 ' A General Formula for the Uniform Flow of Water in Rivers and other
Channels.' E. Ganguillet and W. R. Kutter. Translated from the German by
R. Hering and J. C. Trautwine.
46 General Formula applicable to all Rivers, [chap.
slope is introduced into the coefficient by substituting « + 4?
for a, where m = -00281 is the constant of a hyperbola used in
constructing the formula. The formula thus becomes,
The authors of this complex formula claim that not only
is it applicable to large rivers and small streams, but that it
actually more nearly represents the results of the Mississippi
gaugingsthan Messrs. Humphreys and Abbot's formula. They
also assert its simplicity for practical use ; but this is only
attained by the aid of tables or graphic methods. Its com¬
plexity is a necessary result of its comprehensiveness ; and in
spite of this, it has met with very general approval as a skilful
attempt to condense all varieties of flow into a single formula.
The formula has come into use in Germany and Italy ; and
several persons who have investigated it have testified to its
reliability. Lieut.-Col. Cunningham has stated that it agreed
more nearly with his gaugings on the Ganges Canal than
Messrs. Darcy and Bazin's formula. Mr. Bazin, however,
has indicated that the only discrepancy between the two
formulae is found in cases like the Mississippi1 ; and he con¬
siders that the estimated velocities in the lower depths of the
Mississippi require important corrections, owing to the influence
of the current on the long connecting cord, as was proved in
the later experiments on the Irrawaddi
The permanence of this formula in its present shape will
depend upon the correctness of the gaugings of large rivers
with slight slopes ; for if these should prove to need important
modifications, the formula, in so far as it is based upon them,
will require alteration. The change of the influence of
a variation of the slope on the coefficient beyond a certain
1 Annales des Ponts et Chaussées, 1871 (1), p. 43.
a.] Simplification of Gaugings and Formulée. 47
value of R especially needs further investigation, for it presents
an anomaly which is not clearly intelligible, and which more
extended observations might either elucidate or show to be
incorrect.
Concluding Bemarks. The only way by which the time
and trouble involved in gauging operations could be diminished
would be by establishing a more exact relation, under definite
conditions, between surface velocity and mean velocity, or
between maximum velocity and mean velocity : for then
a single set of observations with a surface-float in the centre
of the stream on a still day, or with a current-meter at the
point of maximum velocity below the surface, would suffice
for obtaining the discharge. Rapid and approximate methods
of gauging the flow of rivers are very valuable in studying the
physical characteristics of a river with a view to works of
improvement, as it rarely is possible to resort to such elaborate
investigations as those carried out with public funds on the
Mississippi and the Ganges Canal.
The hydraulician is interested in observing the flow of
water under the most diverse conditions, with the object
of investigating the laws which govern its motion ; but the
practical engineer has more limited aims. The use of
a formula of discharge is absolutely essential for determining
the discharge of a proposed river enlargement or new cut,
or of a contemplated drainage or irrigation canal ; but such
channels are made regular, and with fairly smooth sides and
bottom, and bear no resemblance to the irregular channels of
large rivers. Accordingly, great as is the interest attaching
to the establishment of a general formula of discharge appli¬
cable to all known rivers and channels, a simpler formula of
more limited scope would generally fulfil the requirements of
engineers.
The most suitable formulae of discharge for general use,
namely those of Messrs. Darcy and Bazin, and Messrs. Gan-
guillet and Kutter, resemble in form the formula adopted by
48 Most suitable Formulée for General Use.
most of the earlier hydraulicians, V=c\/RSJ with the differ¬
ence that c is variable instead of constant. The variation of
c directly with R, and inversely with the increase in roughness
n, adopted in both these formulae, is clearly logical, for the
proportionate influence of the wetted perimeter in retarding
the flow would become less in larger rivers, and a greater
degree of roughness of the bed would diminish the velocity.
The effect of changes of slope on the value of c has not been
decisively determined ; and whatever this effect may be on
the very irregular channel of the Mississippi, it would not
apply to a comparatively small regular, artificial channel.
Accordingly, in ordinary cases, the amended formula of
Mr. Bazin for channels in earth, or Messrs. Ganguillet and
Kutter's formula in its simpler form, with « = -025 to -031,
ought to suffice for estimating the probable discharge in
designing an improved river channel, or a drainage or irriga¬
tion canal.
1 The value of « = • 025 is for brooks or rivers flowing in channels of earth;
whilst 71 — -03 applies to channels containing detritus or aquatic plants.
CHAPTER III.
REGULATION AND CANALIZATION OF RIVERS.
Imperfect natural Condition of Rivers. Variable Flow of Rivers. Progression
of Detritus. Highest and Lowest Water-levels to be observed. Regulation of
Rivers:—Removal of Shoals and Obstructions; Contraction of Channel; Cross
Jetties ; Longitudinal Training Banks ; Protecting and Easing Bends ; Straight
Cuts ; Submerged Cross Dykes ; Influence of Descent of Detritus ; Remarks on
Regulation of Rivers ; Instances of Regulation Works. Canalization of Rivers :—
Primitive Navigation; Stanches for Flashing; Locks, Weirs, and Level Reaches;
Instances of Canalized Rivers ; Remarks on Canalization of Rivers.
A RIVER to be perfect for navigation should be tolerably
straight, without sharp bends, with a fairly uniform breadth
and depth, and consequent regular flow, with neither a scarcity
of water in dry weather, nor too full and rapid a flow in
flood-time. Such a type of river is seldom found. Most
rivers, in their natural state, present constantly varying
conditions of breadth and depth, alterations in fall and flow,
and a more or less winding course. When the fall increases
abruptly, and the bed is rocky, rapids occur ; and when the
fall decreases considerably, and detritus consequently ac¬
cumulates in the channel, the river becomes shallow, and
widens out by eroding its banks in order to maintain its
discharge.
Variable Flow of Rivers. Torrential rivers are not generally
well adapted for navigation, on account of the great variations
in their discharge, and consequently in their depth, at different
periods of the year, and owing also to the rapidity of their
flow, which presents a serious impediment to up-stream traffic.
E
5o Ratios of Maxima to Minima Discharges, [chap.
The maximum discharge of the Loire at Briare has been
estimated to amount to 300 times its discharge at its lowest
stage ; whilst the ratio of the two discharges is over 100 to 1
for the Dordogne at Libourne and the Moselle at Metz, and
90 to i for the Garonne at Langon. Though the Rhone is
a torrential river, its flow is moderated by the Lake of Geneva,
and by some of its tributaries not being of glacier origin, and
therefore not having their greatest flow in the hot weather ;
so that in spite of its rapid fall, the ratio of its extreme
discharges at Lyons is only 40 to 1, considerably less than
the 70 to i of its tributary the Saône ; whilst lower down at
Valence, the ratio is only 30 to 1. The ratio of the extreme
discharges of the Seine at Paris is 37 to 1, owing to the
influence of its torrential tributaries the Yonne and the
Marne ; but it is reduced further down.
The variation in the discharges, which is excessive in
torrential streams near their sources, becomes gradually less
as the area drained increases ; and the flow of rivers also
generally becomes more gentle as they descend their valleys,
on account of the reduction in the general slope of the land.
As a reduction in fall necessitates a larger channel for con¬
veying the same volume of water, and since the discharge
increases, and becomes more regular as the various tributaries,
from districts exposed to different meteorological conditions,
join the main stream, a river is more suitable for navigation
along the lower portion of its course. This, moreover, is
indicated by the reduction in the ratio of the extreme
discharges of large rivers at a distance from their sources ;
for instance, the ratio is 13 to 1 for the Rhine at Kehl, and
less lower down ; it is 8 to 1 for the Danube just above its
delta, and only 4 to 1 for the Mississippi near the Gulf of
Mexico.
The velocity of the current of a river not only varies for
different heights of the river, but also exhibits frequent
changes at different places along the river, owing to variations
m.] Movement of Detritus down Rivers. 51
in fall or alterations in depth. The variations in fall are due
to changes in the general slope of the land, or to hard or
rocky ridges barring the channel. The alterations in depth
result from differences in the erosive nature of the bed, from
modifications in the width of the channel, or from the serpen¬
tine course of the river (Plate 3, Fig. 1).
Progression of Detritus. The shifting nature of the bed
of most rivers must be taken into account in any works of
improvement. A river in its natural condition may present
much the same appearance from year to year ; but the
shingle, gravel, sand, or silt forming the bottom of its bed is
always being carried down during high stages of the river,
and renewed by fresh supplies from above. Just as the
travel of shingle and sand along the sea-coast, under the
influence of the prevalent winds, is hardly perceived till
arrested by the erection of a projecting groyne, on the wind¬
ward side of which the material accumulates ; so the gradual
movement of detritus down a river is not readily noticed till
a modification of the channel disturbs the existing equilibrium.
The matter in suspension is, indeed, apparent ; but the larger
and heavier materials, being rolled along the bed, escape
observation in the turbid current, and often form a con¬
siderable portion of the solid discharge. The noise, however,
of rolling shingle may be heard in a rapid torrent by attentive
listening ; the movement of particles of sand may be perceived
in a clear, shallow, quickly-flowing current ; and soundings in
the Mississippi at Helena and elsewhere have shown that
sand-waves gradually proceed down the river1. The vast
amount of material thus brought down in past times is
evidenced by the wide-spreading alluvial plains through
which many rivers flow, and by the extensive deltas found
near the outlets of several large rivers ; whilst the continuance
of the same action at the present time, on a large scale, is
1 'Report of the Chief of Engineers, U. S. A., 1879,' part 3, p. 1963 ; and 1883,
part 3, p. 2216.
E 2
52 Highest and Lowest Levels of Rivers, [chap.
indicated by the sediment deposited by floods over con¬
siderable areas, of which the annual fertilizing deposit of mud
by the Nile in Egypt is a notable instance, and by the steady
progression of deltas, as proved by careful surveys of the
outlets of the Rhone, the Danube, the Mississippi, and other
delta-forming rivers. Though detritus is mainly brought
down by floods, the scour is so much increased by the
volume and velocity of the discharge, that, in spite of the
accession of large supplies of fresh material, a river is
generally deeper in flood-time than at a low stage.
Highest and Lowest Water-Levels. There are three levels
of a river which it is specially important to note, namely, the
highest known level, the highest level at which navigation can
be carried on, and the lowest known level. A knowledge of
the highest recorded flood level is necessary so that works
which must not be submerged may be raised above that
level, and that due protection may be afforded to low-lying
lands adjacent to the river. All works connected with the
navigation of a river have to be regulated in accordance
with the highest navigable level. The lowest level indicates
the minimum depth of water that is found in a river. The
low-water level, however, upon which works of improvement
are based, is not necessarily the lowest level to which the
river may fall. The absolute lowest level of a river depends
upon the combined effect of several physical causes which
may not occur together, except at very rare intervals ; and,
accordingly, it is not generally advisable to provide, by more
expensive works, for a very rare occurrence of short duration.
The lowest ordinary water-level is usually selected as the
datum below which the desired depth is to be maintained.
Regulation of Rivers.
Removal of Shoals and Obstructions. To render a river
suitable for navigation, its depth and flow should be made
fairly uniform, by lowering the shoals to the requisite
m.] Removal of Shoals and Obstructions. 53
minimum depth, and by regulating the rapids. A simple
expedient, at first sight, for accomplishing this object, would
be to dredge away the shoals, and to remove the obstruc¬
tions causing rapids in the channel. Unless, however, the
shoals are hard permanent ones, which the river has been
unable to erode, they will be soon formed again by fresh
detritus, owing to the enlargement of the sectional area of the
channel by the dredging, and the consequent slackening of
the current. The removal also of the obstructions forming
rapids, whilst improving the flow at those parts, lowers the
water-level of the river above, by the increased facility of
discharge, and therefore diminishes the available depth when
the river is low, and causes the shallowest parts of the upper
channel to become fresh impediments to navigation, and,
moreover, increases the fall of the next rapid above. Ac¬
cordingly, such methods of improvement are liable to prove
nugatory in the one case, and only serve to shift the position of
the impediments to navigation in the other, if supplementary
works are not carried out. Whilst, however, it is practically
useless to attempt to lower soft shoals permanently by
dredging alone, in rivers carrying along large quantities of
detritus, hard shoals, rocky reefs, and other obstructions
must be removed in order to effect improvements in a river ;
and any undue lowering of the low-water level above must
be counteracted by further works.
Contraction of Channel. The only permanent method of
lowering soft shoals, whether caused by an accumulation
of detritus from a reduction in fall, or resulting from the
unstable direction of the main current between two bends,
is by reducing the width of the channel along the site of the
shoal, thus producing its removal by the increased scour,
which also prevents its forming again. This reinforcement
of the current where its scouring capacity is defective, secures
the maintenance of the increased depth by natural means ;
but any contraction of the channel beyond the limits abso-
54 Regulation by Jetties and Training Banks, [chap.
^lutely necessary for securing the requisite depth, must be
avoided, as the additional increase in depth would lead to
a lowering of the low-water level to the detriment of the
channel above, and would cause the deposit of sediment in
the wider channel below.
Hard shoals and reefs have to be removed by artificial
means ; but their removal may render a contraction of the
channel necessary, if the channel has been thereby unduly
enlarged, so as to prevent silting or any serious lowering of
the low-water level.
Cross Jetties. The erection of jetties or groynes, projecting
at right angles from the banks into the channel at intervals,
has often been resorted to for regulating a river, with the object
of increasing its depth in wide shallow reaches. Experience,
however, has shown that cross jetties, if placed far enough
apart to render the system economical, do not generally
deepen and regulate the channel in a satisfactory manner.
The channel tends to adopt a circuitous course between the
jetties, and whilst becoming deeper opposite the ends of the
jetties, it shoals in the intermediate spaces.
Longitudinal Training Banks. The regulation of a channel
is more effectually accomplished by longitudinal embank¬
ments, following the course of the river, and fixing precisely
the widths of the low-water channel along the portions which
require training. In fact, these training works furnish stable
artificial banks, affording a suitable width of channel, in place
of the natural banks where subject to erosion. These banks
are ordinarily formed of continuous mounds of rubble stone,
or chalk, protected if necessary on the face by pitching or
concrete, and secured sometimes at the toe from undermining
by stakes or piles. Where solid materials are not easily
procurable or are costly, fascine mattresses, weighted with
stones or clay, are advantageously employed. In some cases,
where the alluvial matter is fairly light, floating bundles of
brushwood, or fascines, attached to a pole or cord, resembling
m.] Concentrating Discharge ; Easing Bends. 55
large weeds growing from the bed of the river, have been
successfully used for training a river, so as to produce scour
in one part and deposit in another, as for instance on the
Missouri1 and the Isar2.
Where islands, resulting from the accumulation of deposit
in mid-channel, or from a divergence of the current, split the
channel in two, the two branches do not afford the same
capabilities for navigation as the undivided channel above
and below. Accordingly, it is sometimes necessary to con¬
centrate all the discharge during the low-water stage of the
river into one of the channels, by barring the other one up to
a little above low-water level, so that a better depth may be
secured in the navigable channel when the river is low, whilst
leaving the secondary channel available for the passage of
floods.
Protecting and Easing Bends. As the main current of
a river is directed against the concave bank at a bend, the
bank is subject to constant erosion (Plate 3, Fig. 1); and
the sinuosity of a river through alluvial plains tends always to
increase, to the detriment of navigation, till at last sometimes
the loop becomes so tortuous that the river in flood-time
opens out for itself a more direct channel if the intervening
land is flat. To avoid an undue increase in curvature, a con¬
cave soft bank must be protected by fascines, stakes, or
stones ; or the current may be kept off from the bank by
short projecting spurs of stone or brushwood at a low level,
a system possessing the further advantage of straightening
the low-water channel, leading to a reduction of the excessive
depth near the concave bank, and a corresponding diminution
of the shoal at the change of curvature (Plate 3, Fig. 6).
Where the curvature is already excessive, it may be eased by
a flatter training bank in front of the concave bank (Plate 3,
1 ' Report of the Chief of Engineers, U. S. A., 1879,' part 2, p. 1051.
2 'Zeitschrift für Bauwesen,' 1886, p. 515; and 'Wochenschrift des Oster¬
reichischen Ingenieur und Architekten Vereines,' 1888. p. 74.
56 Regulation by Straight Cuts ; Cross Dykes, [chap.
Figs. 2, 4, and 5) ; but such works are liable to be costly,
owing to the depth and scour close to the bank ; and the
width of the channel at a bend must not be reduced with¬
out a corresponding modification of the channel between the
bends, which must be narrower than at the bends to maintain
the depth where the current is shifting over from one bank to
the other (Plate 3, Figs. 3, 4, and 6).
Straight Cuts. Sometimes it is expedient to do away with
tortuous bends in a river by cutting a direct channel where
the current is gentle, and thus restore the river to a primitive
condition. The channel is thus rendered more convenient for
navigation, whilst its length is reduced ; and the increased
fall ensures the maintenance of the depth of the new cut.
The channel, however, should be excavated to the full size
before the diversion of the river into it is effected, as its
enlargement by scour would only result in the deposit of the
eroded material in the channel below the cut. The banks of
the cut at its upper end should be protected from erosion,
and also the banks of the river immediately below the cut, as
any irregularities in the direction of the channel just above or
below the straight portion would lead to the development of
fresh sinuosities, and eventually obliterate the benefit resulting
from the rectification of the channel. Where the current is
rapid, straight cuts are inexpedient for the purposes of
navigation, as they augment the difficulties of the up-stream
navigation by increasing the velocity of the current.
Submerged Cross Dykes. With whatever care a river may
be trained, so as to secure a width suitable to the fall and
the desired depth, the bottom of the river always exhibits
variations in depth, resulting from the changes of the scour
of the current in a winding course, from differences in the
constitution of the bed, and from irregularities in fall. To
obtain uniformity of depth, it is, accordingly, necessary to
protect the bed from undue erosion, just as the sides of the
channel are protected by training banks to secure uniformity
m.] Submerged Dykes ; and Descent of Detritus. 57
of width. This may be accomplished by depositing dykes of
rubble, or mattress sills, across the channel in the deep places,
but kept below the navigable depth in the main channel
(Plate 3, Fig. 5). These submerged dykes should project
a little up-stream from the two banks, and dip down towards
the centre of the channel, so that whilst regulating the
current, they may direct it towards the centre of the river.
The undue lowering of the water-level by the increased
discharge in the deep parts of the channel is, moreover,
prevented by these submerged works ; and the fall at a rapid
is reduced by thus regulating the eroded hollow at the foot
of the rapid. This method of regulating the bed, as well as
the banks of a river, has been adopted for many years on
rivers in Germany, and has been more recently applied to
the Rhone below Lyons.
Influence of the Descent of Detritus. It is evident that
the detritus brought down by rivers seriously enhances the
difficulties attending their improvement ; and these difficulties
largely depend upon the amount, density, and size of the
detritus brought down in proportion to the discharge, and
also upon the fall of a river and its variations. If the velocity
of a river was uniform, or was only reduced in its descent in
proportion as the facility of the transportation of the detritus
is increased by its reduction in size by attrition, a regular
accumulation of this alluvial matter would only be experienced
at the outlet of the river. Any material reduction in velocity,
however, such as is often met with, especially in torrential
rivers, leads to the arrest of some of the detritus and the
formation of a shoal, which cannot be coped with by dredging,
owing to the abundance of the supply in many cases and its
continuity. Any partial regulation of the channel merely
shifts the position of the shoal ; whilst the embankment of
a river, to obtain the increased scour of the flood waters for
deepening the channel, or in order to prevent inundations, cuts
off the lateral areas which served for the deposit of sediment,
58 Arrest of Detritus carried down by Rivers, [chap.
so that a larger volume of solid matter is carried down to be
deposited in the channel, or to accumulate at the outlet.
Accordingly, regulation works in rivers carrying down con¬
siderable quantities of detritus, must not aim at local, isolated
improvements in depth, but must be carried out as a complete
scheme, with a view to the uniform discharge of the detritus,
by securing, as far as practicable, uniformity in velocity.
Adjacent plains also, serving as natural lateral reservoirs for
the deposit of detritus in flood-time, should not be shut off from
the river if possible, especially where a river descending from
the hills into a plain experiencei'an abrupt reduction in fall.
The injury caused to rivers, flowing through plains, by
the descent of detritus resulting from the disintegration of
mountains, has in some special cases been arrested near its
source by placing a dam at a suitable place across the valley
of a mountain stream, behind which the descending detritus
accumulates, another dam being erected at a fresh spot as
soon as the space behind the previous one has become filled
up. This system, which has been occasionally resorted to in
mountainous districts in Europe, has been also employed
in California to arrest the devastation of alluvial lands,
occasioned by the carrying down by mountain torrents of
large quantities of débris left by the extraction of gold from
the mines in the hills, which filled up the channels of the
rivers in the plains at the base of these mountains, wasted
the flat lands by deposits of gravel, and produced extensive
floods1. The great volume of detritus, however, brought
down generally from the numerous mountain sources of
a large river, cannot in practice be dealt with in this way,
owing to the sources being often under a separate jurisdiction,
and sometimes in a different country to the main river, and
also owing to the cost the works would involve, the large
spaces needed for storing up the detritus, and the injury to the
valleys such vast and constantly increasing accumulations
1 ' Report of the Chief of Engineers, U. S. A., 1881,' part 3, p. 2485.
m.] Checking Detritus ; Regulation of Rivers. 59
would entail. Nevertheless, the supply of detritus may be
somewhat checked by stringent enactments against throwing
débris into streams, or placing it so close to the banks as to
be carried down in flood-time ; and it may be reduced by
protecting the mountain slopes from disintegration by pre¬
venting their denudation from reckless clearing, and by
encouraging the growth of trees and vegetation in exposed
parts. This latter course not only diminishes the supply
of detritus, but also reduces the power of the torrents to
carry it down, by equalizing their flow.
Remarks on the Regulation of Rivers. Though regulation
works are valuable in many cases in fixing and deepening
the channel of rivers, and in equalizing their flow, this system
by itself is only suitable for obtaining an adequate navigable
depth in large rivers with a tolerably uniform flow. If the
discharge of a river, and consequently its depth, becomes
very small in the dry season, it may be impossible to secure
a sufficient depth, during that period, by restricting the width
of its low-water channel within reasonable limits. Moreover,
the removal of shoals, by facilitating the discharge, lowers to
some extent the low-water level which has been raised by the
contraction of the channel, and consequently reduces the gain
in depth which might otherwise be anticipated. To obtain
the most satisfactory results, the regulation of the sides of
a river should be accompanied by a regulation of its bed,
so as to secure uniformity in depth and slope as well as
in width.
The large rivers of North America have been regulated in
many places, where deficient in depth, with successful results ;
and some of the large rivers of Germany, and also the Danube,
have been improved for navigation by systematic regulation
works. The Rhone, however, is the only river of France
that has been improved by this method alone, which indeed
has been only rendered practicable in this instance by the
different seasons of the floods of its tributaries, affording
6o Regulation of the Rhine and the Elbe. [chap.
it a comparatively regular flow. This system, accordingly,
has a somewhat limited application ; and other means have
to be resorted to for improving the navigable capabilities of
rivers having smaller, or more variable discharges.
Instances ~8f Regulation Works. The Rhine has a very
moderate fall between Plittersdorf, about 29 miles below
Strassburg, and the sea, varying on the average from 1 in
4,300 down to i in 10,000 near its outlet, with the exception
of the portion between Bingen and St. Goar, where the
fall is increased to 1 in 2,000, and the current is very rapid
in places in a narrow rocky channel. The navigable con¬
dition of the river has been gradually improved by extensive
regulation works below Strassburg, reducing its low-water
channel in width, so that it has a minimum depth of 3 feet
below Strassburg, increasing to 6£ feet above Mannheim,
which continues to Bingen (Plate 3, Fig. a). At the rapid
part of the river, between Bingen and St. Goar, the minimum
navigable depth barely attains 5 feet, in spite of the removal
of rocky obstructions ; but below St. Goar the minimum depth
reaches *¡\ feet, and continues the same down to Coblenz and on
to Cologne, below which a depth of about 8 feet is obtained,
increasing somewhat in Holland. This large, unimpeded
waterway, extending a long distance inland, though with a
very moderate minimum depth, provides the route for a very
large inland trade, reaching altogether about ten million tons
annually in Germany alone, and has enabled a very flourishing
port to be established at Mannheim, 352 miles from the sea,
with a traffic of nearly two million tons annually.
The Elbe has been trained by a combined system of
longitudinal banks and dipping cross jetties, with submerged
dykes at the end of some of the cross jetties to protect them,
and also in front of the longitudinal banks in some places
to reduce the excessive depth at their toe (Plate 3, Fig. 3)l.
1 1 Handbuch der Ingenieurwissenschaften, Band III, Der Wasserbau,' L. Franzius
and Ed. Sonne, plate 7.
m.] Regulation of the Elbe, Niemen, and Rhone. 61
These works have provided a depth, at the ordinary-
water-level, of 5 feet at the German frontier, down to feet
at Hamburg, a distance of 316 miles ; but the minimum
depth of 3 feet at the lowest water-level has been obtained
with difficulty, and a further contraction of the low-water
channel was found necessary. The Elbe has thereby been
constituted an important waterway for inland navigation,
second only to the Rhine in Germany.
The Niemen also has been deepened 15 to i| feet by
similar training works, increasing its navigable depth to 5!
feet at mean low-water (Plate 3, Fig. 4).
The regulation of the Rhone between Lyons and the sea,
along a length of 301 miles, was commenced in i860, and has
been continued up to the present time with important modifi¬
cations 1. The Rhone is not navigable above Lyons at the low-
water stage, its bed even through the upper part of the town
being strewn with shoals of gravel and stones. Even below
Lyons, where its flow is augmented and regulated by the con¬
fluence of the Saône, its fall is rapid in many parts, especially
for some distance below Valence, where it averages about
i in 1,380, and is also irregular ; and its bed being readily
eroded in most places, the deepest channel was frequently
shifted by floods. The works first undertaken were designed
to fix the main channel, and to increase the depth in the
worst places, by shutting off secondary channels, and by
training the river up to a mean water-level by longitudinal
stone embankments, so as to lower the shoals by the scour
of the concentrated current. These embankments, raised
from 6| to 10 feet above the low-water level, had necessarily
to be placed too far apart to train the low-water channel,
which consequently meandered between shoals, crossing over
from one concave bank to the next on the opposite side.
1 'The Training of Rivers, as illustrated by the Results of various Training
IVorks,' L. F. Vernon-Harcourt, Minutes of Proceedings Institution C. E. 1S94,
vol. cxviii, p. 4, and plate 1, tigs. 1-5.
62 Regulation of the Rhone below Lyons, [chap.
Moreover, the high embankments made the current, when
filling the trained channel, scour deep holes along the concave
bends. The river accordingly, even in the trained portions,
though considerably improved, was not uniform in depth ;
and the removal of the shoals, by lowering the water-level
in the trained portions, increased the fall and reduced the
depth above the ends of the training works, so that the
improvements in some parts led to the appearance of fresh
obstacles at others. The positions, indeed, of the shoals and
rapids were altered ; but neither uniformity of depth nor of
fall was attained, and therefore the navigable capabilities of
the river at a low stage were not much improved by these
training works.
Since 1880, a new system has been adopted ; and the
river has been trained as a whole, instead of being dealt
with in isolated portions '. By reducing the height of the
training banks, it has been possible to diminish the width
of the trained channel, so as to correspond more nearly
with the low-water stage, and thus prevent to some extent
the undue lowering of the water-level at low water. In
order, however, to render the fall fairly uniform, which
varied with the differences in depth, submerged dykes
projecting from the training banks slightly up stream, and
dipping down towards the centre of the channel, have
been placed in the deep places, so as to regulate the bed.
and have thus reduced the fall over the shallower parts by
increasing it in the deep places (Plate 3, Figs. 5 and 6).
These submerged dykes, moreover, besides raising the lowered
water-level, and thus increasing the available depth, possess
the advantage of directing the main current towards the
centre of the channel, and thus straighten the navigable
channel, and keep the vessels away from the concave bank
at the bends.
1 1 De l'Amélioration des Rivières navigables à fond mobile,' M. Jacquet.
' Congrès international de l'utilisation des eaux fluviales,' Paris, 1889, p. 159.
m.] Rhone Towage ; Transhipment at St. Louis. 63
The available minimum navigable depth of the Rhone
below Lyons, down to Aries, has been raised by these works
to 3§ feet, and 5i feet below Aries, whilst the river does
not fall to its lowest level every year ; and the annual traffic
has attained 236,000 tons. The trade, however, along the
river does not appear at all commensurate to the very
extensive training works which have been carried out. Un¬
doubtedly, the rapidity of the current must always prove
a hindrance to the up-stream navigation, though various
systems of towage have been proposed for minimizing its
difficulties ; and the cost of the works and their duration
would have been greatly reduced if the improved system of
training had been adopted at the outset. Nevertheless, two
serious obstacles have existed, up to the present time, to
the proper development of navigation on the Rhone. The
one is the absence of competition in conducting the navi¬
gation along the river, as the towage has hitherto been
monopolized by a single company, owning about sixteen
steamboats serving both as tugs and for carrying cargo ; but
towage by means of a fixed submerged chain, wound up or
unrolled by a drum on the tug, is being tried by a new
company at Lyons, with a view of working the navigation
in a series of sections corresponding to the lengths of the
chains laid down. The other obstacle is the necessity of
transhipping the cargoes carried by water between Lyons
and Marseilles at Port St. Louis, which might be obviated
by enlarging the old canal going from Aries to Bouc, and
prolonging it to Marseilles, so as to connect the Rhone
directly with Marseilles without the intervention of a sea
passage, and thus enable river-craft to go from Lyons
right into the port of Marseilles.
The above instances show that a considerable traffic can
be carried on improved inland waterways possessing very
moderate minimum depths, to which, however, the rivers only
fall for short periods, and not every year ; whilst the history
64 Floating of Timber and Rafts down stream. [CHAP.
of these and other inland navigations indicate that the trade
quickly increases with improvements in depth and accessibility.
Canalization of Rivers.
In rivers where the low-water discharge is small, and a
considerable improvement in the minimum depth is required
for navigation, it is necessary to obtain the requisite increase
in depth by raising the low-water level. This system,
moreover, possesses the advantages over training works, not
merely of affording greater scope for improvements in the
navigable channel, and of extending notable improvements
in depth to moderate-sized rivers, but also of rendering it
possible to obtain a further increase in depth without dif¬
ficulty if circumstances make it expedient, and of making
the up-stream traffic almost as easy as the down-stream.
Primitive Method of River Navigation. In old times,
before any alterations had been made in the natural con¬
dition of rivers, trees felled in the forests were floated
down stream, a plan still in common use in mountainous
regions, and in places devoid of roads. Subsequently, the
timber was formed into rafts capable of carrying down
some of the produce of the upper country, where the flow
of the river was adequately regular ; and next small boats
were attached to the rafts, better suited for the safe convey¬
ance of goods, which were taken to pieces and used as
timber on reaching their destination. In France, at the present
time, 629 miles of rivers are classed as floatable. This traffic,
however, can be merely carried on down stream, and only when
there is sufficient water in the river to float the rafts and
boats over the numerous shoals met with in their course.
Stanches for Plashing. Later on the idea was conceived of
producing artificial floods, by damming up the whole dis¬
charge of the river for a certain period, till a considerable
quantity of water had accumulated behind the dam, and then
m.] Stanches for Flashing on Rivers. 65
letting it suddenly escape. This was accomplished by means
of stanches which consisted of spars, planks, or paddles,
supported by the pressure of the water against a sill below
and a movable beam above. These stanches were erected
across a narrow place in the river ; and a man standing on
a foot-bridge above was able to open or close them as
required. The level of the water was raised for a consider¬
able distance above these stanches when closed ; and the
mass of water being set free when the stanches were sud¬
denly opened, the boats, which had previously collected
above, were floated over the shallows below. This operation
of flushing, or flashing, as it is termed, continued in general
use on several French rivers till past the middle of the
nineteenth century, and was also employed on the Thames
and the Severn1. The stanches on the Severn were removed
in 1842, when some improvement works were carried out;
but some stanches are still in existence on the Thames
above Oxford, where they chiefly serve for keeping up the
water-level in summer, and are then generally only partially
opened for the passage of the few boats frequenting that
part of the river.
A novel form of stanch was introduced by Mr. Poirée on
the river Yonne in 1834, consisting of a series of iron frames,
placed about 3Í feet apart across the river and carrying
a foot-bridge, each connected by a bar on the up-stream
side, against which and a sill at the bottom, a row of long
wooden spars rested, square in section, placed close together
nearly vertical, thereby closing the channel2 (Plate 4, Fig. 7).
The slender spars, or needles (aiguilles) as they are termed
in France, were readily raised successively or lowered by a
man from the foot-bridge, thus opening or closing the stanch.
At that period the system of navigation by aid of flashes
was employed on the Upper Seine, the Yonne, and other
1 Minutes of Proceedings Institution C. E., vol. iv. p. III.
1 Annales des Ponts et Chaussées, 1839 C1)» P- 23^-
F
66
Flashing ; and Canalization. [CHAI'.
French rivers; and it is continued at the present time in
France on the Upper Yonne and some small floatable rivers.
Flashing is very inconvenient for up-stream navigation ; and
the danger experienced by descending vessels in the rapid
passage at the stanch, led eventually to the addition of a
lock alongside in some cases, so that vessels might avoid
passing through the stanch whilst making use of the flashing
for the rest of their journey. This system, moreover, only
affords an intermittent navigation for the down-stream
traffic; and therefore this method of river navigation com¬
pared unfavourably with the continuous navigation afforded
by canals, which were extensively constructed during the
latter half of the eighteenth century, and the earlier part of
the nineteenth century.
Locks, Weirs, and Level Reaches. Locks, which are sup¬
posed to have been first employed on some canals in Italy in
the fifteenth century, were eventually introduced on rivers, for
the purpose of securing a continuous navigation with an im¬
proved navigable depth. Instead of increasing the depth of
the river by lowering its bed, the level of the water is raised.
The river is transformed, by dams or weirs, into a series of
nearly level reaches connected by locks whose chambers, being
successively filled with water and emptied, serve to raise or
lower a vessel from one reach to the next. By this means
the requisite navigable depth is obtained with little excavation,
except at the portions of the river just below the locks, and
at old fords ; but the banks have sometimes to be raised for
a certain distance above the locks and weirs to prevént the
river overflowing its banks when its water-level is raised. The
position of the locks, and the length of the reaches, are re¬
gulated by the slope of the valley ; the fall at a lock generally
ranging between five and thirteen feet. The lock is usually
placed where a river is divided into two channels with an
island between, the lock being situated in one channel, and
a weir, keeping up the water to the level necessary for navi-
III.]
Canalization of Rivers.
67
gation, across the other channel (Plate 3, Fig. 7. and Plate 4,
Fig. 4). If no natural division of the river occurs at a point
sufficiently close to the place where a change in the water-level
is desirable, a site is selected where the river makes a bend,
a new straight channel is excavated across the bend, and the
lock is constructed in this channel (Plate 4, Fig. 5). The
ordinary discharge of the river follows the old channel, and
flows over the weir placed in the bend. This latter arrange¬
ment, though necessitating more excavation, is preferable, as
it interferes less with the original channel.
This canalization of a river, converting it into a fairly still-
water navigation, except during floods, greatly facilitates the
up-stream traffic, as the fall of the river is mainly concentrated
at the weirs under ordinary conditions of flow, and the current
of the discharge is reduced by the deepened waterway. More¬
over, the level of the river, and consequently the depth, does
not fall below the amount regulated by the height of the
weir in the driest weather, unless the expenditure of water in
lockage, evaporation, and leakage, exceeds the discharge,
which is very rarely the case.
Instances of Canalized Rivers. The extent to which the
minimum navigable depth has been improved by canalization
is indicated by comparing the former low-water lines on the
Upper and Lower Seine, and the Main, with the present
water-levels retained by weirs (Plate 4, Figs. 1, a, and 6). This
increase in depth, moreover, upon the Seine has been attained
by successive stages, by raising, and in some cases recon¬
structing the weirs. The canalization of the Upper Seine
was commenced in i860, with the object of obtaining
a minimum navigable depth of 51 feet ; but by works under¬
taken in 1878, the depth has now been raised to 6\ feet, ten
of the weirs having been raised, and the two top reaches
deepened1 (Plate 4, Fig. 1) Improvement works were com-
1 ' The River Seine.' L. F. Vernon-Harcourt. Minutes of Proceedings Institution
C. E., 18S6, vol. lxxxiv, p. 228.
F 2
68
Canalization of River Seine. [chap.
menced on the Lower Seine below Paris at the beginning of
the nineteenth century ; but as the dredging of the shoals
lowered the river, little advance was made in its navigable
capabilities, till its canalization was undertaken in 1838, and
completed in 1866, raising its minimum navigable depth to
5 J feet h Directly these works were completed, supple¬
mentary works were determined upon for increasing the depth
to 6\ feet; and in 1878, before the termination of these
works, further works were authorized, which have increased
the minimum depth to io£ feet by raising the water-level at
the weirs, an additional lock and weir having also been con¬
structed, and some shoals lowered by dredging. These works
enable vessels of 800 to 1,000 tons, and with 9 feet 10 inches
draught, to navigate the Seine between Paris and Rouen at
all times, except when floods rise above the navigable limit.
The canalized Seine between Montereau and Martot, at the
limit of the tidal river, a distance of 197 miles, has an inland
navigation trade second only in France to that of the northern
canals which bring the produce of the coal-fields of the north
of France and Belgium to Paris. The traffic, however, varies
in the different sections according to their position in regard
to Paris, as the feeders of the traffic converging mainly to
Paris join the river, and not in proportion to the navigable
depth ; for the traffic on the Upper Seine in 1892 amounted
to 1,114,000 tons between Montereau and Corbeil, and reached
2,295,000 tons between Corbeil and Paris ; whereas, on the
Lower Seine, the minimum traffic of 1,235,000 tons was
on the portion between Rouen and the mouth of the Oise,
and the maximum of 3,337,000 tons in the next section above,
between the mouth of the Oise and La Briche, where the
St. Denis Canal diverts a portion of the up-stream traffic by
affording a more direct route to Paris. The river traffic between
Corbeil and Paris furnishes another illustration of the large
1 ' The River Seine.' L. F. Vernon-Harcourt. Minutes of Proceedings Institution
C. E., 1886, vol. lxxxiv, p. 225.
m.] Canalization of River Main. 69
trade that can be carried along a wide waterway of veiy
moderate minimum depth. Moreover, though the traffic on
the Lower Seine, with a minimum depth of io| feet, is less
in one section than in the most frequented section on the
Upper Seine, with a depth of only 61 feet, the deeper waterway
enables the transport to be effected more economically, and
could accommodate a much larger trade.
The Thames and some other English rivers were gradually
canalized at a considerably earlier period than the Seine ;
whilst a large work of canalization was carried out on the
Main from its junction with the Rhine up to Frankfort,
a distance of 22 miles, as recently as 1883-86 (Plate 4, Figs. 5,
and 6)1. The Main had been previously trained by longitudinal
embankments on the left bank, and cross jetties along the
right bank ; but the minimum navigable depth, at the low
stage of the river, obtained by these works did not reach the
contemplated 3 feet ; and the annual river traffic between the
Rhine and Frankfort was only about 12,000 tons. The
canalization, however, aided by dredging, increased the
minimum depth to 6£ feet, enabling vessels of 700 to 1.000
tons to go up to Frankfort; and the traffic rose in 1887 to
300,000 tons. The remarkable success of these works led to
a decision, in 1889, to increase the depth to 8J feet, which is
being carried out by dredging near the upper end of each
reach ; and the traffic attained 709,000 tons in 1892.
Remarks on Canalization of Rivers. The above examples
of canalization illustrate the increasing difficulty which is
generally experienced in obtaining an adequate navigable
depth in ascending a river, or in its tributaries ; for whereas
a navigable depth of io| feet has been obtained in the Lower
Seine by locks and weirs placed on the average about 135
miles apart, the depth of 65 feet on the Upper Seine has only
1 ' Canal, River, and other Works in France, Belgium, and Germany,'
L. F. Vernon-Harcourt, Minutes of Proceedings Institution C. E., 1889, vol. xcvi,
p. 189.
7o Advantages of Canalizing Rivers.
been effected by locks and weirs placed at average intervals
of only 4! miles (Plate 4, Figs. 1 and a). Whilst also the
Rhine has been adequately improved by training works up to
Mannheim, 45 miles above the confluence of the Main, training
works proved inadequate in this latter river ; and locks and
weirs have had to be placed in it, only 4§ miles apart on the
average (Plate 4, Fig. 6). The length of the weirs is, indeed,
less in the upper part of the river, owing to the reduced width
of the river ; but this affords a very inadequate compensation for
the increased number of locks and weirs required, so that
canalization becomes more costly in the upper parts of a river
valley on account of the increasing slope.
Canalization is most useful where, owing to unfavourable
conditions, training works are unable to realize an adequate
improvement in a river ; and it enables navigation to be ex¬
tended to rivers, and to the upper portions of rivers and their
tributaries, which training works would fail to ameliorate.
Canalization, moreover, affords a much greater certainty of
obtaining a definite increase in depth than training works,
which, in deepening a river, are liable also to modify its water-
level ; and though somewhat impeding the down-stream
traffic by reducing the current and imposing a delay at the
locks, canalization renders it more safe, and greatly facilitates
the traffic up-stream.
CHAPTER IV.
DREDGING AND EXCAVATING.
Definition of Terms. Dredgers:—Bag and Spoon; Bucket-Ladder Dredger,
Stationary and Hopper, details, cost of machines, and cost of Dredging ; Dipper
Bucket Dredger ; Grab Bucket Dredger, varieties, and capabilities ; Sand-Pump
Dredger, instances, capacity for work and cost of Dredging ; Eroding Machines ;
Remarks on Dredgers ; Removal of Rock under Water. Discharge of Dredged
Materials—through a Hopper; by a Chain of Buckets; direct into Wagons;
through Long Shoot ; by Pump and Floating Tubes. Excavators:—Bucket-Ladder
Excavator, method of work, cost, capabilities ; Steam Nawy, varieties, cost,
capacity for work ; Grab Excavator ; Remarks on Excavators.
Dredging is the term applied to excavations carried on
under water ; whilst excavating refers to similar operations
conducted wholly on land. Extensive excavations are carried
out by means of large machines, worked by steam, which are
called dredgers when fitted on barges for removing material
under water, and excavators or steam navvies, when running
on wheels and excavating on land. Dredging is employed
for deepening the channels of rivers, removing obstructions
and shoals, lowering bars at the mouths of rivers, and en¬
larging canals. Excavators are only occasionally used in
river works, for forming a new cut through dry land ; but
they constitute a very important part of the plant in the
construction of large canals. Excavators are made on the
model of some of the principal types of dredgers, being merely
adapted for working on land. Dredging, however, is fre-
72 Dredging with Bag and Spoon. [chap.
quently carried out more economically than excavating, owing
to the comparatively low cost of transport by water.
Dredgers.
Bag and Spoon. The simplest form of dredging machine
is the bag and spoon, which consists of a leather or canvas
bag, having a circular ring of iron round its mouth, fastened
at the end of a long pole (Fig. 10). A
man standing in a barge holds the upper
end of the pole, and guides it ; whilst
another man drags the bag along the
bottom by means of a chain fastened to
the lower end of the pole, and winding
round a crab placed at the far end of the
barge. The mud or gravel is scooped
up by the flat lower part of the ring, and
enters the bag, which is drawn up by the
chain ; and the contents of the bag are
deposited in the barge. The chain is
then slackened, the bag is lowered to
the bottom, and the process repeated.
This primitive plan is still employed for
small operations in tolerably shallow
water, and is constantly used for getting Bag and Spoon,
gravel from the bed of the Thames. Fis-I0-
The aquamotrice is a modification of the bag and spoon, in
which an iron scoop, or small bucket, is hinged to its handle,
so that it can be turned over for discharging on releasing
a catch ; and the dragging of the scoop along the bottom is
effected by aid of the current of the river h The barge
carrying the scoop is moored over the site to be dredged ; and
1 Annales des Ponts et Chaussées, 1874 (2), p. 188.
¡v.] Aquamotrice : Bucket-Ladder Dredger. 73
paddle wheels, on each side of the barge in front, in revolving
under the action of the current, wind up a chain passing over
pulleys and attached to the scoop, and thus drag it along the
bottom, and raise it at the end of its course to be discharged.
Directly the scoop has discharged its load, the windlass
winding up the chain is put out of gear, till a sufficient length
of chain has been released for the scoop to return to the
starting point again, when the operation is repeated. Dredging
in gravel and shingle was accomplished on the Garonne, at
Agen, with this machine, raising 65 cubic yards of gravel per
day of i a hours, at a cost of 1 \d. per cubic yard, including
maintenance and depreciation.
Bucket-Ladder Dredger. The type of machine most com¬
monly employed for extensive dredging operations is the
bucket-ladder dredger (Plate a, Fig. 1 )1. It consists of a con¬
tinuous row of buckets fastened to two parallel endless chains
running on rollers and revolving round two tumblers at the
ends of a long girder, one end of which is fixed to staging on
the deck of the vessel, and the other end can be lowered as re¬
quired by chains attached to it, so that the lowest bucket may
come in contact with the bottom to be dredged. The buckets,
passing in succession along the bottom in a horizontal position,
scoop up the material, and revolving gradually, rise to the
surface in a more upright position, and after passing round
the upper tumbler, turn bottom upwards, and deposit their
contents, through a shoot, into a well in the hold of the
vessel, or into a barge alongside. The dredger is moved slowly
either sideways or forwards during the operation, so that each
bucket may have material to dredge. The steam-engine on
the dredger moves the buckets by causing the upper tumbler
to revolve.
Usually the ladder of buckets is situated in a well along the
centre line of the vessel, and is inclined at an angle which
1 The illustration of a bucket-ladder hopper dredger has been taken from a
drawing furnished me by Messrs. W. Simons and Co. of Renfrew.
74 Stationary, and Hopper Dredgers. [chap.
varies according to the depth to which the buckets are lowered.
Sometimes the ladder is situated alongside the vessel ; and
occasionally a dredger carries two ladders of buckets, one on
each side of the vessel.
The staging carrying the ladder is sometimes provided with
traversing gear, enabling the ladder of buckets to be moved
forwards so as to project in front of the bow of the vessel, and
thus allow the dredger to cut a channel for itself into a shoal
in advance.
Two distinct forms of bucket-ladder dredgers are employed,
namely, stationary dredgers, and hopper dredgers. Stationary
dredgers are so called because they remain stationary at the
spot where they are dredging, and discharge the dredged
material into barges alongside, which carry away the material
to the place of deposit ; but these dredgers are now generally
made self-propelling, so as to obviate the necessity of using
a tug whenever the position of the dredger has to be changed.
Hopper dredgers are provided with a well in the centre of the
vessel, into which the dredged material is discharged ; and
directly the well has been filled, the vessel ceases dredging,
and itself carries the load to the place of deposit. The bottom
of the well is formed by hinged doors which open downwards
when released, and discharge the contents of the well into the
water. This arrangement of a well with a movable bottom is
termed a hopper ; and by this means a dredger and a de¬
positing hopper barge are combined in a single vessel.
The stationary dredger possesses the advantage of being
able to remain continuously at work so long as there is
sufficient water to float it, provided it is served by an adequate
number of barges, depending upon its rate of dredging and
the distance of the place of deposit. The hopper dredger
has, on the contrary, to work discontinuous!}', its dredging
being stopped during the period required for leaving its
moorings, steaming down to the depositing ground, releasing
its doors and closing them again, returning, and picking up
iv.] Relative Advantages of these Dredgers. 75
its moorings ; and during these operations, the dredging
machinery remains idle, and merely increases the load to be
transported. A stationary dredger with hopper barges should,
accordingly, be employed when the amount to be dredged
is large, and in fairly thick layers, when there is ample
space in the channel for the dredger and its barges alongside,
when the place of deposit is distant, and when the cost of an
ample dredging plant is moderate in proportion to the total
cost of the dredging to be carried out. When, however, the
shoals are scattered, the thickness of material to be removed
is small, and the total amount to be dredged is moderate,
when the depositing ground is not far off, when subsequent
maintenance has to be provided for, and when the space
available for the dredging plant to work in is limited, it is
expedient to resort to a hopper dredger. This type of dredger,
though executing the work more slowly than a dredger with
barges, occupies less room, is handled more easily, and with
a smaller crew, is less costly, and is better suited for the com¬
paratively moderate amount of dredging requisite for main¬
tenance. Moreover, it can be supplemented when desirable
by an attendant barge, which it can load and take in tow;
but, on the other hand, its increased draught as its hopper is
loaded, reduces its period of work on a falling tide in a shallow
river. The very extensive dredging operations carried out
during many years on the Tyne, the Clyde, and the Tees
(Plate 8, Figs. 2, 5, and 8), have been effected by bucket-ladder
dredgers with attendant steam hopper barges, or hopper
barges towed out by tugs ; whilst many smaller improvements
in rivers and harbours, of which the approach channel to
Belfast is a notable instance, have been accomplished in recent
years by bucket-ladder hopper dredgers.
The dredger is gradually shifted forwards or backwards,
and sideways whilst dredging, by hauling with steam winches,
placed fore and aft, on chains attached to buoyed moorings,
generally six in number, laid ahead, astern, and on each side
76 Details of Bucket-Ladder Dredgers, [chap.
of the vessel, so that the chain of buckets may successively
traverse the area comprised within the moorings, and lower
it to a uniform depth. A dredger is often provided with
twin screws, so that it may be navigated more easily in
narrow or tortuous channels, and may be more quickly placed
in position for working. One of the largest dredgers hitherto
constructed, which, being built for the port of Bristol, has to
navigate the narrow River Avon, is provided with two pro¬
pellers both fore and aft, so that it can steam in either
direction without having to turn. A similar arrangement has
been adopted in the large hopper dredger constructed for
dredging a channel in the Mersey at the entrance to the
Manchester Ship-Canal (Plate 2, Fig. 1); and the ladder-well
has been carried through the stern, instead of through the
bow. in order to obtain greater seaworthiness and speed by
leaving the bow undivided.
The buckets, tumblers, links, and pins, are now made of
steel : and the bottom of the vessel is advantageously made
of steel if liable to have to ground at low water. The Bristol
dredger was made wholly of steel1. Cast-steel buckets have
been recently introduced, which, being formed in one piece,
are stronger than riveted buckets ; and manganese steel has
proved a very durable material for the pins which are sub¬
jected to great wear. The buckets vary in size according to
the size of the dredger ; and they should also vary according
to the nature of the material, a larger bucket being more
suitable for sand and soft material than for stiff clay and hard
material. A dredger, however, has often to work in various
kinds of soil, for which purpose two sizes of buckets are
occasionally provided. The largest sized bucket has a capacity
of about 23 cubic feet ; the buckets of the Bristol dredger
have a capacity of 17^ cubic feet, those of one of the Clyde
dredgers 15 cubic feet, of the Swansea dredger 13 cubic feet,
and of the Tees dredger 9 cubic feet ; whilst the buckets of
1 Minutes of Proceedings Institution C.E., vol. cvii, p. 316.
IV.j Capacity and Cost of these Dredgers. 77
small dredgers are still less. The buckets should taper out
from the bottom towards the mouth, so as to facilitate the
discharge of their contents. They used generally to be made
with holes in their sides, to let the water escape ; but now the
holes are often omitted, as the water, though adding to the
weight lifted, is useful in facilitating the discharge of the
materials from the buckets, and down the shoot. In dredging
stiff or hard material, ripping claws are sometimes added to
the chain of buckets at intervals to loosen the soil ; but the
buckets with steel lips, in the stronger machines, are capable
of dredging boulder clay, soft sandstone, and rock in layers.
The rate of working varies with the nature of the material ;
and as the buckets are often only partially filled, it is safer
to estimate the capabilities of a dredger by the amount
actually delivered into a hopper in a given time, than to
calculate it from the speed and capacity of the buckets.
Bucket-ladder dredgers are commonly built capable of
lifting from 200 to 1,000 tons of free soil in an hour under
favourable conditions, with maxima depths of working of
from 25 to 35 feet. Hopper dredgers are provided with hoppers
of from 250 to 1,300 tons capacity, according to the size of
the dredger. The cost of these dredgers ranges generally
between £8,000 and £30,000. Thus the four stationary
dredgers employed for several years past on the Clyde, cost
from £6,650 to £17,650 1, and have dredged in depths of from
28Ï to 34 feet ; whilst one of these raised an average quantity
of nearly 200 cubic yards, or about 250 tons per hour in
1890-91. The four dredgers working in the Tees cost from
£10,130 to £18,1002. The Swansea stationary dredger with
two ladders of buckets, raising on the average 320 tons per
hour, and a maximum of about 600 tons, cost £27,500. The
Bristol hopper dredger, with a hopper capacity of 1,000 tons,
1 ' The River Clyde.' James Deas. Minutes of Proceedings Institution C. E.,
1873, vol. xxxvi, p. 160.
2 ' Dredgers and Dredging on the Tees,' John Fowler. Minutes of Proceedings
Institution C.E., 1SS4, vol. lxxv, p. 239.
78 Cost and Speed of Hopper Dredgers, [chap.
and capable of dredging a maximum quantity of 800 tons
in an hour, cost £30,000. Steam hopper barges with hopper
capacity of 200 to 500 tons, cost from about £4,000 to £7,000 ;
and non-propelling hopper barges having similar capacities,
together with a steam tug, approximate to the same cost.
The speed of a hopper dredger is of considerable importance,
and also its draught, especially when working in a tideway,
for on these, and the rate of dredging, depends the number of
trips that can be accomplished in a day. The speed of these
vessels ranges between 7 and 10 miles an hour ; the draught
varies with the size, from about 6| to io| feet when light,
and 105 to 14I feet loaded. Two sets of engines work the
twin screws ; and one, or both if necessary, can be used for
dredging. When dredging plant has to be sent to a distant
country where barges are not available, the hopper dredger
possesses the important advantage of being self-contained,
and a seaworthy vessel ; whereas the transport of barges with
a stationary dredger presents difficulties.
The cost of dredging with bucket-ladder dredgers depends
on the nature and compactness of the material, the concentra¬
tion and amount of the work, the depth of water and the
velocity of the currents, the period during which dredging
is practicable and the exposure of the site, the distance of
the place of deposit, the cost of wages and coals, and the
efficiency of the dredging plant. All these conditions may-
vary with the locality, so that no definite price can be arrived
at ; whilst the problem is complicated by the fact that many
records of the cost of dredging only take into account the
actual working expenses, without making allowance for
repairs, interest on the capital cost of the plant, and its
depreciation. Nevertheless, the actual expenditure in various
dredging works, reckoned per ton or per cubic yard removed,
affords a sort of basis on which to found a rough estimate of
the cost of dredging under similar conditions. The price is
generally given per ton, as this is a measure of the actual
IV.] Cost of Dredging in Various Rivers. 79
work accomplished, and as the contents of the hoppers are
reckoned in tons ; but the improvement effected in a channel
depends upon the volume removed, and therefore should
strictly be estimated in cubic yards measured in place. The
work done, however, is often more than would appear from
comparing the cross-sections before and after the dredging,
owing to the silting which frequently takes place during the
deepening of a river ; but it is desirable to know the relation
between a ton and a cubic yard of material, wherever dredging
is carried out. On the Clyde, 1^ tons of dredging are con¬
sidered to occupy a cubic yard ; whereas on the Tyne, iï tons
are supposed to be the equivalent.
The actual working expenses of dredging with bucket-
ladder dredgers and depositing, are sometimes less than
2d. per ton ; whilst on the Tyne, including maintenance and
repairs, the cost of dredging has varied from 2%d. to 8\d.,
and has averaged 3-39^., the material being deposited in the
sea at an average distance of 11 miles from the workl. The
Bristol dredger has raised mud, sand, and sandstone, and
deposited it 10 miles off, at a cost of $d. per ton, exclusive of
depreciation. It is, however, far more satisfactory when the
price comprises every item of expenditure, including repairs,
interest on cost of plant, and depreciation On the Clyde,
the total cost of dredging boulder clay amounted, in 1872. to
2s. per ton, and in dredging sand and silt 6%d. per ton, which
in 1880 had been reduced to 4'id. per ton, the distance of the
depositing ground being from 9 to 27 miles. At Port Glasgow,
the total cost of dredging and of carrying the dredgings
7i miles, has been 3!d. per ton. The dredging of the direct
approach channel to Belfast through sand and clay, and the
conveyance of the material 10 miles, cost 3-210!. per ton;
whilst the deepening of the Tees in sand, silt, clay, stones,
and boulders, and carrying the dredgings about 9 miles on the
average, cost f^d. per ton. The cost of dredging a channel
1 Minutes of Proceedings Institution C. E., vol. lxxxix, p. 100.
8o Description of Dipper Bucket Dredger, [chap.
to deep water in Swansea harbour1, through sand, clay,
gravel, and boulders, and depositing the materials 7 miles off,
reached nearly 7d. per ton ; whereas at Carlingford Lough,
the removal of hard clay and boulders cost is. 5d. per ton2.
The inexpediency generally of getting dredging done by
contract is indicated by a tender for the work at Carlingford
Lough having amounted to is. 11 d. per ton, and by the
dredging by contract in Boulogne harbour having cost xod
per ton, for mud, sand, and stones, and is. id. in compact
schistose clay, when the place of deposit was only 2 miles
distantFurther instances of the average cost of dredging
in mud, sand, and clay, with hopper dredgers, are i.o6d.
per ton at the Port of Blyth, the material being conveyed
2 miles ; 3d. at Bombay, conveyed 5 miles ; and 4d. on the
east coast of Scotland, conveyed 18 miles4.
Dipper Bucket Dredger. The dipper dredger resembles in
principle the bag and spoon, but in a greatly enlarged form :
and it is worked by a steam-engine. The dredging is effected
by a single large cylindrical bucket, made of iron or steel,
with a projecting lip, fastened to the end of a long beam
carried by a revolving derrick projecting over the extremity of
a barge, and guided by chains. The scope of action of the bucket
depends on the length of its beam, and the projection of the
derrick ; and the whole area within the radius commanded by
the machine can be dredged before any shifting of the barge
is required. Buckets have been made large enough to hold
2J cubic yards of material, and the largest of these dredgers
have raised r,200 cubic yards of silt in a day of ten hours.
The bucket is provided with a hinged bottom opening down¬
wards, by means of which the contents of the bucket are
readily discharged into a barge alongside ; and it is some¬
times provided with pointed teeth projecting in front of the
1 Minutes of Proceedings Institution C.E., vol. ciii, p. 354.
2 Ibid. vol. xliv, p. 135. G Ibid. vol. lxxx, p. 263.
4 'On Dredging and Dredge Plant,' A. C. Schonberg, p. 20, Manchester Inland
Navigation Congress, 1890.
IV.] Description of Grab Bucket Dredger. 81
lip, to break up compact layers. The method of working
of this machine is precisely analogous to that of the steam
navvy which works on land (Plate 2, Fig. 9). This type
of dredger is suitable for dredging materials of any quality
from silt to loose rock, in depths ranging between 5 and
35 feet.
Grab Bucket Dredger. This form of dredger consists
essentially of an iron or steel semicylindrical, or hemispherical
bucket, opening at the bottom into two or three sections
(Plate 2, Figs. 2,4, and 5). The bucket is suspended by chains
from the jib of a crane, or by a spear from a derrick; and it
is lowered, with its jaws kept open, on to the bottom to be
dredged, penetrating the surface by its impact and weight. In
being raised, the jaws are released, and tend to close by the
weight of the sections aided by lever bars attached to them ;
and the sections coming together, excavate the material into
which they have penetrated, and raise it in the closed bucket.
The material lifted is readily discharged from the bucket,
when in position, by slackening the lever chains and opening
the jaws again, leaving the bucket ready for another descent.
The jaws are sometimes furnished with teeth to facilitate the
penetration of the bucket into dense material, for which
purpose the three- or four-bladed hemispherical bucket is also
suitable ; and occasionally rows of claws are substituted for
the sides of the bucket, to disintegrate hard material, and to
lift boulders, draw up piles, and remove large obstacles.
The buckets are made of various sizes according to the
nature of the work, ranging in capacity from about 6 cubic
feet, up to nearly 5 cubic yards.
The grab dredger is best suitèd for dredging silt, soft
sand, and small gravel, or other loose material, and for
raising large loose boulders, blasted rock, or rough débris
such as trunks of trees ; and it is able to dredge at any depth,
and in exposed situations, and is specially well adapted for
dredging in docks, locks, and other places where the available
G
82 Details concerning Grab Dredgers. [chap.
space is limited. The jaws of the bucket do not readily
penetrate compact soils and stiff clay; but this defect may
to some extent be overcome by the addition of claws, by
weighting the bucket, or by pressing it into the soil by
means of a spear or by hydraulic power (Plate 2, Fig. 6),
and thus rendering it capable of dealing with these materials,
where other forms of dredgers are not available, and
where the work is small and scattered. When working in
materials of various sizes, stones or other hard substances
are liable to be caught sometimes between the jaws of the
bucket when closing, so that an aperture is left, through
which the finer materials drop out of the bucket whilst it
is being raised.
These grab dredgers are generally carried on a barge
provided with a central hopper1, as the machinery is com¬
paratively light and occupies little space, and as these dredgers
often have to work in confined spaces (Plate 2, Figs. 2 and 3).
The adaptation of hydraulic power to grab dredgers increases
their power of penetration, and their speed of working; and
the hydraulic power is not only utilized for pushing the
blades of the bucket into the soil, but also for opening and
closing the bucket. The largest grab dredger consists of
a steamer fitted with a central hopper of about 1,200 tons
capacity, and carrying four grab buckets, each holding about
1^ cubic yards, worked by steam cranes, and capable of
raising about 3,000 tons of mud in a day, which is used for
removing the accumulations of silt from the Liverpool docks.
The cost of dredging with a grab varies greatly with the
nature of the material and its compactness or tenacity ; but
under favourable conditions, or with powerful machines, the
actual working expenses of dredging have in some cases
been only between id. and 2d. per ton, whilst the cost of the
plant is comparatively small.
1 The illustrations on Plate 2, Figs. 2, 3, and 4, have been reduced from
drawings furnished me by Messrs. Priestman Brothers, of Hull.
iv.] Working of Sand-Pump Dredger. 83
Sand-Pump Dredger. When the bottom of the channel to
be dredged is composed of sand or silt, the material is readily-
raised by a sand-pump or suction dredger. This machine
consists of a centrifugal pump fixed on a vessel, which draws
up a continuous stream of the sand or silt, mixed with water,
through a pipe dipping slightly into the material to be
raised, and deposits it into the hopper of the dredger, or
into a barge alongside1 (Plate 2, Fig. 8). The material settles
in the hopper; and as the filling proceeds, the water rising to
the top of the hopper, eventually flows overboard at the sides.
To give more time for the deposit of the suspended matter
as the hopper fills up, flap-doors are sometimes raised round
the sides of the hopper, retaining the water temporarily above
the top of the hopper, so that it may discharge its burden
of silt before flowing overboard, and thus enable the hopper
to be filled with deposit. The settlement of the material
in the hopper has also been secured by propelling the
mixture of sand and water through pipes running the whole
length of the hopper, and perforated with a series of holes
at the bottom, through which the sand drops down. Pure
sand is the material best removed by the sand-pump dredger,
for an admixture of silt reduces its mobility ; whilst fine silt
does not readily settle in the hopper, and a large proportion is
carried overboard again in suspension in the outflowing water.
Rough gravel and stones are often raised by sand-pump
dredgers when working on sandy shoals interspersed with
larger materials ; but they wear the pipe rapidly and are
liable to damage or stop the pump, so that sand-pump
dredgers are not commonly set to work to raise such
materials. The ordinary sand pump is not suited for work¬
ing in compact materials ; but when fitted with revolving
cutters, breaking up clay or other stiff soils at the bottom close
to the mouth of the pipe, the dredger is capable of raising the
1 The sand-pump dredger is from a drawing supplied to me by Messrs. J. and H.
Gwynne of London.
G 2
84 Sand-Pump Dredger at St. Nazaire. [chap.
loosened materialh The same object has also been attained
by a sort of plough, fixed to the end of the pipe, which digs
into the material to be lifted. When the suction pipe is fitted
with a telescopic joint, or a flexible end, the sand-pump
dredger can work in exposed situations, when the waves do
not exceed two or three feet in height.
The earliest sand-pump hopper dredger appears to have
been constructed for removing deposits of sand and silt fr»m
the port of St. Nazaire, and commenced working in 1859.
It consisted of a screw steamer provided with two hoppers
in its hold, and having a suction pipe hanging down on each
side from amidships, and slanting down towards the stern,
the two pipes being connected at the bottom by a perforated
pipe resting horizontally in the bed of silt, through which the
sand or silt was drawn up by the pumps2. As soon as the
hoppers were filled, which in the two larger dredgers sub¬
sequently built had a capacity of 360 cubic yards, the engine
which had worked the pumps was put into gear for work¬
ing the screw, to convey the vessel to the place of deposit
nearly a mile off. Each of the larger dredgers cost £6,100-,
and the three sand-pump dredgers raised and deposited the
material at a total cost of only 2-1$d. per cubic yard, or about
2-340?. per ton of thislight material ; whereas with a bucket-ladder
dredger used for removing the more consolidated deposits close
to the quay walls, the cost reached 5-57d. per cubic yard, or
about 5-650?. per ton. The hopper was filled in 3^ hours :
and the depositing occupied 1^ hours, including delays in
passing through the locks. The depth of dredging at
St. Nazaire ranges from 10 to 30 feet; but these vessels
could dredge to over 65 feet in depth with the pipes hanging
down vertically. It was found that if the perforated pipe
was allowed to sink more than 18 inches below the surface
1 Minutes of Proceedings Institution C.E., vol. civ, p. 191.
2 - Le dévasement du port de Saint-Nazaire,' M. Leferme, Annales des Ponts et
Chaussées, 1869 2), p. 15.
IV.] Sand Pumps at Amsterdam Canal. 85
of the sand, it became difficult to haul the vessel forward, and
a crust forming over the hollow produced, as much sand as
water was pumped up. This early example of dredging by
suction is interesting as an instance of remarkably economical
work, and also as being probably the first attempt at combining
a dredger and hopper in one vessel, which does not appear
to have been extended to bucket-ladder dredgers till thirteen
years later.
The more recent types of sand-pump dredgers have a
single suction pipe in front, or at one side of the vessel, with
its end dipping into the layer to be raised. Twelve sand-
pump dredgers employed upon the Amsterdam Ship-Canal
works, had a suction pipe, 18 inches in diameter, supported
by a wrought-iron framework, and projecting in front of
the bow of the vessel, which could be raised or lowered by
a chain attached near the end of the pipe, and suspended from
a projecting beam1. Each of these dredgers could raise
about 1,300 tons of sand in a day, by means of a centrifugal
pump having a fan 4 feet in diameter ; and the best results
were obtained when the mouth of the pipe was kept 3 or
4 feet below the surface of the sand, into which it tended
to bury itself too deeply. The total cost of pumping up
the sand in the entrance harbour, and conveying the material
2 miles out to sea in hopper barges, was 7*230?. per cubic
yard, the cost of each sand-pump dredger being £5,000.
The same work accomplished by a bucket-ladder dredger,
delivering into steam hopper barges, cost altogether 8-33^/.
per cubic yard, the original cost of this dredger having been
£25,000.
A smaller sand-pump dredger, costing only £2,000, and
having a fan 2 feet in diameter, and a i-foot suction pipe,
raised on the average 200 tons of gravel, sand, and stones in
Lowestoft Harbour per hour, which were transported 2 miles out
1 Minutes of Proceedings Institution C. E., vol. lxii, p. 23, and plate 3, figs. 7,
8, and 9.
86 Sand Pumps working on Mersey Bar. [chap.
to sea, at a cost of only 2d. per ton1, which, allowing for
interest and depreciation, would not amount to 3\d. per
cubic yard raised and deposited.
An illustration is given on Plate 2, Fig. 8, of one of
the sand-pump dredgers employed in deepening the channel
across the bar of the Mersey in Liverpool Bay (Plate 7, Figs.
11 and 12). A centrifugal pump and suction pipe have
been fitted to a 500-ton steam hopper barge ; and the
vessel, under favourable conditions, can raise six or seven
loads in a day and deposit them 2 miles away from the
bar, the filling of the hopper sometimes occupying only
half an hour. Two of these dredgers removed 750,000 tons
of sand from the bar in the year ending June 30, 1892,
though during three months of this period only one of the
vessels was at work. A much larger dredger of the same type
has been constructed for the Liverpool Dock Board, with
a hopper having a capacity of 3,000 tons, which it is able to
fill in three quarters of an hour, in order that the deepening
of the bar channel may be pushed forward more rapidly2.
Sand-pump dredgers have been most advantageously em¬
ployed in deepening the approach channels to some North
Sea and Channel ports, across the sandy foreshore in front
of their entrances, more especially at Dunkirk, Calais, and
Ostend. At Dunkirk, the work was at first let by contract
at i s. 9d. per cubic yard raised and deposited 2¿ miles off,
and the price was eventually reduced to 7-28d. ; but as this
price still threw an excessive burden on the port, the engineer
took the work in hand with three dredgers, each costing
£5,600, provided with a pump 5 feet in diameter, and having
a hopper capacity of 315 cubic yards3, whereby the total
cost was reduced to about 2- \d. per cubic yard allowing
for interest and depreciation.
1 Minutes of Proceedings Institution C. E., vol. lxii, p. 45.
2 ' The Sand Dredger Brancker,' A. Blechynden, International Maritime Congress,
London, 1893, Section III, p. 52.
3 Minutes of Proceedings Institution C. E., vol. lxxxix, p. 80.
iv.] Machines for Eroding Shoals. 87
Sand pumps possess the advantage of reducing the lift
of the material to a minimum, as the outlet of the discharge
pipe has only to be raised just high enough to deliver into
the hopper. On the other hand, they have to raise a large
proportion of water with the material, though the actual lift of
this water is only the height of the discharge orifice above the
water-level. Nevertheless, in spite of this disadvantage, they
are able to raise sand very economically in exposed situations.
Eroding Machines. Various contrivances have been tried,
having for their object the erosion, or stirring up of sand or
silt, so that it may be carried away in suspension by the current.
Rakes or harrows have been dragged over shoals, so as
to bring the material forming them under the influence of the
current. This system, when applied to the bars at the mouth
of the Danube, and of the Mississippi, failed to produce any
marked increase in depth, owing doubtless to the enfeebled
current being already fully charged with sand and silt, which
it gradually deposits as its velocity is reduced on entering the
sea. A rake, however, was used with success for cleansing the
river Stour earlyin the nineteenth century1 ; and a similar device,
dragged by a steamer over shoals of shingle in the Danube, in
1879, where the velocity of the current reaches 5 miles an
hour, enabled the river to carry down the loosened shingle,
and thus lowered the shoals about 3! feet2. Scrapers fixed to
a frame, and towed by a steamer, have also been used for
increasing the depth of the Upper Mississippi, the Missouri,
and other rivers of North America : and revolving cylinders
armed with spikes, drawn by horses, or by a barge hauling on
a fixed rope, are used for removing weeds and stirring up the
mud in the rivers and drains of the Fen country.
Screw propellers have been employed, both in Europe and
America, for stirring up sandy shoals, the particles of which
are carried down in suspension by the current.
1 Transactions Institution C. E., vol. ii, p. 181, and plate 15.
2 Minutes of Proceedings Institution C. E., vol. lx, p. 387, and plate 17.
88 Shoals removed by Eroder, and Water-Jet. [chap.
A machine, called an ' eroder,' has been used in the river
Witham for disintegrating and churning up compact material,
so that it can be carried down, in a fine state of division, by
a feeble current having a velocity of only ii miles an hour.
The eroder consists of a circular cutter revolving very rapidly
at the end of a vertical shaft, which is lowered to the bottom
of the river from a steam barge carrying the engine which
works the eroder1. It resembles in principle the disintegrating
apparatus of the von Schmidt sand-pump dredger.
Jets of water, or of air under pressure, have been directed
against muddy deposits and sandy shoals, driving the material
upwards into the current which carries it along in suspension.
All these contrivances possess the advantage of leaving the
actual removal of the material, which has been loosened or
put in suspension, to be effected by the agency of the current.
Accordingly, the system is economical for obtaining small
increases in depth when the current is powerful, or the material
very light ; but it is not adapted for removing dense material
when the current is feeble, nor suited for a river heavily charged
with detritus. Moreover, unless the river has a surplus of depth
lower down, the material transported is liable to form a shoal
elsewhere in settling when the velocity of the current
diminishes. The eroding principle is, therefore, only applicable
under special local conditions, and where the available funds
for the improvement of a river do not admit of more systematic
measures, or where the requirements of navigation are very
limited.
Remarks on Dredgers. The bag and spoon are still service¬
able where the depth is slight, where the amount of work to
be done is very small and scattered about, and where funds do
not admit of a large expenditure on plant.
The bucket-ladder dredger is best suited for carrying out
extensive and continuous dredging operations, such as have
1 ' The Transporting Power of Water, with a Description of the Eroder,'
W. H. Wheeler.
IV.] Relative advantages of Blicket Dredgers. 89
been conducted for a number of years on the Tyne, the Clyde,
the Tees, and many other rivers. It can dredge materials of
very varied compactness, ranging from silt to boulder clay and
soft rock ; and though it absorbs a large amount of power in
working, owing to its cumbrous machinery, and the height to
which it has to lift the material to deposit it into the sloping
shoot, it is able to perform large and regular dredging works
economically and expeditiously. The cost of such plant is
necessarily large, and consequently the system is not well
adapted for a small amount of work, unless the plant can be
hired temporarily, which is generally a somewhat costly pro¬
ceeding. The choice between a stationary dredger with hopper
barges, and a hopper dredger, depends upon circumstances :
the former plant is more costly, but it can work more
continuously, provided the dredger is allowed to ground at
low water, and is therefore best for a large quantity of work.
The hopper dredger being self-contained is cheaper ; it requires
fewer hands ; and it is convenient for a moderate amount of
dredging, and for the relatively small amount of maintenance
which may be subsequently required. The bucket-ladder
dredger is not able to dredge in considerable depths of water,
being limited by the length of its ladder ; and it cannot work
when subjected to wave motion. This form of dredger, however,
is more extensively used than any other ; and it has rendered
inestimable services in river, canal, and harbour works.
The dipper dredger has been largely used in America ; but
though it can deepen a channe uniformly, it cannot work as
continuously as the bucket-ladder dredger, or dredge at the
same rate. Accordingly, the dipper dredger is better suited
for dealing with detached shoals than for deepening a long
length of river ; and as it can only work with its full effect
when making a cut about 4 feet deep, it is not adapted for
the ordinary work of maintenance. It can dredge better
on a rough irregular bottom, with abrupt variations in depth,
than the bucket-ladder dredger ; but, like that dredger, it
9o Capabilities of Grab and Sand Pump. [chap.
cannot adjust its operations to the rise and fall of its vessel,
resulting from waves or a swell ; and the depth of its working
is limited by the length of its handle.
The grab dredger is specially advantageous in raising rough
obstructions, or pieces of blasted rock ; and it can work in
any depth, and in exposed situations. Moreover, it can
dredge in confined spaces, and close to quay walls ; whilst
its cost is comparatively moderate. Its vertical, discontinuous
action unfits it for regular deepening operations in rivers ;
but it can be usefully applied to the removal of detached shoals
of moderate extent. The grab can readily dredge sand,
gravel, or other loose material ; but it requires to be made
specially heavy, and to be furnished with claws, to penetrate
compact or stiff ground, which can be more efficiently dredged
by a bucket-ladder or dipper dredger.
The sand-pump dredger is capable of raising pure sand,
in exposed situations, very cheaply ; and it can even lift
compact material, if this material is previously disintegrated
by a cutter. Where the material is favourable, the sand pump
can dredge more economically than the ladder dredger; and it
can deepen the entrances to ports, and the channels over bars,
in places where ladder and dipper dredgers would often be
unable to work. Moreover, sand-pump dredgers are generally
cheaper than ladder dredgers ; and sometimes a bucket-ladder
dredger carries a sand pump in addition, so as to be able
to dredge the various materials met with in the course of
a long river, by the most expeditious and economical method,
thereby increasing the capabilities of the dredger.
Removal of Roek under Water. As dredging machines are
unable to penetrate hard rock, it is necessary first to shatter
the rock, and then to raise the loose material by dredgers.
The breaking up of the rock is generally effected by drilling
holes in the reef from a barge, and then blasting with
dynamite, a method which was adopted for lowering a rocky
shoal in the navigation channel of the Tees estuary. The
IV.] Rock-breaking Rams on Suez Canal. 91
removal of two extensive reefs in New York Harbour, to
a considerable depth, was accomplished by driving a net¬
work of galleries several feet below the surface of the reef,
and then blowing up the roof and pillars of these galleries
by the simultaneous firing by electricity of a great number
of blasting charges, after which the débris was raised by
grab dredgers1. These systems, however, are tedious and
expensive ; and the latter is only applicable to the ex¬
ceptional case of a large area of rock having to be removed
to a considerable depth.
The widening of the Suez Canal has necessitated the
removal of a considerable quantity of rock under water,
where the canal passes through rocky strata between the
Bitter Lakes and Suez, which has led to the designing of an
apparatus for shattering the rock without using explosives2.
This apparatus consists of a long heavy iron cutter, with
a steel chisel-edge at the bottom, which is let fall from
a height of 5 to 20 feet on to the rock, and is raised
again by hydraulic power for a repetition of the blow
(Plate 2, Fig. io)3. The machine used on the Suez Canal
consists of a bucket-ladder dredger, with staging erected near
the bow, from which five rock-breaking rams, or cutters,
each weighing 4 tons, are worked ; and these rams shatter
the rock by a succession of blows, and the débris is then raised
by the buckets. Sometimes these rock cutters, when placed
singly, or two or three on a barge, are made 8 tons in weight ;
and they can give about thirty-five blows per hour, breaking
up about 2 cubic feet of rock per blow. This system is also
being used for the removal of a portion of the rocky barrier
across the Danube, known as the ' Iron Gates.'
1 ' Blasting Operations at Hell Gate, New York,' L. F. Vernon-Harcourt.
Minutes of Proceedings Institution C. E., 1886, vol. lxxxv, p. 264.
2 ' The Removal of Rock under Water without Explosives,' F. Lobnitz. Minutes
of Proceedings Institution C.E., 1889, vol. xcvii, p. 369.
3 The illustration of three rock-breaking rams on a barge was taken from some
drawings furnished me by Messrs. Lobnitz & Co. of Renfrew.
92 Blasting ' Iron Gates J on Danube. [chap.
The remainder of the obstruction forming the ' Iron Gates '
is being perforated by rock drills worked by steam, inside
iron tubes resting on the rock. When the hole has been bored
to its full depth, it is cleared out by a jet of water, and
a dynamite cartridge is inserted, and a series of charges
are exploded by an electrical battery. During the boring
of the holes, the barge carrying the apparatus is raised slightly
out of water, resting upon four props standing on the bed of
the river, so as to be secure from wave motion h This boring
apparatus has for many years given good results in the
removal of rocks in the St. Lawrence.
Discharge of Dredged Materials.
The actual cost of dredging depends upon two distinct
operations ; first, the cost of dredging and raising the material
to the surface ; and, second, the cost of conveying it to its
destination. It has been already indicated that, in certain
circumstances, economy or convenience may be attained by
accomplishing both operations with the same vessel, by using
a hopper dredger. Generally the dredged material, having
been discharged into barges alongside the dredger, is conveyed
out to deep water and deposited through a hopper ; or, if
the site is at a distance from the sea, it has to be deposited
at a suitable place along the banks, by manual labour.
Discharging by a Chain of Buckets. The cost of unloading
barges by manual labour is necessarily large ; and if the
material is very soft, the operation is difficult. Accordingly,
on the Danube regulation works, at Vienna, the material
was removed from the barges by a chain of buckets, similar
to those employed in bucket-ladder dredgers, supported on
a staging on the bank, and was discharged into wagons running
1 'La Régularisation des Portes de Fer,' Béla de Gonda, Congrès international
de Navigation intérieure, Paris, 1892.
iv.] Discharge of Dredged Materials. 93
on to the staging1. The same system was also adopted at
the enlargement works of the Ghent-Terneuzen Canal. At
the Panama Canal works, a floating bucket-ladder dredger
raised the dredged material out of the barges and discharged
it into a long shoot; and similar plant is at work on the Maas.
Discharging direct into Wagons. Another plan of reducing
the cost of transport of dredged materials, adopted a few years
ago in France, consists in discharging the material, brought
up by the buckets, direct into wagons placed on the barges,
instead of into the barges themselves2. Having erected
«
a suitable landing stage, the wagons on reaching the shore
can be at once run off ón to a railway, thus dispensing with
the double shifting of the material necessitated by the
ordinary processes. A saving of id. per cubic yard was
effected by this system.
Discharging through a Long Shoot. Where the dredged
material has only to be conveyed a short distance to the
side banks, special contrivances have been adopted with great
advantage. Thus on the Suez Canal works, long shoots
were employed for conveying the dredged material, as it
fell from the buckets, away to the banks. The shoots were
.made as much as 230 feet in length (Plate 2, Fig. 7). They
✓
were semi-elliptical in section, 5 feet wide, and were hinged
to the dredger to admit of an alteration in their inclination,
which was accomplished by raising or lowering the supports
■
on the lighter between the dredger and the bank. Mud was
carried away with hardly any inclination of the shoot ; sand
mixed with water was conveyed away with an inclination of
i in 20 ; but in dealing with sand and shells, or clay, a pair
of endless chains were employed, together with a stream of
water, for dragging the material down the shoot.
Similar long shoots were employed on the Panama Canal
works where the banks were low enough. These steel shoots
were 131 feet long, with an inclination of 1 in 20, and were
1 Annales des Ponts et Chaussées, 1875 (1), p. 405. 2 Ibid. ï88o (1), p. 29.
94 Dredgings discharged by Floating Tubes, [chap.
supported by steel wire cables passing over sheer-legs erected
on a vessel fastened alongside the dredger on the side of the
shoot, and counterpoised, on the other side, by a vessel also
fastened to the dredger and loaded with ballast1. The
material was carried along the shoot by a stream of water,
and was discharged n feet above the water-level. Long
shoots have also been made to extend 200 feet out on each
side of the dredger, supported by wire cables from sheer-legs
erected on each side of a wide barge carrying the dredger.
Discharging by a Pump and Floating Tubes. The material
dredged from the sandy bottom of Lake Y and Wijker Meer,
in constructing the Amsterdam Ship-Canal, was discharged
on to the side banks by mixing it with water, and forcing it,
in a semi-fluid state, through long wooden tubes floating on
the surface of the water. A Woodford pump, 3% feet in
diameter and performing 230 revolutions per minute, was
fixed to the side of the dredger at the level of the water,
having a reservoir at the top in which the dredged material
falling out of the buckets was received, and an inlet at the
bottom for admitting the water. The material, mixed with
water, was forced by the pump through the tubes, which were
15 inches in diameter, and were buoyed up by floats on the
surface. The line of tubes, connected together by leathern
joints, extended in some cases to a distance of 300 yards ;
and the material could be delivered 8 feet above the water-
level2. A large dredger was thus enabled to excavate, and
deposit on the banks, about 1,300 cubic yards per day of
twelve hours, at a cost of about 2d. per cubic yard.
A similar contrivance was adopted at the Ghent-Temeuzen
Canal enlargement works ; but in this case a centrifugal
pump was joined on at about the middle of the floating
tube, which increased the discharge, and raised the material,
consisting of sand and silt, to a height of 28 feet, whence it
1 ' Le Génie Civil,' vol. v, p. 17.
2 Minutes of Proceedings Institution C. E„ vol. lxii, p. 6.
IV.] Excavators employed for Ship-Canals. 95
was led into a shoot which conveyed it considerably further1.
By the addition of a second pump to the tube, the material
was eventually discharged about five furlongs away from
the dredger.
At the same canal works, a combination of the pump with
a long shoot was also used for conveying the dredgings
a distance of about 180 feet, and discharging them at a height
of 13 feet above the water-level. The material was forced
through the supported tube, by a pump discharging a rapid
stream of water amounting to about three times the volume
of the material.
Excavators.
The main portion of the excavations for new river channels,
canals, basins, and docks, has generally to be executed before
the water can be admitted ; and dredging is only available for
forming the entrance channels and the waterway through lakes,
and for the final deepening and maintenance. Accordingly,
machines have been designed for effecting large excavations
on land, on the same principles that dredgers accomplish them
under water. Some excavators were designed for the Suez
Canal works ; and though the Panama Canal works have little
to recommend them in their present stage, they undoubtedly
gave a stimulus to improvements in excavators, from which
subsequent works, like the Manchester Ship-Canal, have
largely profited.
Bucket-Ladder Excavator. A machine was employed on
the Suez Canal works, which carried an excavating apparatus
like the chain of buckets of a dredger, and moved on wheels
like a locomotive, being propelled by one steam-engine, whilst
another worked the buckets. Seven of these excavators re¬
moved about eight million cubic yards on the Suez Canal ;
1 ' Le Canal de Terneuzen-Gand, et ses Installations Maritimes.' O. Brnneel and
E. Braum.
g6 Bucket-Ladder Excavators on Canals. [chap.
and similar machines were subsequently used on the Danube
regulation works, the enlargement of the Ghent-Terneuzen
Canal, and the Tancarville Canal. These excavators, sup¬
ported on four axles with three wheels each, ran on three
rails laid on the surface of the ground near the edge of the
top of the cutting, and excavated slices off the slope, like
a plane, with their ladder projecting down the cutting, and
inclined at the angle of the slope. The buckets raised the
material and deposited it in wagons running on a line of rails
alongside the excavators ; and one of these excavators could
remove about 3,000 cubic yards from the sandy cutting of
the Suez Canal, on the average, in a day. Several excavators,
similar in type, though differing in details, were employed on
the Panama Canal works. Seven bucket-ladder excavators, of
French and German make, have been used on the Manchester
Ship-Canal, each costing about ,£2.400, and weighing 70 to
80 tons1. Their buckets cut the slope inwards towards the
machine ; and being merely curved round at the bottom, and
open towards the chains, they discharge their contents down¬
wards into the wagon standing below in turning over the top
tumbler. Each of these excavators can remove 1,400 to
1.500 cubic yards per day on the average, the maximum
amount raised in a day being 2,500 cubic yards. The
excavation with these machines has cost about 1 \d. per
cubic yard, exclusive of interest and depreciation. These
excavators, however, whilst working very economically in
light soft soils, have proved unable to deal with heavy or
stiff materials, such as soft rock, stiff clays, boulders, or
embedded trunks of trees. Owing to their great weight,
heavy steel rails and stout sleepers are required to carry
them ; and the road needs frequent repairs, especially on
soft ground in which some of the machines are in danger of
settling and tipping over into the cutting, an accident which
1 'Mechanical Appliances employed in the Construction of the Manchester Ship-
Canal,' E. Leader Williams. Minutes of Proceedings Institution M. E., July, 1891.
IV.] Steam Navvies on Panama Canal.
97
occurred once or twice on the canal works. Moreover, they
are expensive to remove from one cutting to another. They
possess, however, the advantage of lifting the material out of
the cutting, and thus reduce the cost of haulage.
Steam Navvy. The type of excavator which works a large
cylindrical bucket, with movable bottom, at the end of a long
beam, like a dipper dredger, is known generally as a steam
navvy, and is extensively employed in large canal and dock
excavations (Plate 3, Fig. 9) h Sometimes the bucket is
worked from a strongly framed revolving jib, as shown in the
illustration ; and in smaller types, the bucket is connected to
the jib of a revolving steam crane. The advantages of the
large steam navvy are its considerable capacity for work, and
its power of excavating all kinds of material except hard
rock ; and the smaller machine has the merits of comparative
lightness, so that it can run on the ordinary temporary roads,
smaller cost, and the availability of the steam crane for other
work when not required for excavating.
A large steam navvy of American manufacture, running
on a bogie frame with eight wheels, and armed with a steel
bucket of 3^ cubic yards capacity, was used in the deep
Culebra cutting of the Panama Canal, and excavated 530
cubic yards per day under difficult conditions, its capabilities
being stated to be five times as great2. An English steam
navvy also employed at the Panama Canal, weighing 32 tons
and running on two pairs of wheels, with a bucket holding
cubic yards, excavated nearly 800 cubic yards per day
on the average3. Altogether, not less than seventy-two ex¬
cavators of different kinds were at work on the Panama Canal
shortly before the stoppage of the works, thirty-nine of which
were excavating in the Culebra section only a little over two
miles in length.
1 The illustration of a Steam Navvy has been reduced from a drawing furnished
me by Messrs. Huston, Proctor, and Co. of Lincoln.
2 'Le Génie Civil,' vol. v, 1S84, P' 393' 3 Ibid. vol. xiii, 1888, p. 131.
H
98 Excavators on Manchester Ship-Canal, [chap.
About one hundred excavators were in use at the same
time on the Manchester Ship-Canal, fifty-eight of which were
steam navvies of the larger type, costing about £ 1,200 each.
The buckets of these steam navvies have a capacity of from
11 cubic yards for boulder clay, up to i\ cubic yards for
sand ; and the average output of one of these excavators is
600 to 700 cubic yards per day of ten hours, though a machine
with the largest bucket has excavated a,000 cubic yards in
a day when working in sand. These navvies can excavate
a trench about 35 feet deep, and from 50 to 60 feet wide at
the top. They travel on a gauge of io| feet; their weight
of 32 tons necessitates a very firm road ; and they tend to
settle on soft ground. The steam-crane navvy is readily
moved, it can turn right round and excavate in any direction,
its average output is from 400 to 600 cubic yards per day,
and its cost is £800 to ^"1,050. The excavation with steam
navvies on the canal cost about 2d. per cubic yard.
G-rab Excavator. As the grab dredger is commonly
worked by a crane placed on a barge, it is equally adapted
for working on land from a movable steam crane. The
grab excavator is advantageously used for raising dredged
material from barges, and depositing it on land at the back
of a quay wall, and also for excavating in cramped spaces ;
but its inability to penetrate easily hard or stiff material
renders its application limited compared with the other forms
of excavators. These machines were used on the Manchester
Ship-Canal, mainly for excavating foundations, and for
removing materials under water ; and in suitable soils, they
have been able to raise about 300 cubic yards per day. The
grab bucket is a comparatively cheap machine ; and it can be
readily worked by a steam crane, which is available at any
time for other purposes, and can also be used for loading or
unloading vessels when its work of excavating is terminated.
Remarks on Excavators. The steam navvy is specially
adapted for opening out the advanced trench of a cutting
IV.] Capabilities of Different Excavators. 99
from the bottom ; whilst the bucket-ladder excavator serves
best for enlarging the cutting from the top. Some bucket-
ladder excavators, however, were made for the Panama Canal
works, in which the ladder of buckets, placed nearly hori¬
zontally. projected in front of the truck which carried it at
the bottom of the cutting, so that these machines could
open out the trench of a cutting, 30 feet wide, like a steam
navvy ; but though possessing the advantage of working in
a continuous manner, they were less stable than a steam
navvy when the ladder was turned round towards the side of
the cutting. The grab excavator is only suited for special
work in loose soil in limited spaces ; the bucket-ladder ex¬
cavator has a considerable capacity for work in moderately
soft or loose strata : whilst the steam navvy can excavate any
material which is not compact enough to require blasting.
Steam navvies and bucket-ladder excavators are indispens¬
able for expediting the work of deep extensive cuttings,
such as are involved in the construction of ship-canals ; and
they are especially valuable where manual labour is scarce or
costly, and more particularly where an unhealthy climate
entails peculiar risks to human life, as was encountered at the
isthmus of Panama.
H 2
CHAPTER V.
LOCKS; AND FIXED, AND DRAW-DOOR WEIRS.
Locks:—Description of Parts ; Construction, varieties ; Lock-Gates, single and
double, wooden and iron, strains, support, movement. Weirs:—object, classi¬
fication. Fixed Weirs:—Description, disadvantage; Oblique Weirs, object;
Angular and Horse-shoe Weirs, advantages, forms; Construction; Anicuts,
methods of construction, instances, defects. Draw-door Weirs:—Description,
object; Teddington Weir; Indian Dams; Draw-doors sliding on Free Rollers,
instances, effects, for regulating tidal flow on Thames and Weaver ; Segmental
Gate Weir, description, object, instances. Concluding Remarks.
The object and advantages of the canalization of rivers
have been explained in Chapter III ; and it is proposed in
the present chapter to describe some of the works by which
this canalization is carried out, which consist essentially of
weirs which keep up the water-level of a river in successive
reaches, and locks which enable vessels to pass from the
water-level of one reach to the water-level of the adjoining
reach.
Locks.
A lock contains a chamber, placed at the junction of two
reaches, in which the water can be raised or lowered so as
to be on a level with either the upper or lower reach. The
lock-chamber is usually closed by a pair of gates at each
end ; it is filled by letting in water from the upper reach
through sluices in the upper gates or side walls, and emptied
by letting it out through similar sluices at the lower end.
A vessel, admitted into the lock when the level of the water
in the lock is the same as in the reach which the vessel
Parts of Locks described. 101
has just traversed, can be thus easily raised or lowered to the
level of the adjoining reach.
Description of a Lock. The parts of a lock are shown
on Fig. xi. They consist of two side walls, A A, with
curved ends to facilitate the entrance and exit of boats ; and
two pairs of gates, BB, at each end of the lock-chamber,
C, shutting at the bottom against a sill, DD. The bottom
of the lock-chamber is covered by the invert, E, which, as
its name implies, is generally made in the form of an inverted
arch to support the side walls, exposed as they are to
constant variations in pressure, and to secure the chamber
against injury or leakage from below. The portion of the
floor of the lock, FF, over which the gates move is called
the gate-floor. Beyond the gate-floor, at each end of the
lock, is the apron, G G, which is generally protected from
scour by pitching or masonry. The vertical post on which
each gate turns, called the heel-post, stands upon a stone
called the heel-post stone, and fits into a vertical curved
groove in the lock wall, formed by courses of carefully
dressed stones, which from their shape are called hollow
quoins (Fig. 12, page 105). A recess, HH, is formed in
the lock wall, on each side of the gate-floor, into which
io2 Forms and Construction of Locks. [chap.
the gate is drawn, when opened, so that it may be out of
the way of boats passing through the lock ; it is called the
gate recess, and is terminated at one end by the hollow
quoins, and at the other end by another set of dressed stones
called square quoins.
In locks leading to docks, or on rivers where the fall
of the bed is not abrupt, the upper and lower sills are placed
at the same level ; but on canals, and on rivers where the
original bed is altered, the upper sill is placed on the top
of the lift wall, /, at the level of the bottom of the upper
pool, and the depth of the upper pair of gates is corre¬
spondingly reduced.
Construction of Locks. Formerly locks were constructed
of timber, or sometimes with masonry walls at each end, and
merely a cutting with sloping sides for the lock-chamber, the
slope being frequently terminated about the low-water level
by a line of sheet-piling. Timber, however, is liable to
decay, and to be washed away by the rush of water from the
sluices ; and the sloping sides of the chamber cause a great
waste of water in locking. Vertical walls of brickwork or
masonry are, accordingly, often adopted throughout ; and
in large locks, the invert, gate-floors, and aprons, are con¬
structed of masonry, and the sills and hollow quoins of
granite. Locks are made just large enough to admit the
largest class of vessels employed on the special navigation ;
but where the size of the vessels varies considerably, an
intermediate pair of gates is sometimes introduced, so that
a shorter lock-chamber may be provided for small vessels,
thus saving both time and water in locking. Suresnes lock,
the first lock on the Lower Seine below Paris, furnishes
an instance of this arrangement, the narrower of the two locks
being provided with an additional pair of gates, dividing
the lock into two portions, 133 feet, and 331 feet long
respectively. The large traffic, however, on the Lower Seine
has allowed the same object to be accomplished at other
v.] Sizes of Locks ; and Lock-Gates. 103
places by providing locks of different sizes, as for instance
at Poses, where the three locks are of different lengths and
widths (Plate 4, Fig. 4). In the locks at both ends of the
Amsterdam Ship-Canal, both methods are resorted to for
accommodating the variety of traffic (Plate 12, Figs. 3 and 4).
The five locks on the canalized Main, below Frankfort,
have been recently greatly lengthened by adding a third
pair of gates in the cut below the original lock, providing
a second lock, 820 feet long, capable of admitting six of
the largest Rhine boats two abreast, with their tug occupying
the original lock. These smaller locks, 279 feet long and
34! feet wide, will still serve to accommodate the steamers
of 1,300 tons navigating the Rhine, for which they were
designed (Plate 4, Fig. 5). When a lock is intended to hold
a double row of large vessels, the entrance is only made
wide enough to admit the broadest vessel, and the lock-
chamber is widened out.
Lock-Gates.
The smallest locks are closed by a single gate at each
end. There is, however, an advantage in closing the lower
end of the lock with a pair of gates, as a single gate requires
more space for opening, and thus either reduces the available
length, 01* necessitates an increased length of lock, and a cor¬
responding increased expenditure of water in locking. Except
in small locks, both ends of the lock-chamber are closed
by a pair of gates, each gate being rather longer than half
the width of the passage, and meeting at an angle in the
centre line of the lock. Each gate turns on a heel-post,
and has a meeting- or mitre-post fixed at its other extremity,
which, when the gates are closed, shuts against the mitre-
post of the opposite gate. The mitre-posts of the two gates
are formed so that their meeting surfaces may be in close
contact when the gates are closed ; and a horizontal sill-
piece at the same time shuts closely against the sill.
io4 Materials and Strains of Lock-Gates. [chap.
Wooden Lock-Gates. Lock-gates across small locks, and
in a moderate depth of water, are generally made of wood ;
and at the Amsterdam Canal, wooden gates have been
adopted for the inner gates of the locks having a width
of 60 feet ; but in this instance the variations in the level
of the water are slight, and consequently the pressure is
moderate. English oak is the best material for wooden
gates ; but where the gates are liable to be attacked by the
teredo navalis, greenheart timber must be used, though it
is costly to work and liable to split. The Manchester Ship-
Canal lock-gates have been constructed of greenheart.
Wooden gates are formed of a series of horizontal beams,
connecting the heel-post and mitre-post ; against which
vertical or diagonal planking is fastened, and, being made
watertight, acts as the skin of the gate. The beams are
placed closer together towards the bottom of the gate, in
order to sustain the increased pressure due to the greater
depth of water.
Iron Lock-Gates. Iron gates have generally a double skin
of wrought-iron plates, riveted to horizontal ribs or girders ;
and they are strengthened in addition by vertical ribs. The
skin plates are increased in thickness, and the horizontal
ribs are placed closer together and strengthened towards the
bottom of the gate, to allow for the increased strain. The
lock-gates of the Amsterdam Ship-Canal pointing towards
the sea, have been constructed of iron, to afford additional
security against any inroad of the sea.
Strains on Loek-Gates. The strains on a single straight
gate shutting across a lock result merely from the pressure
of the head of water against its up-stream face, which in¬
creases in proportion to the depth. The horizontal beams
of the gate act like girders, supporting an evenly distributed
load which imparts a strain equal to half the whole pressure
exerted at the centre.
When, however, two straight gates meet at an angle in the
v.] Strains and Rise of Lock-Gates. 105
centre line of the lock, each gate, besides being subjected
to the water pressure on its inner face acting as a transverse
strain, has also to bear a compressive strain in the direction
of its length, due to the pressure of the opposite gate
against its mitre-post. This latter strain is equal to half
the water pressure on the opposite gate, multiplied by the
tangent of half the angle between the lines of the two
gates when closed. This angle depends upon the proportion
which the projection of the sill, AB, bears to the span.
CD (Fig. 12), which is termed the rise of the gates. The
accompanying sketch shows a pair of gates having a rise
of one fourth; and the tangent of half the angle CAD, is
CB-^AB— 2, showing that the rise of a pair of gates
exercises an important influence on the compressive strain
transmitted by one gate to the other, which varies inversely
with the rise. Thus in the old Sparndam lock-gates, with
a rise of only one sixteenth of the span, the compressive
strain amounted to four times the water pressure on one gate.
On the other hand, a large rise increases the length of the
gates and reduces the available length of the lock.
'lledrJBost
Stone,
Fig. 12.
io6 Support, and Working of Lock-Gates, [chap.
Support of Lock-Gates. The weight of a lock-gate is
supported by the pivot on which the heel-post turns, and
by the anchor strap which encircles the top of the heel-post.
The pivot is fastened to the heel-post stone ; and a socket at
the bottom of the heel-post fits over the pivot. The anchor
strap is fastened to the anchor, formed of a cast-iron frame,
in the form of a sextant, bedded into the masonry and further
secured by tie bolts. A roller also is generally placed at the
bottom of large gates to assist in supporting the weight of the
gate. Its axis is placed in the vertical plane passing through
the axis of the heel-post and the centre of gravity of the gate ;
and the proportion it bears of the weight depends upon its
distance from the centre of gravity. In straight gates, the
roller is placed under the gate ; but in curved gates it is
partially clear of the gate, and its bearing is adjusted by
means of a vertical rod over it, at the back of the gate, which
is raised or lowered by a screw or wedges near the top of the
gate. The roller, which is slightly conical, runs on a cast-iron
roller-path fixed to the gate-floor (Fig. ia, p. 105).
Wherever a large portion of the gate remains always
immersed when it is opened and closed, as in the case of
the upper gates of locks, and the Amsterdam Ship-Canal
lock-gates, a roller may be dispensed with even for large
gates, owing to the buoyancy of the immersed gate, and
the small weight, consequently, which has to be supported
by the pivot.
Methods of moving Loek-Gates. The smaller wooden
lock-gates are opened and closed by pushing against the
balance beam, which has a heavy butt end projecting over the
side wall of the lock (Fig. 11, p. xoi), and is fastened to the
top of the heel-post and mitre-post, and serves to balance the
gate on the pivot. The larger gates are moved by chains
fastened to the gate, and worked by capstans or hydraulic
machinery. The employment of rollers renders it advisable
to attach the chains some way down on each side of the gate,
v.] Working of Lock-Gates : Weirs. 107
in order to facilitate the motion and to avoid straining the
gate. This necessitates the construction of chain passages in
the side walls for the opening and closing chains, with rollers
for directing and guiding them. When the gates are con¬
structed without rollers, the chains can be attached near the
top of the gate, and worked from the top of the side walls,
as at the Amsterdam Canal locks. A still better plan in such
cases, where there is ample space at the sides, is to fasten
a long, flat, toothed iron bar to the top of each gate, which
opens or closes the gate on being moved by a pinion on the
top of the side wall.
Further particulars relating to lock-gates, with illustrations,
will be found under the heading •Dock-Gates1' in the part
of 1 Harbours and Docks ' descriptive of docks, to which the
subject of large lock-gates more strictly belongs.
Weirs.
Weirs constitute the barriers which, placed at suitable
intervals in the bed of a river, raise the water-level of the
river, converting it into a series of steps of nearly level
reaches, the differences in level of which are surmounted by
means of locks which form the necessary adjuncts to weirs in
a continuous river navigation. The discharge of the river
passes over, or through the weirs ; whilst the vessels pass in
ordinary times through the locks. The weir stretches across
the main channel of the river ; and the lock is generally
placed in a minor channel separated off by an island, or in an
artificial side cut, so that the vessels may be kept clear from
the rapid current caused by the discharge of the river falling
over the weir. The former plan has been adopted on the
Lower Seine (Plate 4, Fig. 4); and the latter method has
been resorted to on the River Main (Plate 4, Fig. 5). On the
1 'Harbours and Docks.' L. F. Vernon-Harcourt, pp. 443-453, and Plates 15
and 16.
io8 Classification of Forms of Weirs. [chap.
Upper Seine, on the contrary, in the absence of islands, the
locks have been placed close to one of the banks, alongside
the weir ; but the security of the vessels has been provided
for in these cases, by removing the upper end of the lock
as far as practicable from the weir by putting the lower énd
in a line with the weir (Plate 4. Fig. 3).
Weirs may be conveniently divided into three classes ;
namely, 1. Fixed, or Solid Weirs; 2. Draw-door Weirs; and
3. Movable Weirs. No portion of the first type of weir is
removable ; and it forms a solid, permanent obstruction in the
channel of a river. Draw-door weirs, whilst affording an
equally efficient barrier as a solid weir during the low stage
of a river, can be more or less opened in flood-time to facilitate
the discharge of the river. Movable weirs serve equally as
well as solid and draw-door weirs to retain the water-level
of a river during the dry season ; and they possess the
advantage, not only of being able to regulate the flow of
a river, and to facilitate the discharge of flood waters like
draw-door weirs, but also of leaving the river entirely open
in flood-time.
Fixed Weirs.
A fixed weir is a solid wall, or embankment, placed across
a river, which affords no passage for the discharge of the
stream till the water has risen above it. This kind of weir
is very efficient in preserving a fixed minimum water-level
above it ; but it is incapable of regulating the flow, and
is not suitable for the rapid discharge of flood waters. As
the whole drainage of the district above, except the water ex¬
pended in locking, must pass over the solid weir, if no other
outlet is provided, such a weir, raised to the level required
for maintaining the proper navigable depth in dry summer
weather, diminishes considerably the waterway which existed
for the discharge of the river previous to its erection.
Endeavours have been made to mitigate this disadvantage
v.] Oblique Weirs on River Severn. 109
of fixed weirs. Sometimes the sill of the weir is placed
below the summer water-level, and is raised in time of drought
by placing planks along the top of the sill, which are kept
in place by upright stakes. Another method of facilitating
the discharge consists in making the crest of the weir longer
than the average width of the channel, by placing the weir
in a wide part of the river, or giving it a slanting direction
across the river.
Oblique Weirs. Some weirs erected across the Severn,
in 1842 1 (Plate 3, Figs. 7 and 8), were placed obliquely to
the stream, in wide places on the river, so that the length
of the weirs, and consequently the capacity for discharge
over them, was increased. The Severn has high banks and
a rapid fall in the part where the weirs were introduced,
so that solid weirs are less objectionable there than they
would be in rivers having low banks and a small fall. The
river, under these conditions, is able to rise a considerable
height above the top of the weir without overflowing its
banks ; and when it does so, the obstruction to the discharge
presented by the weir is proportionately less. The length
of the weirs on the Severn was made such that the section
of the waterway over the weir in flood-time was equal to
the sectional area of the river above ; but, owing to the less
favourable form of the shallower section over the weir, and
the impediment occasioned to the flow by the change in
form of the channel, and in the direction of the current
at the weir, the discharge must always be somewhat checked
at that part. Also, when the velocity of approach of the
current increases, as it does in flood-time, the advantage
gained in length by the oblique position of the weir is
somewhat neutralized ; for the filaments of the water tend to
flow over the sill of the weir in a direction approximating to
the axis of the river, instead of at right angles to the weir.
Simultaneously with the introduction of these weirs on
1 Minutes of Proceedings Institution C. E., vol. v, p. 340.
no Various Forms of Fixed Weirs. |chap.
the Severn, the channel of the river was greatly improved,
and numerous rocky fords were removed ; so that in spite
of the obstructions presented by the weirs, the discharge
of the river in flood-time was on the whole facilitated ; whilst
a better navigable depth in dry weather was secured by
the deepening of the river, and by the weirs.
Angular and Horse-shoe Weirs. As straight weirs placed
obliquely across a river tend to divert the main current,
and consequently the deepest channel, towards the bank
against which the up-stream end of the weir rests, other
forms of obliqueness are frequently given to fixed weirs.
They are sometimes constructed in two straight lines, joining
in the centre of the river at an angle pointing up stream
(Plate 3 Fig. 9) ; or sometimes a curved form is adopted,
convex on the up-stream side, so that a sort of horse-shoe
fall is formed. These forms, whilst increasing the length
of the weirs, direct the main stream into a central channel ;
but the clashing of the currents is liable to produce eddies,
and cause undue scour below the centre of the weir, so that
the bottom must be protected at this part.
The torrential River Lot in France was canalized many
years ago, for a distance of 185 miles, down to its confluence
with the Garonne, by over seventy fixed weirs in broad
places in the river, with some of the locks adjoining the
weirs, and some inside cuts1. Several mills had been
previously established on the river, retaining its waters by
dams ; and, moreover, the high banks of the river rendered
inundations exceptional, so that under these conditions
fixed weirs were unobjectionable. The earlier weirs erected
on the River Lot were made angular (Plate 3, Fig. 9) ; but
this form proved inconvenient for navigation, owing to the
eddies produced, forming back currents towards the lock
on the down-stream side. An oblique weir, inclined up¬
stream from the bank on one side of the river, and connected
1 Annales des Ponts et Chaussées, 1865 (i),p. 151, and plates 103 and 104.
v.] Methods of Construction of Fixed Weirs, hi
with the lock on the other bank by a portion of weir at
right angles to the stream, directs the currents in a more
satisfactory manner (Plate 3, Fig. 11).
Construction of Fixed Weirs. Fixed weirs are frequently
formed of a mound of rubble stone protected on the slopes
by pitching (Plate 3, Fig. 8). The up-stream slope is made
steep, and the down-stream slope very flat ; and the pitching,
or a layer of rubble stone, is continued beyond the foot
of the down-stream slope to protect the river-bed from
scour. The top of the weir is capped by a sill of timber
or stone. A row of sheet-piling along the line of the sill,
and sometimes also along the toe of the down-stream slope,
secures the weir from being undermined by the water (Plate 3,
Figs. 8 and 10). Sometimes the weir is formed by a thick
masonry wall, battered on both faces, and with a slight
depression towards the current on the top to facilitate the
passage of the water over the sill (Plate 3, Fig. 13).
Anicuts. Large fixed weirs have been erected across rivers
in India, in order to retain and divert the water for irrigation.
These weirs, termed anicuts in the Madras presidency, keep
back the water during the dry season ; and in flood-time
the river flows over the weir. The difference between the
low-water level and the flood-level is so great, that the weir,
whilst high enough to retain a sufficient quantity of water
in the dry season, is submerged to a considerable depth
during the rainy season, so that it offers less impediment
to the flow in flood-time than weirs placed across rivers
experiencing smaller variations in level.
In olden times the natives used to erect embankments of
sand or loose stones at the close of the rainy season, which
served to keep back the water when the river was low, and
were washed away by the first flood. These embankments,
accordingly, had to be reconstructed every year ; but as soon
as these works were entrusted to English engineers, they
were made more durable.
112 Description of Anicuts in India. [chap.
Anicuts are now generally constructed of dry rubble, con¬
crete, or rubble stone set in cement, protected on the face
by pitching or ashlar masonry set in cement. The up-stream
side of the anicut is often built nearly vertical ; and the down¬
stream side has a flat slope, terminated at the base by a stone
apron laid on the river-bed to prevent scour (Plate 3, Figs.
13 and 14). The foundations are varied according to the
nature of the river-bed ; but generally trenches or well-
foundations, filled with masonry or concrete, are carried
some feet below the bed, right across the river, under each
toe of the anicut, and sometimes also under the sill, to
secure the anicut from being undermined. Walls are built
on these foundations, and the intermediate spaces are filled
with loose rubble. The Okhla anicut furnishes an example
of a work in which no foundations were carried below the
bed of the river ; but it is very wide in cross-section,
having been considerably enlarged beyond the original
design h
Some of these anicuts are very long. For instance the
Godavery anicut across the River Godavery (Plate 3, Fig. 14),
with its sill 12 feet above the river-bed, and crossing four
separate branches of the river, has a total length of nearly
miles ; and the Dehree anicut across the River Sone
(Plate 3, Fig. 13), 8 feet high, is 2^ miles long.
Silt and sand brought down by these rivers tend to ac¬
cumulate above an anicut, raising the river-bed ; and sometimes
the deposit eventually chokes up the entrance to the irrigation
canal, which is usually constructed a short distance above the
anicut. Moreover, an anicut, like an ordinary fixed weir,
is not suited to discharge a sudden flood, or to modify the
levels of the water above it. Sluices, accordingly, closed by
gates or draw-doors, have to be constructed in anicuts to
scour away the silt and to carry off any excess of water.
1 'Professional Papers on Indian Engineering/ Second series, vol. vi, p. 239,
and plate 30.
v.] Forms and Object of Draw-Door Weirs. 113
Draw-door Weirs.
A draw-door weir consists of a row of piles, frames, or
piers, placed generally at regular intervals across a portion
of a river, with a series of panels, doors, or planks, to close
the spaces between them. These doors or panels, sliding in
vertical grooves in the piers, can be raised or lowered from
a foot-bridge above, thus opening or closing the weir.
Draw-door weirs, accordingly, provide a better discharge in
flood-time than overfall weirs ; they also furnish a means of
regulating the level of the water above, and of scouring the
channel near the weir.
In mill-streams, a fixed weir is usually placed across the
bank where the artificial cut to the mill diverges from the
old channel ; and the natural fall of the stream having been
reduced, and the sill of the fixed weir placed at a high
level in order to keep up the water above the mill, draw-
door weirs are provided to discharge the surplus water
when the mill-gates are closed, or in flood-time. The doors
of these weirs are generally formed of wooden panels sliding
in grooves between masonry side walls and intermediate
piles ; and they are raised by a rack and pinion, like the
sluice-gates in small lock-gates.
Similar draw-door weirs, on a larger scale, are commonly
placed across a portion of a river channel where a fixed weir
is situated, so as to regulate and increase the discharge.
Several of these weirs with large draw-doors, sliding between
piers and raised by chains passing round overhead pulleys,
have been introduced in the last few years in the weirs on
the Upper Thames, and have proved very valuable in re¬
ducing the height of the floods.
Teddington Weir. The weir on the Thames at Tedding-
ton, at the limit of the tidal flow, furnishes perhaps the
best example in England of a large composite fixed and
draw-door weir placed obliquely across the river. It is
i
114 Draw-Door Weir at Teddington. [chap.
divided into four separate bays, and has a total length
of 480 feet. Along the three down-stream bays, strong
wrought-iron frames, formed of plates and "angle-irons riveted
together, are fixed at regular intervals, and carry a foot¬
bridge on the top (Plate 3, Figs. 15 and 16). The two
side bays are fixed weirs. The two central bays, 17 a! feet
and 69! feet long respectively, are closed by large draw-
doors. The longer bay has twenty-three draw-doors, 6 feet
high and 7§ feet wide, formed of plate-iron strengthened
with angle-irons ; and the other bay has twelve draw-doors1.
The doors slide in vertical grooves at the sides of the frames.
A small tram-road is laid along the foot-bridge, on which a
crab runs which lifts the draw-doors by means of a chain.
The summer flow of the Thames passes over the fixed
weirs, which keep up the river to the requisite navigable
level ; and the discharge during the winter, or in flood-time,
is provided for by raising the draw-doors. The additional
opening thus provided does not extend down to the bed of
the river, for the draw-doors with their frames, are placed
on a rubble mound (Plate 3, Fig. 15) ; but this deficiency in
depth is compensated for by the great width of the weir.
Indian Dams. In constructing weirs for irrigation across
rivers in northern India, bringing down large quantities of
sand or silt, it has been found expedient to erect piers in the
central portion of the weir, with openings between them, about
10 feet wide, which are closed, on the approach of the dry
season, by planks or balks lowered horizontally, or by gates
sliding in grooves down the sides of the piers. The stone
piers rest on a masonry flooring carried across the river ;
and the weir is built solid for a certain distance from each
bank. These partially open weirs are called dams in
India. In more recent dams, the openings between the
piers have been made considerably larger, and are closed
1 ' Fixed and Movable Weirs.' L. F. Vernon-Harcourt. Minutes of Pro¬
ceedings Institution C. E., 1880, vol. lx, p. 25, and plate 1, figs. 11, 12, and 13.
v.] Draw-Doors sliding on Free Rollers. 115
by gates, turning on horizontal axes at the level ol the
flooring, which resemble a form of movable weir, and will
therefore be described in the next chapter. These dams,
whilst generally costing less than solid weirs, present less
impediment, when open, to the flow of the river, and
prevent the accumulation of silt and sand above the weir.
Free Hollers applied to Draw-doors. In raising a draw-
door, a considerable amount of friction is generally experi¬
enced at the surfaces of contact along which the two ends
of the door slide, especially when there is much pressure of
water against the door. This friction has been consider¬
ably reduced by making each end of the sluice-gate slide
upon a vertical row of free rollers hung on a suspended
chain in a recess in each side pier ; one end of the suspending
chain being attached to the sluice-gate, and the other end
to the side wall, causing the rollers to travel up and down
at half the rate of the gate (Plate 3, Figs. 17 and 18).
A watertight joint is made by a suspended rod at each end,
pressed against the side wall on the up-stream side of the
draw-door, so that the door itself merely rests against the
rollers.
Two large draw-door weirs on this principle have been
erected at Belleek and Ballinasloe in Ireland, across the
rivers Erne and Suck, to control the drainage works connected
with these rivers 1 ; and a half-tide weir has been erected
across the Thames a little below Richmond, which is closed
by three draw-doors of 66 feet span working on free rollers.
The Belleek weir, erected in 1883, consists of four openings
between masonry piers, closed by draw-doors, 31 feet long
and 14J feet high, and weighing 13 tons each, which can be
raised 9 feet above the sill on the bed of the river by one
man, against a water pressure of about 100 tons, by means
1 A few particulars about these weirs, or sluices, have been given me by their
designer, Mr. F. G. M. Stoney; and I have also had the opportunity of seeing
them worked.
I 2
ii6 Belleek, Ballinasloe, and Richmond Weirs, [chap.
of gearing placed on a girder bridge resting on the piers.
A turbine on the right bank, worked by the fall of water at
the weir, can raise the four draw-doors simultaneously. This
weir has enabled a rocky barrier to be removed from the
river ; and, whilst holding up the water-level in Lough Erne
and the adjacent Ulster Canal for navigation, this work has
reduced the rise of the floods from 10 feet to 7 feet by the
improved waterway provided, and by raising the draw-door on
the approach of a flood.
The Ballinasloe weir was erected in 1885, across the
up-stream side of an arched masonry bridge crossing the
river Suck, having four spans of 25 feet closed by draw-doors
sliding on free rollers, and counterbalanced so that they can
be very easily and rapidly raised.
The Richmond weir keeps up the Thames between Rich¬
mond and Teddington to half-tide level, in order to cover
the banks in mid-channel, which used to be laid bare at
low tide, owing to the considerable lowering of the low-
water line of the river within the last sixty years, resulting
from the removal of old London Bridge, old Westminster
Bridge, and other obstructions, together with dredging in
the river, facilitating the discharge of the ebb tide. Draw-
doors, 12 feet high, close the three central openings, of 66
feet, of the arched double foot-bridge across the river, when the
river falls below half-tide level ; and being raised from the bridge
as soon as the rising tide reaches that level, they leave the
influx and efflux of the tide unimpeded up to Teddington as
formerly above half-tide level. These draw-doors are counter¬
balanced, as at Ballinasloe, so as to be rapidly raised at the
proper time ; and they have to be raised and lowered every
tide, except when a descending flood keeps up the river
during low tide to the desired level. When raised, the doors
are rotated into a horizontal position, so as to be placed out
of sight between the footways, in order that they may not
intercept or mar the view along the river.
v.] Sluice-Gates on Rollers, Manchester Canal. 117
Sluice-gates working on the same principle have been
erected alongside the locks of the Manchester Ship-Canal,
to regulate the discharge of the rivers Mersey and Irwell which
join the canal, and also alongside the tidal locks at Eastham
to regulate the level of the tidal reach of the canal. The
largest work, however, of this kind on this canal is the
draw-door weir or sluices erected, along the outer side of
the canal, across the mouth of the river Weaver, with ten
openings of 30 feet span, closed by counterbalanced draw-
doors, or sluice-gates, capable of being raised 28 feet, which
whilst maintaining the water-level in the canal, have to
regulate the discharge of the Weaver so that its fresh water
and tidal flow, into and out of the Mersey estuary, shall take
place precisely as it did when the river flowed direct into
the estuary.
This type of draw-door weir is specially well adapted for
regulating the flow through it, and consequently the water-
level above it, owing to the great facility with which the doors
are raised and lowered, especially when counterbalanced ; but as
the discharge commences at the bottom of the weir, under
a head of water, the apron of the weir requires protection
against the powerful scour which takes place directly the door
is lifted off the sill.
Segmental Gate Weir. A peculiar form of weir, erected in
1853, may be seen across a branch of the Seine near the
centre of Paris. This weir of La Monnaie has four openings
of 28I feet, separated by masonry piers, each closed by
a wrought-iron cellular gate made in the form of a segment
of a cylinder. The gate revolves on a horizontal axis, round
pivots fixed in each pier, with which it is connected, at each
side, by six radiating iron rods (Plate 3, Fig. 19). The gates
descend into a curved depression in the apron when the weir
is to be opened, and are counterbalanced by weights rising
or falling in a hollow in the piers. They are by this means
easily tinned through a small arc, by chains worked from the
ii8 Segmental Gate Weir on Seine and Nile. [chap.
piers, closing or opening the weir according as they are raised
or lowered.
This form of weir was adopted with the object of its being
out of sight when the weir is open, and in order to dispense
with any intermediate frames which would be liable to arrest
floating rubbish, giving an unsightly appearance in the centre
of the city. Each gate weighs tons; and the cost of each
opening was £750. This weir, which retains the water up
to about 8 feet above the apron, has worked satisfactorily,
and has not at all impeded the passage of ice and floating
substances.
Similar gates were erected about the same period for
closing the weir erected across the Rosetta branch of the
Nile for the purposes of irrigation h This weir has sixty-one
openings, each i6J feet wide; and the gates, which, unlike
the Paris weir, were raised out of the water for opening the
weir, were intended to retain the water up to a height of
19 feet above the apron. The foundations, however, were
subject to settlement in the alluvial ground ; the water
escaped through gratings under the gates, placed there in
order to keep the gates clear of silt ; and the water could
never be retained to the desired height. Since 1884 the
foundations of this weir have been repaired, and the gratings
have been closed ; and for the future this weir, and the
Damietta weir, are to be controlled by iron draw-doors
sliding against free rollers.
Concluding Remarks. The discharge of most navigable
rivers generally suffices, even in the dry season, for supplying
water for locking, so that the chief objects in designing river
locks consist in adapting their sizes to the vessels using the
navigation, and in expediting the passage through the locks,
by providing intermediate gates or smaller locks for the small
vessels, and by constructing ample sluice-ways fitted with
revolving, or cylindrical sluice-gates, so that the lock may be
1 'Egyptian Irrigation.' W. Willcocks, p. 152, plate 20.
v.] Fixed and Draw-Door Weirs compared. 119
rapidly filled or emptied. The sills also of the locks in an
important, or developing navigation, should be placed at
a lower level than the existing available depth of the river,
in order that any subsequent deepening of the channel may
be effected without entailing the rebuilding of the locks.
Fixed weirs possess the advantages of requiring no looking
after, and of rarely needing repairs ; but they more or less
block up the natural channel of the river, and they only
maintain the river to a certain level above, without at all
regulating the discharge at any higher water-level. They
should always be placed in a wide portion of a river, so as
to ensure the greatest available capacity of discharge over
their crest ; but they are only properly suitable for torrential
rivers flowing between steep banks. Even when forming
irrigation dams across rivers which rise to a considerable
height in the rainy season, and in their natural condition
inundate large tracts of country, solid weirs are advantageously
supplemented by sluice-ways which can be opened, in order
that accumulations of deposit above the weir may be carried
down.
Draw-door weirs form important adjuncts to fixed weirs,
enabling the discharge to be regulated, and facilitating the
passage of flood waters. They are, indeed, more complicated
structures than fixed weirs, and necessitate superintendence
for opening and closing them. No canalized river, however,
subject to inundations, can be considered in a satisfactory
state which is devoid of appliances for regulating its water-
level, and of provision for the rapid discharge of floods at the
weirs. Draw-door weirs afford a method of accomplishing
these objects, but they generally only supplement fixed weirs,
their sills are commonly raised above the bed of the river,
and their doors are not lifted high enough to enable vessels
to pass through the weir in flood-time. Accordingly, draw-
door weirs, whilst furnishing a very valuable means of
regulating a river, and capable of being placed under perfect
i2o Déficiences in Draw-Door Weirs.
control by the adoption of free rollers for reducing friction,
generally extend across a portion only of the channel, and
are not suited for navigable passes. Moreover, the frequent
piers necessary for small draw-doors obstruct the channel ;
and the piers and overhead bridges required for large draw-
doors involve a considerable expenditure. Consequently,
draw-door weirs, though constituting a great advance on the
rough simple system of fixed weirs, do not afford a perfect
system of weirs for rivers in which it is important to leave the
channel as open as possible in flood-time, and in which the
navigation has to pass through the weir when the lock is
submerged, as for instance on the Seine, the Main, the Meuse,
and other large rivers.
CHAPTER VI.
MOVABLE WEIRS.
Definition; Classification. Frame Weirs:—Needle Weir, description, method
of working, instances, improvements, advantages, limit of height ; Sliding Panels,
object, advantages, instances ; Hinged Curtain, description and method of working,
compared with panels ; Suspended Frames, object, description of Poses weir, ad¬
vantages, cost compared with other forms. Shutter Weirs :—Bear Trap, description,
method of working ; Thénard's Shutters, description, working, modifications
adopted in India; Hydraulic Shutters, description and cost; Chanoine's Shutters,
modifications and working, introduction of foot-bridge, cost, adoption in America,
system of double grooves, advantages ; Tilting Shutter, description. Drum Weirs :
—Description, at Joinville, on the Main, at Charlottenburg, merits. General remarks,
summary of respective types, and comparison of systems.
A movable weir is a barrier placed across a river for
keeping up the water above it to any desired level, and
capable of being so lowered on to the bed of the river, or
removed, as to present no obstruction to the waterway in
flood-time.
The primitive stanches used for flashing on rivers, as
described in Chapter III, furnished the first rough types of
movable weirs ; and the needle weir, which is a common form
of movable weir at the present time, was first introduced
on the river Yonne for facilitating navigation by means of
flashes. The addition of locks at the weirs, for the purposes
of a continuous navigation, obviated the necessity of a frequent
opening of the weirs, or stanches, for the passage of vessels.
Nevertheless, movable weirs, besides serving for regulating
the discharge, are very valuable for enabling floods to pass
down in an unrestricted channel, and are in many cases
122 Classification ofi Movable Weirs. [chav.
indispensable for the passage of vessels through them when
the locks are submerged, and when the flood has not risen
so high as to put a stop to navigation. A movable weir is
generally divided by piers into two or more passes, one or
more deep passes near the lock serving for the passage of
vessels when the river is high, and one or more shallow passes
in a shallow part of the river, or raised for economy and
facility of working above the river-bed, for regulating the
discharge and affording an outlet for small floods.
A variety of forms of movable weirs have been erected
or experimented upon ; and considerable modifications and
improvements have been introduced from time to time ; but
movable weirs may be classified under three distinct types,
namely, ( i) Frame weirs closed by needles, panels, or a rolling-
up curtain ; (2) Shutter weirs ; and (3) Drum weirs. All these
types, with some of their varieties, are illustrated in section
on Plate 4, drawn to the same scale for the purpose of com¬
parison, the first type being represented by Figs. 7, 8, 9, and
11, the second type by Figs. 10, 12, 13, 14, and 15, and the
third type by Fig. 16.
1. Frame Weirs.
Frame weirs consist of a series of iron frames placed in
a row across a river, carrying a foot-bridge on the top, and
hinged to the apron at the bottom, or suspended from an
overhead bridge, against which the actual wooden barrier
of needles, panels, or curtains rests. The needles, or square
spars, can be removed successively by hand from the foot¬
bridge, the panels can be raised, or the curtains rolled up ;
and then the frames can be gradually lowered on to the bed
of the river, or raised, in the most recent examples, some
height out of the river, so as to leave the channel entirely
open during floods for the discharge of the water and the
passage of vessels.
vi.] Form and Method of Working Needle Weirs. 123
Needle Weir. This form of frame weir derives its name
(barrage à aiguilles) from the long square spars, rounded off
at the top to form a handle, which, standing almost vertically
in the water, side by side in a long row across the river, close
the weir (Plate 4, Fig. 7). A row of light, wrought-iron
frames, placed end on to the current at regular intervals
apart, are connected and fastened together, when standing
upright, by movable, horizontal, iron bars, and can be
lowered flat down across the stream, being hinged to the
apron of the weir. Each frame forms a trapezium braced
horizontally and diagonally ; the frame and braces being
formed of bar iron, T irons, channel irons, or angle-irons.
The lowest horizontal piece of the frame has iron pins at its
extremities, which turn in cast-iron sockets fixed to the
apron, and form the hinges of the frame. The horizontal
pieces at the top of the frames support a foot-bridge of
planks, or iron plates, on which the weir-keeper stands for
removing or replacing the needles. The needles, when in
place, rest against a sill on the apron at the bottom, and
against the front horizontal bar connecting the frames at the
top. The needles are almost always made square in section,
as this form has proved more convenient in practice than
oblong and hexagonal sections. The frames are generally
placed at intervals of only between 3 and 4 feet, so that
they may not be too heavy to be raised or lowered by the
weir-keeper.
To open the weir, the needles are first raised one by one
and removed, the movable bars connecting the frames are
unfastened and withdrawn, and the frames are then gently
lowered one after the other by the help of chains, the con¬
necting bars being withdrawn and the foot-bridge taken up
length by length as the frames are lowered, commencing
from one end of the weir.
The weir is reinstated by raising each frame successively by
means of the chains attached to them, the bars are fastened on
124 Examples and Details of Needle Weirs, [chap.
each pair of frames directly they are raised, the foot-bridge is
then relaid, and lastly the needles are put in place one by one.
Needle weirs were originally exclusively adopted for the
weirs of the Lower Seine ; and some of the largest examples
of this form of weir still exist on that part of the river, though
several of the weirs have been reconstructed, and their form
modified, to raise the water-levels of the respective reaches.
Needle weirs have also been put across the regulating passes
of the Upper Seine weirs, where the water-level of the reaches
above them was raised by the latest improvement works.
Needle weirs have, moreover, been erected on the Meuse and
several other rivers ; and this system has been used for the
principal portion of the weirs recently constructed for the
canalization of the river Main. The sizes of the frames and
needles depend upon the height of the water-level in the upper
reach above the sill of the weir, which must be nearly level
with the bed of the river in navigable passes, and also upon
the head of water which may have to be retained at the weir.
At Martot, the last weir on the Lower Seine (Plate 4, Figs. 2
and 7), the frames are 8 feet long at the base and 11 feet high,
and weigh 467 lbs. each ; and the needles are 13 feet long
and 3 inches square, the head of water retained occasion¬
ally reaching nearly 10 feet. The frames on the Belgic
Meuse are 8£ feet long at the base, and 13 feet high, and
weigh 798 lbs. ; and the needles are 4 inches broad, and
4 inches thick at the bottom, increasing to qf inches at the
centre, and weigh 55 lbs., reaching about the limit which
a man can handle, so that contrivances have been devised
for facilitating their manipulation. Thus on the Meuse and
the Main, the bars supporting the needles along the top of
the frames are hinged at one extremity to a frame, and the
other end is held in position by a vertical spindle on the
adjacent frame, so formed that by giving it a quarter of
a revolution the end of the bar is released, and turning on
its hinged end sets free the series of needles it supported,
vi.] Contrivances adopted at Needle Weirs. 125
which float down stream and are fished out by means of
a long rope passing through iron eyes fastened to each needle1.
In the needle weirs on the Upper Seine, each needle has
an iron hook attached near its handle, which encircles the
bar against which it rests, so that the needle can rotate on
the bar2. The needle is put in place by pushing it up stream
across the foot-bridge till the hook catches the bar ; the needle
then projecting out from the bridge, descends by its own weight,
and, being pushed well down, the stream carries it into position,
its end being arrested by the projecting sill. The needle is
released by lifting it at its hook by a lever or a crab, till its
end rises clear of the sill, when the current makes it rotate
on the bar till its end floats in the current pointing down
stream. By thus raising the needles successively, the weir
is opened to any desired extent ; and the needles can be un¬
hooked and removed by aid of a boat on the up-stream side,
when the current through the weir has slackened.
Needle weirs are simple, comparatively cheap, and easily
worked and repaired. Moreover, by removing a few needles
the discharge can be regulated to any extent ; or sometimes
the simple expedient of pushing out some of the needles
at the top away from the frames, and keeping them in that
position with blocks of wood, affords a sufficient outlet for
the discharge, without the trouble of removing and replacing
any of the needles. There is always a certain amount of
leakage beween the needles ; but it rarely exceeds the discharge
of the river that must be allowed to pass, if the needles
are properly placed and in good condition. Means, however,
have been adopted for reducing undue leakage through these
weirs in dry weather, such as closing the interstices with hay,
putting a row of covering needles in front, or stretching a tarred
canvas curtain along the upper face of the needles.
1 ' Canal, River, and other Works in France, Belgium, and Germany.' L. F. Ver-
non-Harcourt. Minutes of Proceedings Institution C. E. 1889, vol. xcvi, p. 186.
3 Annales des Ponts et Chaussées, 1S81 (2), p. 220, and 18S3 (1), p. 622.
126 Needle Weirs unsuited for great Heights, [chap.
The limit of height to which these weirs can be carried has
probably been reached at the needle weirs on the Lower
Seine. the Belgic Meuse, and the Main. As the needles rest
merely against the sill and the top connecting bar of the
frames, their section has to be considerably increased when
the height of the frames is raised. An attempt was made to
reduce the strain on the needles in high weirs by introducing
an intermediate supporting bar ; but this arrangement, whilst
considerably relieving the needles, prevented the adoption of
contrivances for releasing them, and also materially increased
the difficulty of raising them, owing to their being pressed by
the stream against the central bar. The rotating bar adopted
on the Meuse and the Main, only releases the needles, and
affords no assistance in replacing them. The hook on the
back of the needles used in France is of considerable assistance
in replacing the needles, as soon as the hook is in position ; but
though the system has been tried with heavy needles, the
process of hooking on must always be somewhat difficult ; and
hitherto, in France, the size of the needles has been limited to
3 inches square, which is too slight for high weirs.
The increase in the navigable depth of the Lower Seine,
from 6^ feet to io| feet, has necessitated the raising of several
of the weirs, as well as the construction of additional weirs,
retaining a greater head of water than the original needle
weirs. As, however, the suitable limit of size for the needles
was considered to have been already reached on the Lower
Seine, it became necessary to devise some other form of barrier
for closing the higher weirs, which could be more easily
handled than needles.
Sliding Panels on Frame Weir. As the movable frames
of the needle weirs are only between 3 and 4 feet apart, the
distance between the points of support of a transverse barrier
resting against the frames is much less than the distance
between the sill of the weir and the upper bar against which
the needles rest, which, moreover, increases with the height of
vi.] Panel Weirs on River Seine.
127
the weir. Accordingly, the thickness of a transverse barrier
resting against two adjacent frames may be considerably less
than the thickness of the needles ; and, moreover, its thickness
may be reduced with the diminution in the water pressure from
the bottom towards the surface. Consequently, by adopting
a series of wooden panels sliding between, and resting against
each pair of frames, for closing the space between them, instead
of needles, the thickness of the barrier can be considerably
reduced without diminishing its strength, and proportioned to
the variation in pressure.
The weight of each panel is regulated by its height, which
is made dependent on the limit of weight that can be easily
lifted by a crab on the foot-bridge, winding up a chain
hooked on to an iron loop attached to each panel. This
system was experimented on by Mr. Boulé at Port-à-1'Anglais
weir 1 on the Upper Seine in 1874, and introduced by him at
Suresnes weir on the Lower Seine on its reconstruction in
1880-85 (Plate 4, Fig. 8)2. The navigable pass of Suresnes
weir has frames about 20 feet high ; and the opening between
each pair of frames is closed by four panels sliding in grooves
on the up-stream faces of the frames, each panel having
a height of 4I feet and a width of about 4 feet. The panels
are pushed down into their places by a pole, and are raised
by the crab travelling along the foot-bridge on the top of the
frames, which lifts each panel in succession by means of
a hook and chain. The number and length of the joints in
the panel weir are much less than with needles, so that the
leakage through the weir is considerably reduced ; and the
foot-bridge can be raised high enough to be secure from
submergence, without increasing the height of the panels.
The discharge of the weir can be regulated by removing some
of the upper panels ; and this system has the advantage,
1 Annales des Ponts et Chaussées, 1S76 (1), p. 320.
3 ' Les Écluses de Suresnes et de Bougival," A. Boulé. Mémoires de la Société
des Ingénieurs Civils de Paris, 1884 (2), p. 246.
128 Panels, and Curtains for Frame Weirs, [chap.
over draw-doors and needles, of not producing a scour along
the apron on partially opening the weir, since the discharge
takes place over the top of the panels still remaining in place.
The system of superposed panels for closing a frame weir
is simple, economical, and easily worked ; and the height
which such a weir could attain is only restricted by the limit
of weight imposed on the frames by the necessity of their
being able to be raised without difficulty from the bed of
the river, when the weir has to be reinstated on the approach
of the summer season.
Before this form of frame weir was employed at the recon¬
structed Suresnes weir, completed in 1885, it had been
adopted with success for the regulating portion of the Mulatière
weir across the Saône at its confluence with the Rhone below
Lyons : and it was also employed in 1876 for six weirs on the
river Moskowa between Moscow and Kolumna, but in this
case the height of the panels was made only x foot, so that
they could be easily raised by a boat-hook.
Hinged Curtain on Frame Weir. Another form of trans¬
verse barrier, supported by the movable frames, was designed
by Mr. Caméré for closing the openings between the frames
of Port Villez weir, the first of the new weirs erected on the
Lower Seine for increasing the navigable depth to ioj feet,
and which was opened in 1880. It resembles in principle the
panel systenx, in being composed of transverse pieces of wood
resting on the frames, and proportioned in thickness to the
depth of water ; but it forms a continuous curtain hanging
down the up-stream side of a pair of frames, consisting of a series
of strips of wood joined by hinges, which covers the opening
from top to bottom (Plate 4, Fig. 9). The curtain is rolled
up, from the bottom, by an endless chain turning round a crab
travelling along the foot-bridge on the top of the frames.
The rolling up or unrolling of the curtain can be stopped at
any point, so that the extent to which the weir is opened or
closed can be regulated with great nicety. As, however, this
vi.] Port Villez Weir on Lower Seine. 129
form of weir is opened from the bottom, its apron is exposed
to a powerful scour when the raising of the curtain is com¬
menced. The weir at Port Villez, 700 feet long, is placed along¬
side the lock which is close to the left bank, the two navigable
passes placed nearest the lock having frames across them,
18 feet high and 3 feet 7i inches apart, each weighing 37 cwts.,
and capable of being raised or lowered in ten minutes1. Frames
of half the height are placed across the two shallow regulating
passes adjoining the right bank of the river ; and the foot-bridge
is only 11 feet above the upper water-level. The curtains
which close the whole of this weir have worked satisfactorily,
and have proved watertight ; and similar curtains have been
adopted at the other more recently constructed weirs on the
Lower Seine (Plate 4, Fig. 11), with the exception of Suresnes
weir, where however, in one of the passes, panels and curtains
have been placed side by side to test their relative durability.
The hinged curtain possesses the advantages of being readily
raised or lowered to any extent, and of being capable of
exactly adjusting the discharge through the weir ; whereas
the panels are not so readily raised when the water is flowing
over them. On the other hand, the panels are cheaper and
less complicated, and will probably prove more durable ; and
with them, the apron is not exposed to scour.
Suspended. Frames with Curtain. Though the height of
the barrier of a frame weir could be considerably extended by
the substitution of panels or rolling-up curtains for needles,
the size of the frames which can be conveniently raised and
lowered has been practically reached at Port Villez and
Suresnes weirs. By suspending the frames from above, how¬
ever, instead of hinging them on the apron at the bottom of
the river, it is possible to raise all the movable parts of the
weir out of water in flood-time, and to render the frame
weir applicable to greater heights (Plate 4, Fig. u). This
1 Minutes of Proceedings Institution C. E., vol. lx, plate 4 ; and vol. lxxxiv,
plate 3, fig. 7.
K
130 Suspended Frames at Poses Weir. [chap.
principle was proposed many years ago by Mr. Tavernier for
weirs on the Rhone, where ordinary frame weirs, with the
frames lowered in flood-time on to the bed of the river, would
be impracticable, owing to the large quantities of gravel and
shingle carried down by the rapid current, which would bury
the recumbent frames. The system which has been carried
out at Poses and Port-Mort weirs on the Lower Seine1,
consists of a series of frames suspended vertically from a wide
overhead foot-bridge, and hinged so that they are capable
of being raised to a horizontal position under the bridge,
sufficiently high at the navigable passes to afford the requisite
headway for vessels underneath, so long as the river does not
rise above the highest navigable level. The frames, when
lowered, rest at the bottom against a sill ; and the spaces
between them are closed by hinged curtains, rolled up or let
down by the weir-keeper from a small foot-bridge formed
of brackets hinged to the back of each frame.
Poses weir is separated from the locks by an island,
and is divided by masonry piers into seven passes (Plate 4,
Fig. 4). The two passes adjoining the left bank are navigable
passes, each 1065 feet wide, with their sills 165 feet below
the upper water-level ; and they leave a headway of 17I feet
above the highest navigable flood-level when the frames are
raised, which is I3 feet higher than the ordinary upper water-
level, but 3^ feet below the maximum height reached by the
great flood of March 1876 (Plate 4, Fig. 11). The five other
passes are each 99 feet wide; and the girders across them,
carrying the overhead foot-bridge, are placed at a lower level
than across the navigable passes, as no headway is needed,
so that the platform on the top of them is at the same level
as the platform placed between the girders across the navigable
passes, making the footway level throughout. Each frame
consists of four long ribs braced together, across which
1 ' The River Seine,' I- F. Vernon-Harcourt. Minutes of Proceedings Institution
C. E., 1S86, vol. lxxxiv, pp. 234 and 236 ; and plate 3, figs. 1, 3, and 4.
vi.] Working and Cost of Poses Weir. 131
a curtain, 7 J feet wide, extends, an interval of ij inches being
left between the curtains of adjoining frames for clearance.
The rolled-up curtain rests on the frame when the frame is
raised, which is effected by means of a winch and chain from
the foot-bridge ; and the curtain is only removed for repairs.
The rolling-up or unrolling of a curtain takes fifteen minutes ;
and a frame is raised in twenty minutes, or lowered in ten
minutes ; so that a pass can be fully opened in five hours.
This system of frame weir is very readily opened or closed,
and secures the frames from the injuries to which they are
exposed when laid lapping over one another on the river-bed,
and left under water during the winter, and also from the
strains to which they are subjected in the process of lowering
them into the river, or raising them from the river-bed.
Moreover, the suspended frames have only to be made strong
enough to bear the strain of being raised, and the water
pressure against the curtain when the weir is closed. On
the other hand, the system involves the costly addition of
a wide foot-bridge spanning the openings, and also long
frames and high piers at the navigable passes on account of
the headway required. Nevertheless, the cost of Poses weir
has been stated by Mr. Caméré, the engineer who designed
it, to have been .£15* 5s- Per lineal foot; whereas the cost
of Port Villez weir amounted to £163 Js. per lineal foot1,
though more than half the cost in both these weirs was due
to the very deep foundations necessitated to prevent filtration
under the weirs through the alluvial subsoil. Probably, how¬
ever, the cost of Port Villez weir was considerably increased
by experimental systems tried there, for on comparing the
section of Poses weir with the sections of Suresnes and Port
Villez weirs, drawn to the same scale, (Plate 4, Figs. 8, 9, and
11), it is impossible to avoid the conclusion that, under
At Suresnes, the weir across the neighbouring pass, with frames 19! feet high,
cost £148 per lineal foot ; and the regulating weir, with frames 13! feet high, cost
if 93 per lineal foot.
K 2
132 Frame Weirs, and Shutter Weirs. [CHAP.
similar conditions, for heights not exceeding those of these
weirs, the cost of suspended frames, with their wide overhead
foot-bridge, must be more costly than simple frames. The
Martot needle weir (Plate 4, Fig. 7) cost only £49 per lineal
foot, though it is one of the largest examples of ordinary
frame weirs. The chief merits of the suspended frame weir
are ease and security in working, exemption from injurious
influences, facility of maintenance, and its applicability to
greater heights than those reached at Suresnes and Port Villez.
2. Shutter Weirs.
Shutter weirs are composed of a series of hinged panels,
gates, or shutters, which, when raised, form a barrier across
the river ; and the weir is opened by making the shutters
revolve on a horizontal axis, and fall flat on the bed of the
river. Various methods have been devised for raising and
supporting the shutters, which are generally made of wood
and occasionally of iron ; and in some types the shutter
revolves on an axis at the bottom (Plate 4, Figs. 12, 13, and
15), though in the most common form of shutter weir, the
axis is nearly central (Plate 4, Fig. 10).
Bear-Trap Weir. A shutter weir, called a bear trap, was
erected, in 1818, on the Lehigh river in Pennsylvania. It
was formed by two wooden gates revolving on horizontal
axes on the apron, the up-stream gate pointing down stream,
and the down-stream gate pointing up stream. When raised,
the up-stream gate was supported on the top edge of the
down-stream gate. This weir is not now in existence; but
the type has been reproduced, with improvements, in France,
at the weir of La Neuville-au-Pont on the river Marne (Plate 4,
Fig. 15). The pass in which the gates are placed, is 29 J feet
wide. Culverts have been constructed in the side walls, with
outlets into the upper and lower pools ; and a side passage
from each of the culverts opens into the space under the
gates. For closing the weir, a small shutter, revolving on
vi.] Bear-Trap Weir at La Neuville. 133
a horizontal axis on the apron above the upper gate, is
raised, rising with the stream. This shutter stops the flow
of water through the weir, and removes the pressure from
the recumbent gates. Then by opening the upper sluice¬
gates, the lower ones being closed, the water flows from
the upper pool into the space under the gates and raises
them. The weir is opened in three minutes, by releasing the
fastenings of the gates, closing the upper sluice-gates, and
opening the lower ones, so that the water under the gates
flows away into the lower pool, and the gates fall down on
losing their support. At this weir, the water in the lower
pool is liable to fall too low for navigation before the requisite
head of two feet is obtained in the upper pool for closing the
weir, and consequently the gates are raised by chains.
This type of weir requires a greater pressure of water for
working it than is available at La Neuville; and it is costly
with its large gates, long side walls, and culverts, and is not
suitable for wide passes, so that other forms of shutter weirs
have been since preferred. A drift pass, however, has been
recently constructed on this principle in the Davis Island
weir on the Ohio, 52 feet in width, to afford a convenient
outlet for the large quantities of drift brought down by the
river, as the bear trap is more readily raised again after
the discharge of the drift than one of the shutters of the weir
during the low stage of the river.
Thénard's Shutter Weir. The earliest type of shutter weir
was erected across a shallow pass on the river Orb in the
eighteenth century. It consisted of a shutter, turning on
a horizontal axis at the bottom, supported by a prop when
raised against the stream, and falling flat on the apron
when the prop was drawn aside ; but it was difficult to raise
the shutter against a head of water. Accordingly, when
Mr. Thénard erected similar shutters across three shallow
passes on the river Isle, between 1832 and 1837, he placed
a second set of shutters on the up-stream side of the weir,
134 Thênard's, and Indian Shutter Weirs, [chap.
which, rising with the stream, were retained in an upright
position by chains, and stopping the flow of water through
the weir, enabled the down-stream shutters to be easily raised
and propped up1. Valves were then opened in the up-stream
shutters ; and the level of the water being equalized on both
sides of them, they could be lowered and secured to the
apron, the other shutters remaining upright and forming the
actual weir. The weir was readily opened by drawing aside
the prop of the down-stream shutter. The St. Antoine weir,
erected in 1843, having seven shutters 5 i feet high and 4 feet
wide, was an extension of the same system. A narrow foot¬
way was constructed along the top of the up-stream shutters,
from which the weir-keeper could lift the others.
This form of shutter weir has not been further employed
in France, for it was found that the river had to drop to
a lower level than suitable for navigation before the weir
could be raised. Moreover, the up-stream shutters, in rising
with the stream, were liable to come up too rapidly and tear
away the fastenings of their chains.
Indian Shutter Weir. A similar type of shutter weir has
been employed on a large scale for closing the openings in
some Indian irrigation dams. In the Mahanuddee dam, there
are ten openings, 50 feet in width, each closed by seven
pairs of shutters, the down-stream shutters forming the actual
weir being 9 feet high2. The sixty-six openings in the Sone
dam, 201 feet wide, are each closed by a single pair of
shutters, the down-stream shutter being 9! feet high (Plate
4, Fig. 13). As the up-stream shutters of the French weirs
were liable to injury, though only raised when the discharge
of the river was becoming small, the larger shutters of the
Indian weirs, at the head of irrigation canals, are naturally
exposed to considerably greater strains, as they have to be
1 Annales des Ponts et Chaussées, 1841 (2), p. 45.
2 ' Movable Dams in Indian Weirs,' R. B. Buckley. Minutes of Proceedings
Institution C. E., 1880, vol. lx, p. 44.
vi.] Soné, and Brûlée Island Weirs. 135
raised when a sudden fresh in the river threatens to discharge
too rapid a current into the canal, leading to a frequent
breakage of the chains or their fastenings. To regulate the
rising of the large shutters in the Sone weir, Mr. Fouracres
fitted a hydraulic brake to their up-stream side, which
consists of a piston drawn along a cylinder full of water,
with small orifices in the side. As the shutter rises with the
current, the motion of the piston is checked by the cushion
of water in the cylinder, resulting from the inability of the
water to escape rapidly through the small orifices. Moreover,
as the piston travels along the cylinder, the number of orifices
available for the escape of the water is ^reduced ; so that
whilst the tension on the piston rod is increased as the
shutter rises, the checking force is also augmented.
Hydraulic Shutter Weir. In a shutter weir erected across
a shallow pass at Brûlée Island, on the Yonne, in 1870-72,
Mr. Girard dispensed with protecting shutters, and raised the
shutters against the stream by hydraulic power exerted by
a press under each shutter (Plate 4, Fig. 12)1. A piston,
working in the cylinder of the press, is fastened at its upper
end to a cross-head carrying three connecting rods by which
it transmits its motion to the shutter, the rods being fastened
to an axis placed along the central line of the underside of
the shutter. The water pressure is supplied to the presses
from an accumulator on the bank, the water being pumped
into the accumulator by a turbine which is turned by the fall
of water at the weir. The seven shutters of this weir, each
11 è feet long and 6 feet high, can be raised in five minutes.
The working of this weir has proved quite satisfactory ; and
the large cost of this type of weir, amounting to £36 6j. per
lineal foot, about three times the cost of a needle weir, or
ordinary shutter weir of similar moderate height, is the only
reason which has prevented the extension of the system.
Chanoine Shutter Weir. At a weir erected at Confians on
1 Annales des Ponts et Chaussées 1873 (2), p. 360.
136 Description of Chanoine Shutter Weir. [chap.
the Upper Seine, in 1858, Mr. Chanoine completelytransformed
the arrangement of the shutters, so as to render the system
suitable for navigable passes as well as shallow passes h The
axis on which the shutter turns is placed horizontally, a little
above the centre of pressure, on the down-stream side of the
shutter, and is supported on a wrought-iron tressel hinged to
the apron of the weir (Plate 4, Fig. 10). When the weir is
closed, the tressel stands upright ; and the shutter, having
a slight inclination down-stream at the top, abuts at the
bottom against a projecting sill on the apron, and is supported
in that position by a wrought-iron prop which is connected
with the shutter and tressel, and turns, together with the
tressel, on the same axis as the shutter. The prop, when
supporting the shutter, is inclined at an angle of about 45o ;
and its lower extremity rests in a cast-iron shoe, from which
it can be released by a sideways pull by a bar with projecting
teeth, called the tripping bar, which is laid on the apron and
worked by a man from the bank. When the prop has been
drawn clear of the shoe, it slides down an iron groove on the
apron ; and the tressel and shutter, being unsupported, are
pushed over by the stream, and fall, over the prop, flat down
on a recessed part of the apron.
The shutter is raised, in a horizontal position, by a long
hook from a boat, or from a foot-bridge above the weir, by
a chain attached to its lower extremity. As soon as the
tressel has been raised, and the prop is in the shoe, the lower
extremity of the shutter is released, and it closes, either under
the pressure of the water alone, or, if necessary, by the aid
of a push at the bottom, or by a pull from a chain attached
to the top.
The original intention was that the small shutters across
the shallow passes should automatically regulate the water-
level of the river, their axis of rotation being so adjusted, near
a third of the height up, that a rise of the river above the
1 Annales des Ponts et Chaussées, 1859 (2), p. 197.
vi.] Shutter Weirs on the Upper Seine. 137
desired level would make the shutters rotate and open the
weir, till the drop in the water-level produced by the increased
discharge reversed the balance of pressure on the upper and
lower portions of the shutters, and closed the weir. All the
weirs on the Upper Seine were constructed on this principle ;
but it was found that, though the regulating pass of the weir
opened when the river attained a certain level above it, the
increase in the discharge was too suddenly produced ; and,
moreover, the river fell too low before the shutters righted
themselves and closed the weir h Accordingly when most of
the Upper Seine weirs were raised for increasing the navigable
depth, needle weirs were substituted for shutter weirs across
the shallow regulating passes, it having been found previously
necessary to add a foot-bridge on movable frames to the
regulating shutters, in order to control their movements by
means of chains2.
The axis of the large shutters across the navigable passes
is placed sufficiently above the centre of pressure to prevent
their opening with a rise of the river, even when the water
rises above its ordinary level on the down-stream side ; but
provision is generally made for increasing the discharge along
these passes when the river rises, by inserting a butterfly
valve, resembling a diminutive shutter, in the upper panel of
the shutter, turning on a horizontal axis near the centre of
pressure so as to open when the water-level attains a few
inches above the top of the shutter. These valves also, which
in the Upper Seine weirs are 33 feet high by feet wide,
facilitate the raising of the weir. The shutters across the
navigable passes on the Upper Seine, nf- feet high and 4 feet
i inch wide, are raised from a special boat, according to
the original design ; but a foot-bridge on movable frames has
in many cases been placed alongside the shutters across
1 Annales des Ponts et Chaussées, 1861 2 , p. 209 ; 1868 (1), p. 282, and (2;,
pp. 50, and 366; and 1873 (2% p. 159.
2 Ibid., 1883 (t), p. 622.
138 Shutter Wetrs in France and America, [chap.
navigable passes to facilitate their working, as for instance on
the Marne, the Saône, and the Ohio. Some large shutters
were placed across a deep pass of Port-à-1'Anglais weir on
the Upper Seine, in 1870, which had been formed through
a portion of the shallow pass in order to supply the place of
the lock during repairs 1 ; these shutters can retain a height
of 12 feet of water above the sill of the weir, they are 13 feet
high and 35 feet wide, and they are raised or lowered from
a foot-bridge on movable frames 15! feet high (Plate 4, Fig.
10). The cost of this high weir, with its foot-bridge, was
^67 i2i. per lineal foot ; whereas the average cost of the
weirs, as originally constructed across the navigable passes on
the Upper Seine, io feet high, was £^J 3J. per lineal foot, and
across the shallow regulating passes 6i feet high, together
with a foot-bridge, was £21 ï6s. per lineal foot. The shutter
weirs erected across some navigable passes on the Marne, 9!
feet high, together with the foot-bridge on movable frames,
cost £51 4s. per lineal foot.
The Chanoine system of shutter has been adopted for
weirs for improving the navigation on the Great Kanawha and
Ohio rivers in the United States, three movable weirs and
two fixed weirs having been erected in 1880-87 on the Great
Kanawha river above and below Charleston, for securing
a navigable depth of not less than 6 feet throughout the year,
and one movable weir having been constructed at Davis
Island on the Ohio, five miles below Pittsburgh2. The weirs
on the Great Kanawha river differ from the system described
above, in substituting iron rectangular frames hinged to the
apron for the tressels, in which the shutters revolve ; but the
weir on the Ohio was made on the Chanoine model, with
the addition of a foot-bridge on movable frames across the
navigable pass, 559 feet wide, to facilitate its working, as well as
along the three shallow passes having a total width of 664 feet.
1 Annales des Ponts et Chaussées, 1873 (2), p. 98.
2 ' Reports of the Chief of Engineers, U. S. A.' 1881, 1889, and 1892.
vi.] Weirs on Great Kanaioha and Ohio. 139
The shutters across the navigable pass at Brownstown
on the Great Kanawha river, nine miles above Charleston,
are 13Ï feet high and 3! feet wide, slightly larger therefore
than the large shutters at Port-à-PAnglais weir. Two more
weirs are in course of construction below Charleston, and
several others are in contemplation lower down. It has been
found necessary to dispense with the foot-bridge at the weir
on the Ohio, as the frames were liable to be fractured by
accumulations of drift and floating ice before they could
be lowered ; and the weir is now raised and lowered from
a boat. The cost of this weir, together with a lock, was
about ^"191,000. Though this weir is very much hampered
by the large quantities of drift, and the frequent inability to
lower it before a flood has actually arrived, for fear of stranding
fleets of coal vessels which congregate above it, its utility is
so great for the navigation that the construction of four more
similar weirs has been recommended, to extend the slack-
water navigation 25 miles below Pittsburgh 1. The weirs on
these rivers have to be lowered and raised several times in the
year, to provide for the passing down of sudden floods, and
also in order to afford a passage for the vessels, too numerous
to be accommodated by the lock, as soon as the river below
is sufficiently high.
The tripping bar enables the shutters to be lowered quickly
from the bank or from a pier ; and in order to prevent too
great a strain being thrown upon the bar, its projections for
drawing aside the prop are so spaced that on first moving the
bar. the shutters are tripped singly, on account of the pressure
of water against them ; and then as the pressure becomes less
by the opening of the weir, the shutters are tripped in pairs,
and lastly three, or even four together. In spite, however, of
various precautions, stones are liable to impede the action of
the tripping bar, and some teeth are occasionally broken off,
delaying the lowering of the shutters. To obviate this difficulty,
1 Report of the Chief of Engineers, U. S. A., for 1S89,' part 3, p. 1869.
140 Grooves for Prop at Mulatière Weir. [chap.
Mr. Pasqueau devised a special form of double groove for
the foot of the prop to travel in, which was adopted at the
Mulatière weir at the confluence of the Saône and the Rhone
just below Lyons1. When the shutter is raised, the prop is
drawn along the groove into its shoe in the ordinary way ;
but for lowering the shutter, the prop is drawn a little further
up by a pull on the shutter, till it drops into a second groove,
down which it slides directly the shutter is released, causing
the shutter to descend on to the apron. Thus by standing on
the foot-bridge and pulling upon a chain attached to the lower
end of the shutter, the shutter is raised, and the prop brought
into position ; and it is lowered again by a little further pull
up-stream, without the aid of a tripping bar. This system was
followed at Davis Island weir, where the drift and gravel
brought down by the river seemed to preclude the use of a
tripping bar, which, moreover, could not have been applied
to a pass of the width of 559 feet without intermediate piers ;
and it is also being tried at the weirs on the Great Kanawha
river. The partial destruction of the foot-bridge at Davis Island
weir has led to the adoption of a pusher, worked from a boat
below the weir, to give the necessary movement of the prop
up-stream for effecting the lowering of the weir. At the
Mulatière weir, which was first worked in 188a, the size of the
iron shutters, 14^ feet high and 45 feet wide, together with
the great width of 340 feet of the navigable pass, and its
peculiar position at the junction of two rivers differing greatly
in the periods of their floods, rendered it expedient to dispense
with a tripping bar. The system of two grooves has worked
there with perfect success, enabling the sixty-nine shutters to
be easily raised and gently lowered from the foot-bridge in any
order ; and when any débris accumulates under a shutter, it is
readily scoured away by lowering and raising the shutter, an
operation which would be impracticable with a tripping bar.
1 'Barrage de la Mulatière, Notice sur le nouveau système de Hausses.' A.
Pasqueau.
vi.] Shutter Weir on River Irzvell. 141
The shutter weir possesses the advantage over the frame
weir of being capable of more rapid lowering, especially by
the use of a tripping bar, which has rendered it peculiarly
suitable for the rapid and frequent floods of the Great
Kanawha and Ohio rivers. It is also superior to the needle
weir in allowing a moderate increase in the discharge to pass
over its crest, and also in providing for the passage of small
floods by the ready opening of the butterfly valves, which in
the Mulatière weir, attain the dimensions of 5 feet high by
3 feet wide. The irregularity, however, in the automatic
closing of the regulating shutters across the shallow passes,
leading to the addition of a foot-bridge, and the adoption of
a foot-bridge in many cases for working the shutters of the
navigable passes, render the system more complicated and
costly than the frame weir.
Tilting Shutter Weir. Another form of shutter weir was
erected, in 1882, across the River Irwell at Thostlenest, close
to Manchester, by Mr. Wiswall, consisting of fourteen wooden
shutters, 12 feet long and from 9 to 10 feet wide, turning on
a nearly central axis supported on a fixed open framework
extending across the weir1 (Plate 4, Fig. 14). When the weir
is closed, the shutters are inclined down-stream at an angle of
35° to the vertical, abutting at the base against the flat apron
of the weir ; and they were designed to tip over and assume
a horizontal position when the river rises 2§ feet above their
crest, thus automatically opening the weir. The weir is,
however, usually opened by chains fastened to the centre of
the bottom of each shutter, passing round guiding pulleys and
worked by two crabs, one on each side of the river, each con¬
trolling the seven chains of half the shutters, as the automatic
opening is too sudden in its operation, and is liable to cause
undue erosion of the river banks. The shutters are retained
in their horizontal position by ratchets ; and when released
on the lowering of the water-level, they are closed by the
1 'The Engineer.' Sept. I, 1882.
142 Description of Drum Weir. [chap.
stream. The shutters themselves when raised offer hardly any
impediment to the discharge, as they are protected by the upper
transverse beam of the framework ; but the permanent frame¬
work reduces the sectional area of the waterway through
the weir by rather more than one fifth. This form of shutter
weir, accordingly, though regulating the discharge with pre¬
cision, offers some impediment to the flow when open ; and,
moreover, it would not be suitable for very wide openings on
account of the multiplication of the controlling chains, and it
is not applicable to navigable passes.
3. Drum Weir.
The drum weir was designed by Mr. Desfontaines for the
shallow regulating passes of twelve weirs on the River Marne,
erected between 1857 an<^ 1867, to which river this type of
weir was for many years restrictedl. It consists of an upper,
and an under iron paddle, capable of making a quarter of
a revolution round a central horizontal axis. The upper
paddle forms the weir, which it closes when upright ; and the
under one, slightly larger than the other, revolves in a closed
recess, below the sill of the weir, having the form of a quarter
of a cylinder, which has led to the name drum being given to
this type of weir (Plate 4, Fig. 16). The upper paddle is
straight ; but the under paddle is curved in such a manner,
near its junction with the axis, that a space is left between
its upper face and the top of the drum, when both paddles
are horizontal, through which water can be admitted so as
to press upon its upper face. A similar space is left, by a
prolongation of the drum, between the down-stream face of
the under paddle and the vertical wall of the drum when both
paddles are vertical.
Culverts are provided in the abutment of the weir, com-
1 Annales des Ponts et Chaussées, 1868 (2^, p. 482.
vi.] Drum Weirs on Marne, Main, and Spree. 143
municating with the upper and lower reaches of the river
and the drum ; and the flow through these culverts is con¬
trolled by a see-saw arrangement of sluice-gates, whereby
communication can be opened and closed between the upper
or lower part of the drum and the upper or lower reach,
thus producing or removing a pressure of water on the upper or
lower face of the under paddle, and thereby closing or opening
the weir to any extent with great facility and precision.
The largest weir of this type in France was erected across
the regulating pass of the last weir on the Marne at Joinville,
in 1867, which consists of forty-two upper paddles forming
the actual weir, 3! feet high and 41 feet wide; and this weir
can be completely opened or closed in three or four minutes
by one man.
Bigger drum weirs, precisely similar in type, have been
placed across the timber passes of the four weirs constructed
for canalizing the Main in 1883-86 (Plate 4, Fig. 5). The
pass at each weir, 39^ feet wide, is closed by a single paddle
retaining a head of 5 feet 7 inches above the sill of the weir.
In this case, the upper paddle constituting the weir is slightly
inclined down-stream when closed, so that when it is lowered
into a recess in the apron in opening the weir, the straight
under paddle, which is vertical when the weir is closed, may
leave a sufficient interval between its upper face and the top
of the drum to admit the water from the upper reach when the
weir is to be closed x. These weirs serve to regulate the water-
level in the reach above them, as well as for the passage of
timber ; and they are readily closed against the full discharge
of the river rushing through the pass, when the rest of the
weir is closed.
Another still bigger drum weir has been placed across the
navigable pass of the Charlottenburg weir on the River Spree
(Plate 4, Fig. 16), retaining a depth of 9-^ feet above the sill
1 ' Some Canal, River, and other Works in France, Belgium, and Germany,' L. F.
Vernon-Harcourt. Minutes of Proceedings Institution C.E., 1889, vol. xevi, p. 191.
144 Vnine of Drum Weir, and Disadvantages, [chap.
of the weir, which is 2j feet above the bed of the river, the
normal difference in the water-level of the upper and lower
reaches being 4 feet The width of the single upper paddle
closing this weir is 32^ feet ; whilst the arrangement of the
upper and under paddles and the drum resemble very closely
the drum weirs on the Marne on an enlarged scale.
The great merit of the drum weir is the perfect control
maintained over its movements, and the facility with which
it is raised against a rapid current. As a hydraulic con¬
trivance it is, indeed, superior to any other form of movable
weir ; and its sole drawback is that the drum containing the
lower paddle, extending to a greater depth below the sill than
the actual height of the weir closed by the upper paddle,
involves costly foundations, rendering this type of weir
expensive and preventing its ready adoption for deep passes.
In fact, except at Charlottenburg, it has only been applied to
shallow passes ; and even at Charlottenburg the pass is
moderate in depth. Thus the cost of the drum weir at
Joinville was £27 12s. per lineal foot for an available height
of weir of only 3! feet ; and it reached £144 i6j. per lineal
foot at the Charlottenburg navigable pass. Nevertheless, the
great facility and rapidity of working, and the very small cost
of maintenance, compensate to some extent for the large first
cost ; whilst the preciseness with which the drum weir can
regulate the discharge, and its capability of being raised
against a rapid current, render it a most important adjunct to
a movable weir under certain conditions.
General Remarks on Movable Weirs. Though various
kinds of movable weirs have been experimented on, and
different forms of shutter weirs have been occasionally con¬
structed, there are only three types that can be said to have
entered into ordinary use ; namely, the Frame Weir with
modifications merely of the barrier adopted, the Chanoine
1 Minutes of Proceedings Institution C.E., vol. xcvi, p. 192 ; and Zeitschrift für
Bauwesen, 1886, p. 338.
VI.]
Remarks on Frame Weirs.
145
Shutter Weir, and the Drum Weir ; and of these, the drum
weir has been somewhat limited in its application.
The frame weir with needles, the simplest form of movable
weir, was the first type at all generally employed ; and though
it was introduced more than half a century ago, it is still in
very common use in France and in Belgium, and was selected
in recent years for the weirs on the Main. The only objections
that can be urged against it are, its incapacity for regulating
sudden rises of a river, which may submerge the foot-bridge
before an adequate waterway can be provided by removing
a sufficient number of needles ; and the limit to its height
imposed by the increasing weight of the needles, and of the
frames, necessitated by augmentations in height. The first
difficulty was provided against, at the needle weirs on the
Meuse below Namur, by the addition of a long solid weir
parallel to the stream, over which any increased discharge
could flow without unduly raising the water-level till the
weir-keeper could release some needles ; and it has been
more effectually obviated by the drum weirs on the Main.
The second objection has been met by introducing panels or
curtains in the place of needles, and lastly by suspending the
frames from an overhead bridge, which renders the frame
weir available for any height of weir likely to be required in
a river.
The Chanoine shutter weir, though introduced sixteen years
after the Poirée needle weir, was fortunate in coming into
use at a period when navigation by aid of movable weirs was
being rapidly extended in France ; and it was largely em¬
ployed on the rivers of the Seine basin above Paris. In its
primitive form, by dispensing with a foot-bridge and pro¬
viding an automatic regulating pass, it naturally became a
powerful rival of the needle weir, especially as by means of a
tripping bar its opening could be rapidly accomplished. This
system was specially well adapted for navigation by flashes,
which was still to some extent in use on its introduction.
L
146 Remarks on Chanoine Shutter Weir. [chap.
When, however, continuous navigation by aid of locks became
general, the delay in the automatic closing of the regulating
shutters proved inconvenient ; and the addition of a foot¬
bridge to remedy this defect rendered the system more
complicated and costly than the needle weir, which has,
accordingly, in some cases replaced the shutter weir when
improvements for navigation necessitated the reconstruction
of the weir. The insertion of a butterfly valve in the upper
panel of the shutters has materially facilitated the regulation
of the discharge, and the raising of the shutters ; whilst the
provision of a double groove for the prop to move along has
enabled the tripping bar, with its risks of breakages, to be
dispensed with, and the shutters to be lowered without any
shock, though involving generally the addition of a foot-bridge.
The Chanoine shutter weir, like the needle and drum weirs,
was first employed in the Seine basin ; but it has also been
adopted on the Loire, the Saône, the Meuse above Namur,
and in the United States. After due consideration of the
different types of movable weirs in use, the Chanoine shutter
weir was selected for the Great Kanawha and Ohio rivers,
on account of the rapidity with which it could be lowered,
an operation which is necessary on those rivers several times
in the year, the tripping bar arrangement being used in the
one case, and the double groove in the other. The partial
destruction, indeed, of the foot-bridge of the Davis Island
weir, by the accumulated drift coming down the river, shows
that the frame weir would not be suitable under such
conditions. The shutter weir, accordingly, in its primitive
form without a foot-bridge, possesses distinct advantages over
the frame weir in certain circumstances, and even with a foot¬
bridge in special cases such as at the Mulatière navigable
pass, owing to the facility with which it can be lowered ; but
under ordinary conditions, the frame weir with rolling-up
curtains or sliding panels appears preferable to the shutter
weir.
VI.]
Remarks on the Drum Weir.
147
The drum weir is unquestionably the most perfect type of
movable weir, owing to the perfect ease, regularity, and
rapidity with which it can be raised or lowered, giving the
weir-keeper absolute control of the discharge by merely
moving the handle regulating the sluice-gates, without the
need of a foot-bridge or other accessories. Its sole defect is
its cost, resulting from the deep foundations required by the
drum below the sill of the weir, which for some years limited
its application to the shallow regulating passes of the Marne
weirs. The extension, however, of the system to the timber
passes on the Main, where it performs the double office of regu¬
lating the discharge and providing at any time a passage for
rafts of timber, and still more the drum weir at Charlottenburg
across the navigable pass, show that the system can be
applied to all the purposes of a movable weir. Moreover,
the success which has attended the working of these enlarged
drum weirs, suggests the idea that this system might have
been as applicable to the improvement of the Great Kanawha
and Ohio rivers as the shutter weir, considering the moderate
depth of water to be retained and the obstacles to the
maintenance of a foot-bridge, and that the larger cost
might have been compensated for by the safety and rapidity
of working, the freedom from accident, the exemption from
heavy costs for maintenance, and the facility with which the
floating drift could be passed down.
Each of the three systems has its special advantages ; but
it appears probable that, in Europe, the frame weir, and
perhaps the drum weir, especially for regulating weirs, will be
more extensively used in the future than the shutter weir.
L 2
CHAPTER VII.
PREDICTION OP FLOODS ; AND PROTECTION PROM
INUNDATIONS.
Causes of Floods. Prediction of Floods :—by Rainfall Returns ; various cir¬
cumstances producing Flood Rise ; Observations of Rise of Tributaries ; accuracy
attained. Protection from Inundations :—necessity for, in certain cases ; Methods
adopted; Extension of Vegetation and Forests; Regulation of Torrents, and
Arrest of Detritus; Catch-water Drains; Reservoirs for impounding Floods;
advantages, objections; Removal of Obstructions; Utility of Movable Weirs;
Enlargement of Channel ; Straight Cuts ; Embankments, low banks, high banks,
respective advantages, conditions necessary for high banks, high banks overtopped
by rise of flood-level ; Instances of Embanking Rivers, Mississippi, Po, Loire, and
Theiss ; Rising of River-Bed, in Japan, on Yellow River, causes ;• Pumping ;
Protection of Upper River Valleys. Ports of Refuge. Concluding Remarks.
Floods in rivers are dependent on the amount and con¬
tinuity of the rainfall over the river basin, the period of
the year in which the rainfall occurs, and the nature of the
soil and the extent of its previous saturation, as explained
in the first chapter. The channel of a river, being formed
by the current, is generally only large enough to convey
the average discharge of the river, and is quite inadequate
to carry off the large floods which occur at intervals in the
cold or rainy season. Rivers, accordingly, during these
periods, are liable to overflow their banks, and inundate the
adjacent low-lying lands. Within the tropics, and in coun¬
tries exposed to periodical winds, floods are confined to
the rainy season ; but in the temperate zones, though the
great floods, except those of glacier origin, are limited to
the cold season, an exceptional coincidence of unfavourable
circumstances leads occasionally to the occurrence of a
Difficulties in Predicting Floods. 149
moderate flood in the warm half of the year, especially where
the river basin is mainly composed of impermeable strata.
Prediction of Floods.
The actual fall of rain, in a given period, over any drainage
area furnishes by itself a somewhat incomplete indication
of the probable rise of the river draining the district, owing
to the uncertainty which exists as to the correct ratio between
the actual, and available rainfall in any special basin at any
particular period, in consequence of its variation with the
season of the year, the previous condition of the soil, and
the precise distribution of the rain. In the neighbourhood
of the sources of a river, however, the rainfall is the only
available index of the advent, and extent of a flood ; whilst
the interval between the receipt of the observations of rainfall
and the arrival of the resulting flood, is necessarily small,
especially near the hilly districts which generally border
the limits of a basin, where the valleys are steep and the strata
impermeable. Fortunately in these districts the area of land
subject to inundation is generally restricted, and its agricultural
value comparatively small.
Lower down in the valley, where rich, extensive alluvial
plains, formed by the deposits of centuries, border the river
on either bank, other means are happily available for giving
timely warning of an approaching flood, in addition to rainfall
observations. A flood in the lower part of a river is the
combined effect of the floods of the upper portion of the
main river and the several tributaries, and does not attain
its maximum till a considerable time after the maxima
have been reached in the branches in the upper part of the
basin, the period depending upon the distance to be traversed
and the rate of propagation of the flood. The floods, how¬
ever, of the various tributaries do not generally coincide in
the time of their arrival at any given point on the main river,
owing to variations in the time of the occurrence of the
150 Conditions affecting Rise of Floods. [chap.
flood, partly due to an absence of coincidence in the period
of the rainfall over a large basin, and partly to the difference
in the nature of the strata, and also owing to the differences
in the distances to be traversed, and in the rates of propaga¬
tion of the floods, which depend upon the fall of each tributary
and the condition of its channel. The floods, accordingly,
from some of the tributaries are generally passing off when
the floods of others come down ; and consequently rivers,
in the lower portion of their course, are not usually exposed
to such sudden and proportionately high floods as their
torrential tributaries. Where the distances to be traversed
are not very dissimilar, the floods from the torrential tribu¬
taries arrive first, owing to the rapid flow of the rainfall off
the sloping impermeable strata, and the considerable fall
of such rivers ; whilst the floods of the tributaries draining
permeable strata descend later, owing to their more gentle
rise and slower rate of propagation. Rivers therefore draining
basins composed of partly permeable and partly impermeable
strata, like the Seine, are less exposed to very high floods
than rivers whose basins are mainly impermeable, such as
the Loire, owing to the less liability of coincidence in the
floods of the several tributaries. The actual height, moreover,
attained by a flood depends upon the height of the river
previous to the arrival of the flood, so that the highest floods
generally occur when a second flood comes down before
a previous one has passed off". This is well illustrated by
the flood diagrams of the Seine basin of Nov.-Jan., 1882-83
(Plate X, Figs. 7—t6), where the second flood, though smaller
than the earlier one, attained in most places a greater height.
This circumstance, combined with the influence of evaporation
and the generally low state of the springs during the warm
season, renders high floods extremely rare between May and
October in basins situated like the Seine, containing some
extent of permeable strata (Plate 1, Figs. 3-6); whereas
a winter devoid of floods, such as 1881-82 in the Seine basin,
vu.] Method of Predicting Floods. 151
is quite exceptional. Where, however, the break-up of the
winter is somewhat later, as in the upper part of the Mississippi
valley, floods sometimes extend into June. A deficiency
of rain in the winter, followed by dry weather in the early
spring, affords a prospect of a low river towards the close
of the autumn, whatever the intervening weather may be,
owing to the lowering of the springs by the deficiency in
rainfall, which a subsequent increase in rainfall, the effects
of which are reduced by active evaporation, is unable to
compensate for fully.
By erecting gauges at suitable places on each tributary
of a large river, noting the heights attained by floods, and
telegraphing the results to stations on the main river, and
by observing the periods occupied by the floods on the
various tributaries in reaching the several points on the main
river, and the corresponding rises at these places in a number
of instances, it is possible after a time to predict with con¬
siderable accuracy the time of the arrival of the top of the
flood at a given place on the main river, and the height
likely to be reached by the river some two, three, or more
days before the advent of the flood. This system of pre¬
diction of floods was established in the Seine basin by
Mr. Belgrand in 1854, and has also been adopted in the Loire,
the Garonne, and the Saône basins, and arranged for the Ohio ;
whilst the floods of the Elbe are predicted at Tetschen and
Dresden. It has been brought to such a state of perfection
on the Seine, that during the great flood of March, 1876 (Plate
4» Fig. 2), the maximum height of the flood at Paris was pre¬
dicted, three days beforehand, to within half an inch1. On the
same occasion, the flood, though greater than any since 1807,
did little injury below Paris, owing to the timely warning given
of its approach. Thus, in one place an embankment was
raised above the predicted flood-level ; and the inhabitants
1 ' La Grande Crue de la Seine en Mars, 1876.' E. Belgrand and G. Lemoine.
Annales des Ponts et Chaussées, 1877 (1), p. 435.
152 Vahle of Predictions of Floods. [chap.
of the low districts had time to remove with their cattle and
goods to places of safety. The great value of predictions of
the time of the arrival, and probable height of floods when
inundations are threatened, is manifest ; and where practicable,
this system of warnings should be established on every large
river which is liable in flood-time to inundate considerable
tracts of cultivated land, and still more for towns and villages
which are exposed to danger from floods.
Observations of the rises of the upper portion of a river
and its tributaries are also important, in order that the weir-
keepers lower down may receive timely notice of the approach
of a flood, and thus prepare for its passage by adequately
opening their weirs beforehand, a warning specially needed
in the summer months when the weirs are naturally closed to
maintain the water-level.
Protection from Inundations.
Rivers in their natural condition generally overflow their
banks when floods come down, owing to the size of their
channels being approximately proportionate to their average
discharge ; and, moreover, the material they carry down
tends to deposit in the river-bed on the decrease of a flood,
thus further reducing the capacity of the channel till scoured
away by the next flood. The extent of the inundation
depends upon the amount of low-lying land bordering the
river, and the height attained by the flood ; it is generally
small in the hilly parts of a river basin, owing to the restricted
width of the valleys ; but it becomes large in the flat, fertile,
alluvial plains stretching out at the base of the hills, where
the torrents rushing down from the mountains have their
fall abruptly reduced. Floods are not generally injurious
when spreading quietly over the land in the cold season,
when vegetation is dormant, and where the crops are of
a nature to be benefited by the layer of mud which the flood
deposits on the land. Some kinds of crops, however, cannot
vu.] Means of Protection from Inundations. 153
stand a temporary inundation, and most crops are ruined
by a summer flood ; whilst some lands lie so low that they
cannot be cultivated without protection from floods ; and
villages, and even towns sometimes spread over areas within
the natural limits of inundation. Works therefore for pro¬
tection from floods have been often undertaken by the
proprietors of the lands bordering rivers, or by the State
for the public benefit.
Protection from inundations may be sought, (1) by
measures for directly reducing the volume of flood water
passing down a river ; (2) by facilitating the discharge by
the improvement of the channel ; or (3) by works for excluding
the river from the low-lying lands. The first object may
be attained by encouraging the growth of vegetation, and
planting forests on steep mountain slopes, and thus retarding
the descent of the flood waters, or by the formation of catch-
water drains, or by the construction of impounding reservoirs.
The discharge of a river may be facilitated by the removal
of obstructions and shoals, by increasing the waterway at
weirs and under bridges, and by enlarging the channel.
Lastly, the river may be shut off by embankments raised
above the highest flood-level, and the drainage of the protected
district effected artificially by pumping.
Extension of Vegetation and Eorests on Mountain Slopes.
In several mountainous regions, vegetation has been reduced
by excessive pasturage, and forests have been cut down,
producing a denudation of the steep slopes, and a consequent
increase in the floods along the valley. Exactly the opposite
course should be followed to diminish inundations ; for, as
pointed out in the first chapter, the extension of vegetation,
and the growth of trees equalize the flow of rivers, by retard¬
ing the flow of the rainfall into the watercourses. The general
cultivation of districts has similar effects, for the rain sinks
more readily into cultivated land ; but the drainage of swampy
plains has an opposite influence, for it destroys natural
154 Regulation of Flow : Catch-water Drains. [chap.
reservoirs of rainfall, and leads the rainfall more or less directly
into the nearest stream to swell the volume of the floods.
The influence of increased vegetation is necessarily slow in
manifesting itself, but its operation is sure and permanent ;
and encouragement has been afforded in some districts,
by special legislation, to this means of gradually reducing
inundations.
Regulation of Torrents, and Arrest of Detritus. The
rapid descent, and erosive influence of torrents may be
checked by erecting barriers across their channels. These
small cross dams retain the water to some extent, and there-,
fore reduce the velocity of flow and the erosion of the bed.
Protective works also against landslips diminish the supply of
detritus. In special cases, embankments across the narrow
valley of a torrent provide kinds of reservoirs for the storage of
detritus ; but this system has to be extended from time to
time, by erecting additional embankments to provide fresh
sites for the accumulation of detritus, when the spaces
originally provided have become filled up.
Catch-water Drains. As the discharging capacity of any
channel is proportionate to its fall, it is very important to take
advantage of any available fall in river valleys where the
general fall is extremely small, as, for instance, in the Fen
districts of England. This extensive tract of very low-
lying land, formerly simply marshes, portions of which were
gradually reclaimed from the sea, has been drained by
numerous channels leading the water into the enlarged and
artificially straightened rivers which flow into the Wash,
aided by subsidiary straight drains. Even these straightened
rivers have a fall of barely four inches in a mile, and have
had to be enlarged in section, and supplemented by drains,
in order to convey the drainage waters to the sea. Some
portions, however, of the district are at a higher level ; but
formerly the available fall, due to the higher elevation of
these portions, was lost by letting their rainfall flow directly
vu.] Catch-water Drains : Flood Reservoirs. 155
down into the main river, which, by its more rapid descent,
rendered the drainage of the lower lands more difficult than
if the land had been uniformly flat. By forming catch-water
drains, however, contouring the slopes, which collected the
rainfall of the higher ground before it reached the flat plains
bordering the river, the fall could be utilized ; and a smaller
channel sufficed for conveying this water to a lower point
of the main river which, together with the plains, was re¬
lieved higher up from the influx of this water. Catch-water
drains, accordingly, are very useful in providing for the
drainage of the higher lands of a flat district in an economical
manner, and in relieving the adjacent low-lying plains, and
the river itself, from the direct descent of the rainfall of the
high lands, which aggravated the danger of inundations.
Eeservoirs for impounding Floods. It has frequently been
proposed that reservoirs should be formed in river valleys for
retaining the excess of flood waters, which the river channel
is too small to discharge, and thus prevent their flooding the
riparian lands, till the subsidence of the flood enables the
impounded water to be quietly discharged without inundating
the land. Masonry dams were erected at Pinay and la Roche
in the valley of the Upper Loire, early in the eighteenth cen¬
tury, restricting the river to a narrow rocky channel, and
thus checking the passage of floods, which appear to have
been useful in keeping down the height of large floods in the
valley below them. The very high masonry dam across the
valley of the river Furens, a torrential tributary of the Loire,
at the Gouffre d'Enfer, was designed with the object of
preserving the town of St. Etienne from inundation 1 ; but
previously to its completion, it was determined to utilize
the reservoir thus formed for supplying St. Etienne with
water, for which purpose a second reservoir has also been
constructed higher up the valley. The establishment of
1 Annales des Ponts et Chaussées, 1866 (2), p. 184.
156 Objections to Reservoirs for Floods, [chap.
a regular system of reservoirs along the valleys of the
Loire and of the Yonne has been suggested for the purpose
of mitigating the floods in these rivers. Both the Loire
and the Yonne, however, are torrential rivers flowing over
impermeable strata, and with a rapid fall in their upper
portions. It might be possible to find sites in the valleys
of these rivers, or their tributaries, where short high dams
might retain a large volume of water ; but unless the water
thus stored up could be utilized, it is doubtful whether,
even on this class of river, the results would justify the
expenditure
In most cases, the establishment of reservoirs for the
control of floods is open to serious objections. Many river
valleys, as for instance those of the Thames, the Seine, and
the Mississippi, are not suitable for the formation of storage
reservoirs ; and in such cases, the cost of making adequate
reservoirs would be far greater than the value of any benefits
they could confer. Also, supposing proper reservoirs to have
been formed, they would be unable to cope with a rapid
succession of floods, for when once filled they would be
unavailable for the succeeding floods. Reservoirs, indeed,
are not economically applicable for dealing with floods,
except under specially favourable local conditions, and unless
they can be also employed for supplying irrigation canals,
or for the supply of water to towns, or as a source of water-
power for electric lighting, or other industrial purposes.
Natural reservoirs may occasionally be obtained, where
barren tracts of low-lying land are capable of improvement
by being flooded by a river and receiving a deposit of
alluvium. The tendency, however, of agricultural improve¬
ments lies generally in an opposite direction, namely, the
reclamation of marshy districts formerly serving as natural
reservoirs, which has led to a rise in the level of the floods
in the adjacent valleys, and has increased the damage they
effect. These natural reservoirs form valuable outlets for the
vu.] Flood Reservoirs : Removing Obstructions. 157
flood waters, and should be retained for this purpose, unless
the increased value of the reclaimed land can be charged
with the requisite compensating enlargement of the channel.
The system of reservoirs for the storage of floods, if it
could be extensively applied, would have a similar regulating
influence on the flow of a river as a large lake ; but con¬
siderations of expense preclude the due application of the
system, especially in the basins of gently-flowing rivers,
where the extent of land required for reservoirs would bear
no inconsiderable proportion to the lands to be preserved
from inundation. Reservoirs, however, might be established
with advantage in many mountain valleys, where they would
fulfil the double duty of affording an abundant wholesome
water-supply to the nearest towns, and regulating the flow
of a torrent, thus exercising some influence in mitigating
the floods of the main river.
Removal of Obstructions. Any natural or artificial re¬
striction of the channel of a river, produced by rocky shoals,
trunks of trees, the piers of a bridge, or other obstructions,
checks the flow, and consequently raises the water-level and
intensifies the height of the floods above the obstruction.
Each obstruction produces a loss of fall, and therefore a
reduction of the discharge in the river above ; for the river
has to rise above the obstruction till a sufficient fall is
obtained at the place to enable the discharge to take place
through the diminished waterway. The removal of these
obstructions distributes the fall in the channel, and con¬
sequently increases the discharge, resulting in a lowering of
the water-level and a diminution in the height of the floods.
If, however, the obstructions are only removed from one
portion of a river, the improved discharge in that part will
aggravate the floods in the portion below ; and therefore,
in all such works, a river should be dealt with as a whole,
and should consequently be under a single jurisdiction from
its source to its outlet, or at any rate to its tidal limit. The
158 Importance of Removing Obstructions.
improvement in the discharge of a river that can be obtained
by the removal of rocky shoals, was forcibly shown by the
results of the Severn improvement works in 1842, referred
to on page no, where the height of the floods was reduced
by the lowering of numerous rocky shoals, notwithstanding
the simultaneous erection of a series of solid weirs for the
purposes of navigation.
The removal of obstructions from the channel of a river
is one of the primary duties of river conservancy, for a river
cannot efficiently drain a district if the maintenance of its
channel is neglected ; and care has to be taken to prevent
riparian landowners from restricting the channel for their
private interests, to the detriment of other parts of the river,
as for instance by the erection of water-mills with inadequate
discharge outlets, or by placing fish-traps in the channel, the
wire meshes of which soon become choked by weeds, leaves,
and other débris, or by encroaching on the river with jetties,
quays, or bridges with wide piers and small openings. In
mining districts, the refuse is not unfrequently tipped into,
or close to the streams, to be carried down by the floods,
in reckless disregard of the maintenance of the river below
and its outlet, and of the agricultural and commercial interests
involved. Every river, with its tributaries, should be pro¬
tected against such encroachments and injuries, by powers
granted to the conservators of the river representing the
agricultural and navigation interests of the locality.
Utility of Movable Weirs. The objects aimed at in
improving a river for navigation appear somewhat at variance
with those for mitigating floods, for the water-level of a river
is raised for navigation by means of weirs, and provision is
made against an insufficiency of water ; whereas, for the miti¬
gation of floods, the discharge is facilitated, and the water-level
lowered as far as practicable. Improvements, however, for
navigation generally comprise the lowering of shoals, which
is also advantageous in promoting the discharge of floods ;
vu.] Movable Weirs for Passage of Floods. 159
and the unfavourable influence on floods of raising the water-
level by weirs, can be practically obviated by the adoption
of movable weirs, coupled with warnings of the approach of
a flood, enabling the flood on its arrival to pass down an
almost unrestricted waterway by a timely opening of the
weirs. This compensating influence of movable weirs, renders
them the equitable accompaniment of improvements for
navigation, in order that the prior rights of drainage may be
duly safeguarded. Movable weirs, accordingly, should be
strictly regarded as proper preventive measures against
obstruction of the discharge of a river by improvements for
navigation, rather than as direct works for protection from
inundations. Nevertheless, navigation works have been so
often carried out without due regard to the interests of
efficient drainage, that the enlargement of draw-door weirs,
and the introduction of movable weirs, have often been
subsequently undertaken to mitigate floods. The great value
of movable weirs consists in their removing the antagonism
between the claims of navigation and drainage, and thus
enabling an unrestricted extension of navigation to be carried
out without prejudice to the drainage of a district.
Enlargement of River Channel. An obvious, and very
efficient method of mitigating floods is the enlargement of
the channel of a river, thereby increasing its discharging
capacity and lowering the flood-level. This system is only
applicable to the prevention of inundations in the summer
months, when the floods are generally moderate, and the
protection of the crops of most importance ; for the cost of
coping in this manner with the larger floods of the cold
season would be too great, and inundations in the winter
are generally harmless, and sometimes decidedly beneficial.
The only objections that can be urged against this method
of mitigating floods in the winter, and affording protection
to the summer crops, are the cost of maintenance necessitated
by a channel enlarged beyond the size maintained by the
i6o Enlarging Channel ; Straight Cuts. [chap.
average discharge, and the probable lowering of the summer
low-water level. The system, however, which has been
adopted in the Fens, and has been partially carried out on
other rivers, is sure in its operation, and can be gradually
extended to any desired degree of efficiency.
Discharge increased by Straight Cuts. The adoption of
straight cuts in a winding river reduces the length of the
channel, and consequently increases its fall and facilitates
the discharge. These cuts are most advantageous where the
fall is very slight, as for instance on the Fen rivers, where
the slightest gain of fall is of importance ; in other circum¬
stances they must be adopted with caution, as the variation
they introduce in the flow, and the deviation they produce
in the currents, tend to neutralize their good effect. Cut-offs,
as they are called in America, have been tried on the
Mississippi ; but they are not regarded with much favour,
as, though they reduce considerably the length of the channel
in the large bends of the river, the labour involved is large,
they develop currents unfavourable to navigation, and the
changes they produce in the velocity of the flow lead to
the formation of shoals below. Moreover, the increase in the
discharge is confined to the cut itself, and raises the water-
level below it ; whilst the capacity of the channel is somewhat
reduced by the diminution in its length. Accordingly,
straight cuts are only useful for the mitigation of floods
under somewhat exceptional conditions.
Embankment of Rivers. The method of protecting land
from floods by erecting embankments along each side of
a river possesses several advantages. It provides an enlarged
section above for the discharge of the flood waters, and leaves
the existing water-levels unaltered. Moreover, by placing the
embankments back a certain distance from the banks, the avail¬
able section of the river can be enlarged to any extent, at the
mere cost of the land and of the embankments. Many rivers
have been controlled in this manner, as for instance the Po
vu.] Embankments against Inundations. 161
and the Theiss, which, possessing only a small fall in flowing
across extensive tracts of level plains, have been systematically
embanked. Where low-lying lands border the tidal portion
of a river, or where a river flows through flat marshy districts
such as the Fens, embankments afford the only available
means of preventing inundation. Thus the tidal waters are
kept out by banks from the low-lying lands bordering the
estuary of the Thames ; and the river Witham is embanked
from Lincoln to the sea. The Loire also furnishes an
instance of a torrential river whose valley is protected from
inundations for a considerable length by embankments ; but
owing to the comparatively small width of the valley, the
amount of land protected by the Loire embankments, in
proportion to their length, is only one-fifth of the area
protected by the embankments of the Po.
Two systems of embanking rivers may be adopted ; either
low banks may be formed, rising only to a sufficient height
to exclude ordinary summer floods, and allowing exceptional
summer floods, of rare occurrence, and the high winter floods
to pass over ; or high banks may be constructed, raised above
the highest flood-level, with the object of entirely securing
the land from inundation.
The first system has the advantage of involving only a
moderate outlay, and of securing the land from inundation
at the period when the crops arc especially liable to be
ruined. Moreover, by allowing the winter floods to pass
over the banks, the riparian lands receive each year that
deposit of fertilizing alluvium which is beneficial for many
crops, and which, if confined to the channel, is liable to
reduce its depth on the subsidence of the floods, or to settle
at the outlet of the river. Accordingly, low banks along
a river are, in most cases, wholly beneficial in providing
protection from inundation when most needed, and in not
arresting the beneficial distribution of the alluvial matter
brought down. Moreover, the landowners being accustomed
M
iÓ2 Objections to High Embankments. [CHAV.
to inundations in winter, grow crops suited for such conditions ;
whilst the very exceptional summer floods which may at
intervals overtop the banks, involve less loss than the cost
of higher banks, and the additional expense of maintenance
they would entail. The suitable height of these banks must be
deduced from records of the rises of the river in the warm season ;
and due allowance must be made for the increased rise which
the confinement of the river within banks will occasion.
High embankments constructed along a river to secure the
surrounding country from inundation, like those erected along
the Po, the Loire, the Theiss, and other rivers, have to be
made perfectly watertight, and strong enough to resist the
water pressure at the highest level ; and they are designed to
be raised above the limit of the highest possible flood, for if
the water once overtops the banks, a breach is soon formed
from the washing away of the material by the rush of water
down the high bank. Unfortunately this latter condition of
security is rarely fully attained, for it by no means suffices to
raise the embankments above the highest recorded flood-level.
In the first place, floods in a main river result from a com¬
bination of the floods of its tributaries ; and at rare intervals,
an exceptional widespread and long-continued rainfall, fol¬
lowing upon a wet season or a heavy fall of snow, together
with a coincidence in the arrival of floods from all the
tributaries, produces a flood surpassing all previous records.
Secondly, the embankments, by confining the river and pre¬
venting its flood waters dispersing themselves by flowing
over the plains as formerly, materially augment the height
of the floods to an extent which is generally underestimated
beforehand. Lastly, the detritus brought down by the torrents
from the mountains, which previously was distributed over
the plains by the floods, is kept within the channel by the
embankments, and is liable to settle in the channel as the
velocity of the current slackens, in some cases permanently
raising the river-bed, and consequently raising the flood-level.
vu.] Levees along the Mississippi. 163
From one or other of these causes, or from a combination of
them, these high embankments are sooner or later overtopped ;
a breach is formed at a weak spot ; and the liberated flood
rushes in an impetuous torrent over the adjacent country,
carrying everything before it. and causing widespread injury
to property, occasionally accompanied by serious loss of life.
Instances of Embanking Rivers. Longitudinal embank¬
ments, or levees as they are termed, have been long resorted
to along the Mississippi, for protecting the extensive low-
lying alluvial plains through which the river flows between
its junction with the Missouri and its outlet in the Gulf of
Mexico The most fertile portions of these plains are below
the flood-level of the river ; and levees were commenced in
1717 for securing New Orleans, on its establishment, from
inundation. These levees were gradually extended by settlers
in the district ; and the total length of main levees now
amounts to about 1,300 miles, their prolongation, consolida¬
tion, and maintenance having been taken in charge by the
Government. They are essential for the protection of towns
built within the area naturally liable to inundation, and they
have proved very useful in reclaiming lands liable to be
annually submerged ; whilst they have been beneficial for
navigation during floods, by marking the line of the channel,
and by serving as a quay for local traffic. The system, how¬
ever, does not as yet form a perfectly continuous line of
protection ; nor have the levees hitherto been sufficiently
consolidated throughout to retain the river completely during
floods, so that they are always breached in some weak spots
when the river is in flood. Moreover, exceptional floods,
occurring on the average at intervals of about ten years, such
as those which took place in the spring of 1882 and 1890, are
liable to overtop the banks in places, causing the formation of
wide breaches, or crevasses as they are termed, in the levees,
through which the liberated flood rushes, inundating very
large tracts of land. In 1890 the crevasses in the levees
M 2
164 Breaches in River Embankments. [chap.
attained a total length of 6^ miles, a small length compared
to the total length of the levees, but sufficient to set free a very
large volume of water, flooding in one district 1,350 square
miles of country h Owing to the changes which have taken
place in the channel of the river, the levees are in some places
at a distance from the present course of the river. If it was
desired to secure the land permanently from floods, the levees
would have to be regulated, made continuous, and raised
considerably in places, or placed an ample distance apart, for
in proportion as the gaps are closed, the flood-level rises.
The old levees are being strengthened gradually, and the new
levees are being more substantially constructed ; but it appears
to be considered undesirable to go to the expense of raising
the banks to such an extent as to secure the land from excep¬
tional floods.
The river Po has been embanked for centuries below
Cremona ; and the embankments have been gradually extended
towards its outlet, to protect the rich flat alluvial plains of
Lombardy through which it flows on descending from the
Alps. Its valley, however, has been frequently devastated by
floods, several considerable inundations having occurred during
the nineteenth century. One of the worst of these took place in
October, 1872, when the river forced its way through thirty-
three breaches in the embankments which had been over¬
topped ; and the inundated land was not entirely free from
water till the following April. Conflicting views have been
held as to the raising of the bed of the Po by the deposit of
sediment in its embanked, channel ; but the silting up of its
channel has undoubtedly been slight, and the rise in its flood-
level, amounting to about 7 feet in the last two centuries, must
be mainly due to the increased restriction of the flood waters
by the extension of the embankments along the river and its
tributaries, aided by improved drainage and the cutting down
1 "Report of the Mississippi River Commission for 1890 p 'Report of the Chief
of Engineers, U. S. A. for 1890,' p. 3083.
vu.] Inundations in Italy, and Remedies. 165
of forests. The Commission appointed by the Italian Govern¬
ment, after the flood of 1872, to consider the best means of
preventing the inundations of the Po, considered that it was
unadvisable to endeavour any longer to confine the floods
within the river channel by adding to the height of the
embankments, as such a course tended to aggravate the
injuries which it was designed to prevent. Nevertheless,
the raising of the embankments up to the flood-level of 1872
has since been partially carried out.
Since 1872, large breaches have been formed in the embank¬
ments of the Po in 1879, of the Adige and other Venetian rivers
in 1882, and of the Reno in 1889. Moreover, floods rising
higher than the great flood of 1872 have occurred on all these
rivers and their tributaries since that period. Thus the Po
rose slightly higher at Pisa in 1879 than in 1872, 1 foot
5 inches higher in 1886, and if feet higher at Milan in 1887;
the Adige rose 4^ feet higher at Verona in 1882 than in 1872,
and the Reno 2J feet higher at Ferrara in 1889; whilst some
of the rivers of this part of Italy have exhibited a still greater
increase in height since 18721. To remedy this perilous
condition, the Italian Government are attacking the evil at
its source by retarding the influx of torrential affluents by
constructing small cross dykes, by endeavouring to reduce
gradually the quantity of detritus brought into the rivers by
planting forests on the denuded slopes of the hills, and by
executing works to prevent landslips. The sum of £1,150,000
was also expended on embankment works in 1888-90, since
which period little has been done owing to financial difficulties ;
and out of a total length of embankments on these rivers of
4,012 miles, there are still 223 miles of embankments liable
to be overtopped by the highest floods.
The Loire, below its junction with the Allier, flows through
a plain subject to inundations, which has been protected by
1 I am indebted to Cav. Luigi Luiggi, of the Ministry of Public Works at Rome,
for these particulars about recent floods.
166 Inundations of Loire and Theiss.
embankments. In most cases, a bank on one side only has
been required, as high ground generally borders the opposite
side of the river. Great floods, however, occurring about every
ten years, come down with a rapid flow owing to the fall of
the river, and usually burst the banks and overflow the valley.
The remedy proposed consists in forming openings in the
banks at suitable places, protected by pitching, with their sills
just above the height of ordinary floods, so that the higher
floods may flow comparatively quietly through these openings,
and in this way the devastation caused by the sudden bursting
of the river through breaches may be avoided 3. The land¬
owners have, however, hitherto refused to allow outlets to be
formed in front of their lands.
The river Theiss, on descending to the plains of Hungary,
has a very small fall, amounting to from 6 to 9 inches per mile
above Szegedin, and to only i*a inches between Szegedin and
its junction with the Danube. In consequence, the plains
through which the river winds were frequently flooded ; and
the extent of country covered by these floods was 6,000 square
miles. Of this area, 4,200 square miles have been protected
by the regulation and embankment of 740 miles of river.
The works of improvement, however, having been arrested in
1867, and the necessary works for counteracting the gradual
rise of the flood-level, resulting from the straight cuts formed
above the town, having been neglected, the waters of the
Theiss rose above the protecting embankments in March,
1879, and, forming a breach, destroyed a portion of the town
of Szegedin, and inundated two hundred square miles of
country. The raising of the embankments, and the completion
of straight cuts below the town have since secured Szegedin
from inundation, though the floods of 1881 and 1888 rose
higher than the flood of 1879, nearly reaching the top of the
embankment on the right bank in some parts below the
town.
' ' Rivières et Canaux.' P. Guillemain, vol. i, p. 247.
vu.] River-Bed raised by Deposits. 167
Raising of River-Bed. A point of considerable importance
in considering the expediency of protecting districts from
inundation by longitudinal embankments, and the prospects
of permanent protection afforded by such works, is the possi¬
bility of deposits of alluvium in the river-bed gradually raising
the bed, and consequently producing a rise in the flood-level,
which would necessitate successive additions to the height of
the embankments, or would result eventually in the rupture of
the banks by the flood overtopping them. The Rhine has
been systematically embanked for long distances without
any elevation of its low-water level having been observ¬
able ; and if any raising of the bed of the Po has occurred,
it has unquestionably been slight, though this possibly
may have been due to the periodical scouring out of the
channel by the escape of the river through breaches in
the banks. On the other hand, the river Reno in northern
Italy raises its bed in the plains with the detritus it
brings down from the mountains, which it deposits when its
velocity is checked by the reduction in its fall, and is a source
of anxiety in consequence.
Some rivers in Japan, descending from the mountains, and
embanked across flat alluvial plains, have so raised their beds
by the gradual deposit of their sediment, followed by a raising
of the protecting embankments, that in some cases the surface
of the water is 40 feet or more above the level of the plains over
which the rivers flow. Under these peculiar circumstances, in
constructing railways across these plains, it has occasionally
proved advisable to drive a tunnel on the level underneath the
bed of the river, rather than raise the railway to cross over the
river1. It is evident also from an account by Mr. Morrison,
an English engineer residing at Shanghai, of a personal in¬
spection of the Yellow River of China, in September 1888,
after the great inundation which resulted from a breach in
1 1 Railway Work in Japan.' W.F.Potter. Minutes of Proceedings Institution
C. E., vol. lvi, p. 10, plate I, figs. I and 2.
i68 Deposit in Bed of Embanked Rivers. [CHAP.
the embankments of the river in the autumn of 1887, that
a similar raising of the river-bed occurs on this river h
The embanking of a river tends rather to improve its scouring
capacity by concentrating its flood waters, thereby generally in¬
creasing its hydraulic mean depth and consequently its velocity ;
and therefore the embankments somewhat diminish a river's
tendency to deposit its sediment. The limits, however, of deposit
are restricted by the embankments ; and the alluvial matter
which,under natural conditions,would spread over the inundated
plains, is confined within the flood channel by the embank¬
ments. Accordingly, torrential rivers descending from the
mountains, heavily charged with detritus, when their velocity is
diminished by the reduction in the fall, deposit the heavier
material in their bed which they formerly strewed over the plain,
in spite of a larger proportion of the alluvium being carried
down by the concentrated current. The rising of the beds of the
Japan rivers referred to, and of the Yellow River, is clearly due
to this cause ; and under such conditions, the raising of the
embankments to correspond with the rise of the bed, infallibly
leads sooner or later to a breaking of the banks, and the dis¬
charge of the flood waters of the river into the plains below,
a catastrophe which periodically occurs in Japan and in the
valley of the Yellow River, spreading devastation far and wide.
Pumping for Drainage of Land. In some places the land
is at such a low level in relation to the watercourses of the
district, that it cannot be drained by gravitation ; and the
drainage waters have to be raised over the protecting banks
by pumping, and discharged into the nearest watercourse.
This system is extensively adopted in Holland, and it has
been applied to portions of the Fens in England. Low-lying
land can thus be reclaimed and secured from inundation,
provided its value is sufficient to bear the cost of the pumping,
which is sometimes reduced by the use of windmills to
supplement steam power.
1 'North China Herald.'
vu.] Protection of Valleys : Refuge Ports. 169
Protection of Upper Valleys of Rivers. Embankments and
regulation works should not generally be undertaken in the
upper part of river valleys, for they hasten the descent of floods
and promote the carrying down of detritus, and consequently
intensify the floods in the lower parts of the valleys, and thus
protect the less valuable lands above at the expense of the
alluvial plains below. The Rhone valley, indeed, above the
Lake of Geneva, has been protected from inundation by
regulation works and embankments, without injury to the
river below, owing to the intervention of the moderating
influences of the Lake of Geneva. This, however, is an
exceptional instance, where the peculiar physical conditions
rendered such a course practicable without detriment to
the lower parts of the valley. The Rhone also is a river
along whose valley protection from inundation is specially
important, as, owing to its glacier origin, its floods occur
in the summer season.
Refuge Ports on Rivers. Where the navigation on large
rivers is liable to be endangered during floods by débris and
floating ice, ports of refuge are constructed at suitable places
near towns, where vessels can take shelter during the winter.
A protecting bank or jetty is constructed enclosing a portion
of the river alongside one of the banks, having a suitable depth,
thus forming a sheltered haven with an entrance at its down¬
stream end, in which vessels can lie in safety. Moreover, by
the construction of a quay along the land, the port furnishes
a convenient sort of dock for loading and discharging vessels.
A port of this kind has been established on the Main, a little
below Lrankfort (Plate 4, Lig. 5) ; and many ports of refuge
have been constructed on German, and most American rivers.
Concluding Remarks. In matters relating to river con¬
servancy and works for protection against floods, every main
river basin should be under the control of a single body,
either the State or conservators representing the several
districts, who alone can undertake systematic works for the
170 Public Conservancy of Rivers. [chap.
general benefit of the whole watershed. This controlling
authority should be vested with full powers for preserving
the channel of the river from injury, and to prevent works
being undertaken by private individuals ; for refuse thrown
recklessly into a river is injurious to the maintenance of
the channel lower down, and isolated embankments or cuts,
affording protection or relief at one part of the valley, serve
to intensify the floods elsewhere. The management of these
matters, on the continent of Europe and in the United States,
by competent officials under the direction of the State, affords
evidence that the control of rivers may be fairly regarded as
of national importance, to be carried out for the public
benefit. The comprehensive nature of the works involved,
necessitating their being taken out of the hands of private
landowners, obliges public funds to be employed for their
execution, either drawn from the taxes or raised by a general
rate levied on the district. In any case, however, it would be
equitable for the proprietors of lands liable to inundation to
contribute a considerably larger proportion of the cost of
works for the mitigation of floods, than the owners of the
higher lands, who are merely interested in the general
healthiness of the district and the preservation of the means
of communication. Absolute security from inundation should
be provided for towns in the neighbourhood of rivers ; but
often, in the case of cultivated land, the cost of works
providing entire immunity from floods would be greater than
the loss resulting from partial inundations during exceptional
floods. The removal of obstructions, and the improvement of
the channel are the first works to be undertaken for the
mitigation of floods. Further protection against inundation
can be provided to any extent by continuous embankments
along the main river and its tributaries, aided sometimes by
straight cuts where the river is tortuous and the fall very
slight. Embankments, however, must be adopted with
caution, for they often have to be raised much higher than
vu.] Works for Mitigation of Floods. 171
anticipated at the outset, owing to the rise in the flood-level ;
and they cannot readily be abandoned when the riparian
landowners have become accustomed to their protection,
even when, as in the case of the Loire, periodical disasters
counterbalance their advantages. Embankments are very
advantageous where wide tracts of low-lying land border
a river, such as the alluvial valley of the Mississippi ; but they
are generally inexpedient in the higher parts of a valley,
owing to the increase they produce in the floods below, or in
the somewhat narrow valley of a torrential river which
periodically resumes its sway over the valley. This system
of protection should be avoided, if possible, in the case of
rivers which raise their beds in the plains with the detritus
they bring down from the mountains, for under these circum¬
stances the struggle with nature is continuous ; and she,
sooner or later, avenges herself for the opposition to her laws,
by bursting her bonds and inflicting widespread disaster, the
struggle being subsequently renewed, to be terminated in
a similar manner.
Mitigation of floods near the mouth of a river results from
improvement works at the outfall, though generally the object
of such works is solely the benefit of navigation. Regu¬
lation works, indeed, at the confluence of two rivers, to prevent
the injurious conflict of their currents, are sometimes under¬
taken to promote the discharge, as well as for navigation ;
and the new cut for the outfall of the Witham was carried out
with this double object. Most of the improvement works,
however, at the mouths of rivers, such as the dredging of the
Tyne, merely incidentally reduce floods by facilitating the
discharge through the enlargement of the waterway. Outlet
works would usually be too costly, and their influence would
not extend sufficiently far up a river, to be undertaken for the
mitigation of floods; but in benefiting navigation, they also
exercise a beneficial influence in moderating floods.
CHAPTER VIII.
DELTAS OF TIDELESS RIVERS; AND
IMPROVEMENT OF THEIR OUTLETS.
Contrast between Deltas of Tideless Rivers, and Tidal Estuaries. Deltas :—For¬
mation, Form, Bars; Advance; Rhone Delta; Nile Delta; Danube Delta;
Volga Delta; Mississippi Delta; Comparison of these Deltas; Methods of Im¬
provement. Dredging and Harrowing on Bar : —Objects ; at Outlets of Danube
and Mississippi; at Volga Mouths, results, proposals. Parallel Jetties at Outlet:—
To scour the Bar ; Conditions of success, and permanence ; Rhone Jetties, previous
conditions of Mouths, works carried out, their effects, causes of failure ; Danube
Jetties, Sulina Mouth selected, form of works, results, causes of success, compared
with Rhone works, gradual shoaling in front, improvement of Sulina branch;
Mississippi Jetties, South Pass Outlet selected, works described, results, works at
Head of Pass, shoaling in front of outlet, dredging, prolongation of jetties imminent ;
Mississippi and Danube works compared. Ship-Canal instead of a Delta Channel :—
Object, at Ostia, San Carlos Canal, Mahmoudieh Canal, St. Louis Canal, pro¬
posed for Danube and Mississippi. Conclusions.
All rivers, till they approach the sea into which they
finally discharge the rainfall they have brought down, present
somewhat similar characteristics ; their current always flows
in one direction ; their channels bear a proportion to the size
of their basins ; and the variations in them are merely due to
differences in climate, rainfall, stratification, and the fall of
their bed. The detritus they bring down is gradually carried
along to their outlet, or strewn over the adjacent plains ; and
their channel is more or less uniformly maintained by the
average discharge. Rivers, however, on approaching the sea
exhibit a notable difference in their condition, according as
they flow into a tideless, or a tidal sea. In the first case, the
river continues its downward flow with a diminishing fall and
velocity, till, on encountering the inert waters of the sea, its
current is checked, and finally arrested as it mingles its waters
with the ocean. In the second case, the tide flows and ebbs
Contrast between Tidal and Tideless Outlets. 173
for some distance up a river, the extent of the tidal influence
depending upon the inclination of the river-bed. the rise of the
tide at the outlet, and the volume of fresh water coming
down the river. Accordingly, in a tideless river, the outlet
channel is wholly maintained by the fresh-water discharge, as
in the portions higher up ; whereas in a tidal river, the water
brought down by the river is largely reinforced by a great
volume of sea-water, which enters the estuary during the
flood tide, and flows out, together with the accumulated
volume of fresh water, during the ebb. These differences
produce a very marked contrast in the form and constitution
of the outlets of sediment-bearing tideless, and tidal rivers ;
and they necessitate the adoption of essentially different
methods for the improvement of these two classes of rivers.
A comparison of the tideless mouths of the Rhone, the
Danube, and the Mississippi (Plate 5, Figs. 1, 2, and 5) with
the estuaries of the Thames, the Plumber, the Mersey, the
Clyde, the Seine, the Loire, the Weser, and other tidal rivers
(Plates 7, 8, and 9), fully exemplifies the great contrast which
they present. There are, indeed, as pointed out in Chapter I,
p. 19, some rivers whose outlets occupy a sort of intermediate
position between tideless and tidal rivers, in consequence of
either a great fresh-water flow densely charged with alluvium,
like the Ganges, or a small tidal rise in proportion to the
fresh-water discharge, like the Rhine ; or a tideless river may
be free from a delta, owing to the small proportion of sediment
in relation to the discharge. In dealing, however, with the
improvement of the outlets of rivers, it is expedient to draw
a broad distinction between the two extreme classes of rivers,
and to consider them quite separately, reserving a modification
of the principles of improvement for those river mouths which
hold a somewhat intermediate position.
Deltas.
A very marked peculiarity of rivers bringing down large
174 Origin and Advance of Deltas. [chap.
quantities of alluvium, collected in the upper part of their
valleys and eroded from their banks, is the deltas which they
form at their outlets, when flowing into tideless seas. The
detritus brought down by a river, partly in suspension and
partly rolled along the bottom, at length reaches the sea, and
is gradually deposited at the outlet, when the current of the
river is arrested on emerging out of a confined channel into
the open sea. This deposition is promoted by the much more
rapid settlement of clayey alluvium in salt water than in fresh,
owing probably to the aggregation of the particles produced
by saline solutions. The denser material is first deposited ;
whilst the lighter matter is carried on further by the fresh
water which spreads out like a fan on the top of the salt
water, but is eventually deposited, if not carried away by any
littoral current. Accordingly, the continually accumulating
mound of sediment advances in a fan-shaped form into the
sea, projecting out from the coast-line, and reduces the fall
of the river near its mouth by prolonging the outlet channel.
The enfeebled current finds its way to the sea, through the
mass of sediment which it deposits, in several shallow,
diverging channels, which, with the coast-line, present some
resemblance to the Greek letter A ; and, in consequence, the
name delta has been given to the flat, low-lying alluvial land
formed by these rivers. The discharge through each of these
channels, in depositing its burden of sediment on emerging
into the sea, forms a bar with the denser material in advance
of the outlet; so that these tideless rivers offer serious
obstacles to navigation at their mouths, though often possess¬
ing an excellent navigable channel above their deltas. These
bars progress seawards with the advance of the delta ; and
their distance from the mouth of the channel depends upon
the volume and velocity of discharge through the channel.
The rate of advance of a delta depends upon the amount of
sediment brought down by the river and its density, the depth
of the sea in front, and the influence of disturbing causes, such
VIII.]
Description of Rhone Delta.
175
as winds, waves, and littoral currents. The advance of the
delta at each mouth of a tideless river is proportionate to the
discharge through the mouth, for the sediment conveyed
through each outlet channel of any particular river is pro¬
portionate to the discharge.
Delta of the Rhone. The Rhone divides into two branches
a little above Aries, at which point its delta is considered to
commence by the divergence of the Little Rhone from the
main river. The Great Rhone flows past Aries in a channel
500 feet wide, and having a maximum depth of 46 feet ; and,
though Aries in Roman times owed its importance to its
proximity and ready access to the sea, it is now 32 miles
distant by river from the mouth of the Great Rhone, owing to
the steady advance of the delta. Previously to 1852. the
Great Rhone, which is the navigable branch, and conveys
over four-fifths of the discharge of the river, flowed into the
Mediterranean Sea, at the Gulf of Lyons, through six mouths ;
and the advance of its delta has been estimated at 140 feet
annually (Plate 5> Fig- -0- The Rhone brings down large
quantities of mud, sand, and shingle ; but the shingle finally
disappears at Soujean, 4^ miles above Aries ; and the sedi¬
ment below this point consists entirely of mud and sand, of
which the whole delta is composed. The main branch of the
Rhone brings down about 23! million cubic yards of alluvium
annually, as estimated by Mr. Guérard, or about stVö of the
volume of water discharged ; and this alluvium for the most
part consists of fine sand rolled along the bed of the river, less
than a fourth of the volume of silt brought down having been
found in suspension in the current h
The sea-bottom along the Mediterranean coast in front of
the delta has a fair dip seawards, a depth of 5 fathoms being
reached within 5 furlongs from the shore ; but the slope is
considerably flatter in the inner bay, called the Gulf of Foz,
1 Rapport sur l'amélioration des embouchures du Rhône. A. Guérard, 18S8 ;
and Minutes of Proceedings Institution C. E., vol. Ixxxii, p. 309.
176 Delta of Nile, and Annual Rise of River, [chap.
into which the existing outlet channel discharges. The
depth over the bar in front of the principal mouth, in 1852,
previously to the commencement of any improvement works,
averaged 4 feet 7 inches, increasing to a maximum of 7$ feet
in flood-time, and falling to a minimum of 1 foot 7 inches
during the low stage of the river. Access from the river to
the sea being thus practically barred, the principal port of
France has been established on the sea-coast at Marseilles,
out of the reach of the alluvium of the Rhone, where the
necessary shelter has been provided artificially by large
breakwaters, which rivers commonly afford on tidal coasts.
Delta of the Nile. The Nile furnishes one of the best
known instances of a delta-forming river ; and to its outlets
the Greeks applied the name of delta, which has come into
general use to denote similar excrescences of other tideless
rivers. The river begins to divide at Cairo ; and about
12 miles lower down it splits into two principal branches,
called Rosetta and Damietta, through which, in conjunction
with several minor channels, it finds an outlet for its waters
into the Mediterranean, some ancient branches and mouths
having silted up. The delta of the Nile extends over an area
of about 9,000 square miles, having been gradually formed by
the alluvium of the river, which during high Nile constitutes
about °f the discharge.
The annual rise of the Nile is very regular, being caused
by the tropical spring rains over the mountains of Abyssinia,
from whence the alluvium is derived. The river begins to
rise about the beginning of July ; it reaches its maximum
height in September, averaging about 25 feet at Cairo ; and
it falls again more gradually to its minimum flow in April.
Its average maximum discharge at Cairo is about 11,000
cubic yards per second, and its minimum, 525 cubic yards.
Accordingly, the flood discharge is twenty-one times the
low-water flow, and raises the river 3 or 4 feet at its mouths ;
so that whilst there is a depth of about 7 feet over the bars at
vin.] Description of Delta of Danube. 177
the outlets when the flood is at its maximum, the entrance
is barred, even to small vessels, during low Nile.
Delta of the Danube. The Danube flows past Isaktcha in
a single channel, 1,700 feet wide and 50 feet deep; but 15
miles lower down, its delta commences by the splitting up
of the river into two channels, about 4,7 miles from the Black
Sea in a direct line (Plate 5, Fig. 2). The southernmost
channel divides into two branches 11 miles lower down, so
that eventually the river flows through its delta in three main
branches to the sea. The northern or Kilia branch, carrying
nearly two-thirds of the whole discharge of the river, after
meandering in places through several channels, forms an
independent delta near its outlet ; and, after a course of
61 miles, it flows into the Black Sea through twelve mouths,
encumbered by a continuous shoal of alluvium in front of
their outlets, with depths over it of only from 1 foot to 6 feet,
and advancing seawards 200 to 300 feet annually h The
central Sulina branch, receiving rather less than one-thirteenth
of the discharge of the river, flows into the sea through
a single outlet, 63 miles from the head of the delta, having
a bar in front, over which the depth varied from 7 to 12 feet
between 1829 and 1857, the bar attaining its greatest height
on the subsidence of high floods. The yearly advance of the
delta in front of this mouth averaged 94 feet before the
commencement of the improvement works ; this much smaller
rate of advance, as compared with that of the Kilia delta,
being due to the much smaller discharge of the Sulina channel,
and the greater depth of the sea in front of its outlet. The
southern, St. George's branch carries rather more than two-
sevenths of the discharge, and possesses a good navigable
channel with a minimum depth of 16 feet ; but it splits up
near its outlet into two branches, and discharges into the
1 ' Description of the Delta of the Danube.' C. A. Hartley. Minutes of
Proceedings Institution C. E., 1862, vol. xxi, p. 278.
N
178 Deltas of the Danube and the Volga, [chap.
Black Sea, 75 miles from the head of the delta, through two
mouths, having depths of only 3 feet and 7 feet respectively
over their bars, which are further out from the shore than the
Sulina bar.
The area of the delta of the Danube comprised between the
northern and southern branches is 1,000 square miles. The
total volume of water discharged by the Danube averages
4,630 cubic yards per second at low water, and 12,000 cubic
yards at high water ; but at an extremely low stage, the
discharge falls to nearly half the first discharge, and in an
extraordinary flood it reaches over three times the latter
discharge, whilst the mean discharge is 8,240 cubic yards per
second. The alluvium carried in suspension by the river is
naturally very small during the low stage of the river, and
very large proportionately during floods. The suspended
matter amounts, on the average, to °f the volume of
water discharged ; but a considerable amount of denser
material must be rolled along the bed during floods, which
materially increases the proportion of alluvium actually
brought down.
Delta of the "Volga. The delta of the Volga commences
about 31 miles above Astrakhan, where the Buzan branch
diverges from the main channel ; and its area amounts to
about 5,300 square miles. Near Astrakhan and below, other
channels branch off, which again subdivide, so that eventually
the Volga enters the Caspian Sea through at least two hundred
mouths1. The Volga has an uniisually moderate fall, even
in the upper portion of its course ; and the large quantity of
alluvium it brings down appears to be in great measure due
to the erosion and falling-in of high cliffs which border its
channel in some parts. Uncertainty exists as to the pro¬
portion of alluvium brought down ; but, according to some
1 ' Les Embouchures du Voiga.' V. E. de Timonoff. Vme Congrès Interna¬
tional de Navigation intérieure, Paris, 1892.
vin.] Navigable Channels of Volga Delta. 179
observations, the river has carried along 26,000 cubic yards
per day during a flood, which bears a somewhat small
proportion to the maximum discharge of the river in flood-
time, averaging about 27,000 cubic yards per second. At the
low-water stage, the discharge falls to 2,600 cubic yards per
second. The rate of advance of the delta of the Volga, which
necessarily varies at the different mouths, is not accurately
known; but it has been estimated at 1,270 feet per annum,
a very rapid advance, which, if not considerably exaggerated,
can only be accounted for by the shallowness of the Caspian
Sea in front of the delta. The northern portion of the Caspian
Sea is, indeed, very shallow, and there has been an average
annual decrease in its depth of 3 to 9 inches in the last fifty
years ; and the level of the sea is affected to the extent of
some feet by the wind. Out of the numerous changeable
branches of the Volga traversing the delta, two only at the
present time serve for the navigation between Astrakhan and
the sea, namely, the Bakhtemir and Kamysiak channels. The
former channel is the most frequented, as a normal depth
of 8 feet over its bar has been obtained by dredging, whereas
the depth over the Kamysiak bar is only feet ; and these
depths are liable to be raised or lowered two or three feet by
the wind. In other respects, however, the Kamysiak is the
better channel, for it is much straighter and shorter, and
generally deeper than the other ; and sea depths of 3 fathoms
approach much nearer the Kamysiak mouth than the other
outlets. The advance also of the delta must be much more
rapid in front of the Bakhtemir outlet, as rather over one-
third of the whole discharge of the Volga is conveyed by
this channel.
Delta of the Mississippi. The average rise of tide in the
Gulf of Mexico opposite the mouths of the Mississippi is only
14 inches, and there is only one tide in a day. Accordingly,
the Mississippi entering a practically tideless sea, and bringing
down large volumes of sediment during floods,, forms a delta
N 2
i8o Channels of Mississippi Delta. [chap.
like the European rivers previously described, which flow into
tideless inland seas. The delta of the Mississippi is con¬
sidered to commence just below the confluence of the Red
River, 316 miles from the outlet, where the first small branch
separates from the main river ; and it stretches over an area
of 12,300 square miles3. The regular splitting up, however,
of the main channel does not take place for a long distance
further down ; and at Fort St. Philip, within 35 miles of the
Gulf of Mexico, the river still flows in a fairly uniform channel,
2,470 feet wide and 120 feet deep (Plate 5, Fig. 5). Twenty
miles below Fort St. Philip, the river, having attained a width
of about if miles, and its depth being reduced to 30 feet,
divides into three very diverging channels, two of which
eventually split up again, so that at last it discharges into
the Gulf of Mexico through seven mouths. The delta projects
about 200 miles in front of the natural shore-line of the gulf ;
and its formation has extended over at least 4,400 years. The
mean discharge of the Mississippi into this main delta amounts
to about 23,000 cubic yards per second ; and the average
quantity of alluvium discharged into the Gulf of Mexico in
a year reaches 299! million cubic yards, giving a proportion
of sediment equivalent to about of the discharge. The
sediment consists of silt, sand, and clay ; and the heavier
materials are rolled along the bed of the river, the quantity
thus transported being estimated at one-tenth of the whole.
The annual advance of the delta at the three principal outlets
is, 300 feet at the mouth of the South-West Pass, which
conveys one-third of the whole discharge ; 260 feet at the
mouth of the Pass à l'Outre, which carries one-fourth of the
discharge ; and it was 100 feet at the mouth of the South Pass
before any works were commenced, only about one-tenth of
the discharge passing through this outlet, the advance being
1 1 Report on tlie Physics and Hydraulics of the Mississippi River.' Humphreys
and Abbot, p. 452.
vin.] Bars at Mouths of the Mississippi. 181
naturally somewhat proportionate to the discharge. Though
the average depth in the passes is from 34 to 58 feet, the
bars in front of the outlets reduced the available depth to
a maximum of about 13 feet, which is the depth over the bar
of the South-West Pass about 5 miles seawards of the outlet,
constituting formerly the only navigable channel between the
river and the gulf, after the deterioration of the outlet of the
North-East Pass, one of the branches from the Pass à l'Outre,
during the first half of the nineteenth century. Owing to the
smaller discharge through the South Pass, the depth over its
bar was only 8 feet ; but the bar was only 2} miles in advance
of the coast-line, less than half the distance seaward of the
bar of the South-West Pass. The sea-slope in front of the
delta of the Mississippi is about the same as that of the
Mediterranean Sea in front of the Rhone delta, and about
four times as steep as the slope of the Black Sea in front of
the Danube delta.
Comparison of Deltas. The physical conditions which it
is important to compare in the foregoing examples of deltas,
are the proportionate volume and density of the sediment,
the inclination of the sea-slope in front, the rate of advance
of the delta, the natural depth over the bar, and the extent
and direction of any littoral current across the face of the
delta.
The Rhone appears to carry along about one-eighth more
sediment, in proportion to its discharge, than the Mississippi,
and probably between two and three times more sediment
than the Danube. This sediment, moreover, seems to be
considerably denser in the case of the Rhone than that of the
Mississippi ; whilst the alluvium of the Danube is, for the
most part, much lighter. Accordingly, the condition of the
Rhone as regards sediment is more unfavourable, both as
to quantity and density, than that of the Mississippi ; whilst
the condition of the Danube is superior in both these respects
to the Rhone and the Mississippi. The flatness, however, of
i82 Comparison of Conditions of Deltas. [CHAP.
the sea-slope in front of the Danube delta places the Danube
in this particular at a disadvantage in comparison with the
Rhone and the Mississippi ; but the Volga is considerably
more unfavourably affected by the shallowness of the northern
part of the Caspian Sea. The Volga, on the other hand,
appears to carry down a comparatively small proportion of
sediment; whilst the Nile seems to be more heavily charged
than the other rivers described. The advance of a delta
depends on the volume of sediment deposited, and the sea-
slope ; and thus the rates of advance of the Mississippi and
Danube deltas are fairly similar, owing to the different pro¬
portions of sediment being compensated by differences of
sea-slopes. Though the proportion of sediment in the Volga
appears to be considerably less than even in the Danube, the
advance of its delta is probably more rapid, owing to the
flatness of the sea-slope in front of its mouths. The advance
of the several branches of a delta varies with the relative
discharges ; and though a larger discharge provides generally
a better depth over the bar, it is accompanied with a more
rapid advance of the mound of deposit, and with a greater
protrusion of the bar out to sea.
A littoral current retards the advance of a delta by sweeping
away the lighter alluvial materials brought within its influence.
An easterly littoral current carries the turbid waters from the
Damietta mouth of the Nile across the entrance to Port Said
harbour (Plate 13, Fig. 4) ; a southerly current exists in front
of the Danube delta, which is aided by the prevalent north
winds ; and a permanent current, extending to depths of
about 330 feet, travels from east to west along the Mediter¬
ranean shore of the Rhone delta, which may account for the
smaller rate of advance of this delta in former times as com¬
pared with the average advance of the Mississippi delta, in
spite of similarities as regards proportionate sediment and
sea-slope. There appears to be no natural littoral current
along the northern coast of the Caspian Sea ; and the pre-
vin.] Improvement of Outlets of Deltas. 183
valent winds, blowing up and down the Volga delta, can¬
not create one. A westerly current flows across the outlets of
the Mississippi, mainly produced by the prevalent winds, aided
somewhat by the feeble tidal wave entering the Gulf.
Methods of Improvement. The above descriptions of
deltas show what serious obstacles to navigation these allu¬
vial accumulations, formed by the rivers themselves, present.
These natural deposits cannot be prevented ; and their gradual
prolongation increases the difficulties of the situation by
lengthening the channels, and reducing by degrees the fall
of the river near its outlet, promoting the subdivision and
shoaling of the branches. The largest rivers possessing an
excellent navigable channel for great distances inland, may
thus be practically cut off from proper access to the sea, and
thereby debarred from external communications. The re¬
moval, therefore, of these barriers to ocean-going trade is of
the highest importance to the extensive districts served by
such rivers.
Three methods have been adopted for securing improved
navigable communication between these rivers and the sea,
namely: (1) Dredging or harrowing on the bar ; (2,) Prolonga¬
tion of outlet channel by parallel jetties ; and (3) Construc¬
tion of a ship-canal. The first method aims at the direct
lowering of the obstruction at the bar by dredging, or by
stirring up the material so that it may be carried away by
the outgoing current. The second system effects the increase
in depth by the scour of the concentrated current across the
bar. By the last plan, the difficulties of the delta are evaded
by connecting the river, above the delta, with the sea by
a ship-canal emerging at a point of the coast beyond the
influence of the alluvial deposits. The deepening of the
channel over the bar of one of the outlets was, in the first
instance, attempted, chiefly by harrowing, at the Danube
and Mississippi deltas ; and dredging has been adopted
for improving one of the main outlets of the Volga delta.
184 Dredging Danube and Mississippi Bars. [chap.
Embankments were constructed for prolonging one of the
outlet channels of the Rhone, and thus directing the issuing
current across the bar ; and jetty works have been carried out
with the same object at the mouth of one of the branches of
the Danube, and of the Mississippi. Lastly, a ship-canal was
proposed for affording an outlet for the trade of the Danube,
and also of the Mississippi, beyond the zone of deposit of the
delta ; and ship-canals have been made for connecting the
Rhone and the Nile with the sea beyond the limits of their
deltas.
Dredging and Harrowing on the Bar.
The direct removal of the crest of the bar in front of the
outlet of one of the channels of a delta, to a depth suitable
for navigation, by dredging or harrowing, was naturally the
expedient first resorted to.
Dredging and Harrowing at Outlets of Danube and
Mississippi. The Sulina branch of the Danube was the only
navigable channel to the sea in 1857 ; and as a first attempt
at improvement, a rake was dragged across the bar, according
to the old Turkish method for providing a temporary deepen¬
ing ; and a dredger was also employed for throwing up and
putting into suspension the material forming the bar. These
measures, however, produced only a transitory improvement
of the outlet channel ; for the river current, becoming enfeebled
on emerging into the sea, was unable to retain in motion all
the material it had brought down, and therefore was incom¬
petent to bear an increased burden into deep water. A
similar system was resorted to at the bar of the South-West
Pass of the Mississippi delta, on a larger scale, between 1852
and 1875, harrowing on the bars having been tried as early as
1726 ; but an increase in depth, from 13 to 18 feet, was with
difficulty obtained, for a width of 300 feet, by large dredging
operations ; and the channel soon reverted to its original
average depth of 13 feet whenever the dredging was suspended.
vin.] Dredging Outlet Channels of Volga. 185
Dredging at the Mouths of the Volga. The depth over
some banks obstructing the mouth of the Bakhtemir branch
of the Volga delta was formerly little over 4 feet at the
normal water-level, and was reduced to 2 feet by land winds.
These banks have been lowered by dredging, since 1874, so
as to provide a depth of 8 feet at mean level, for a width of
420 feet, the total length of dredged channel over the several
banks amounting to 5! miles ; and the maintenance of this
depth is effected by raising 76,000 cubic yards annually with
one dredger1. To obtain, however, an additional foot in
depth in the outlet channel of this branch, it would be neces¬
sary to dredge along a length of at least 17 miles. Moreover,
the advance of the delta in front of this branch must be
more rapid than at other parts, as one-third of the total
discharge of the river flows through this outlet ; the branch
is narrow and winding in places ; there are some shoals in it
which have to be dredged ; and the distance by this route
between Astrakhan and the sea is about 67 miles. Accord¬
ingly, it is now proposed to revert to the smaller Kamysiak
branch, at one of the mouths of which a futile attempt was
made, in 1858-69, to deepen the channel over the bar by
directing the current between training banks of fascine
mattresses aided by dredging, and concentrating a larger dis¬
charge in the selected channel by closing the minor outlets.
The Kamysiak branch has a minimum depth of 14 feet
between Astrakhan and its bar ; its channel is fairly straight ;
and its bar, with a normal depth over it of 4! feet, is about
12 miles wide, and therefore considerably narrower than the
banks which impeded the mouth of the Bakhtemir branch.
Moreover, the 3-fathom line is much nearer this outlet ; and
the distance from the sea to Astrakhan by this channel
is only about 46 miles. Owing also to its much smaller
discharge than that of the Bakhtemir branch, the advance
' Les Embouchures du Volga.' V. E. de TimonofF. Vme Congrès International
de Navigation intérieure, Paris, 1892.
i86 Delta Outlets improved by Jetties. [chap.
of the delta at its outlet must be much slower. It is,
accordingly, believed that a 14-foot channel might be
obtained at one of the mouths of the Kamysiak branch by
adequate dredging plant, which would suffice for vessels
navigating the Caspian Sea, and dispense with the present
necessity of transhipping cargoes, in the open sea, to vessels
of small draught in order to convey them to Astrakhan.
Parallel Jetties at Outlet.
The second method adopted for deepening one of the
outlet channels of the delta of a tideless river, consists in
prolonging the channel beyond the coast-line by a jetty on
each side, so that the current, instead of spreading out at
the mouth, may be continued out to the bar, and increase
the depth over it by scour. The advantage of this system
is that the additional depth is maintained by the same scour¬
ing force which created it, instead of necessitating constant
works of maintenance, such as are required in the case of
improvements by dredging in channels exposed to sedi¬
mentary deposit. It is essential, however, for the success of
this system, that the sea-bottom should shelve down beyond
the bar ; for the deposit removed by scour from the bar is
only carried somewhat further out to sea by the prolonged
current, and deposited when the current slackens beyond the
ends of the jetties, except in so far as it may be partially
borne away by any littoral current. If, therefore, the sea
continues shallow beyond the bar, the effect of the jetties
is merely to push the bar a little further out in proportion
to their length ; and this, in the case of the Volga jetty works,
together with the leakage through the interstices of the
jetties, and the increase of the volume of alluvium by the
barring of the other outlets of the Kamysiak branch, explains
the failure of these works.
The amount of permanence of the improvement in depth
effected by parallel jetties is dependent upon four conditions;
vin.] Conditions of Improvement by Jetties. 187
namely, the slope of the sea-bottom beyond the bar ; the
density of the alluvium brought down ; the power and extent of
any littoral current or wave-action ; and the rate of advance of
the delta at the outlet selected. A good depth seawards defers
the period when the accumulation of deposit further out will
rise high enough to form a new bar. Alluvium of too great
density to be carried in suspension is less affected by the
concentrated current than the lighter materials, and therefore
comes to rest again nearer the coast-line, and is brought less
under the influence of any littoral current. A strong littoral
current and powerful wave-action carry away some of the
alluvium, and therefore delay the formation of a new bar
beyond the jetty channel. The rate of advance of the shore
in front of the outlets of any special delta is proportionate to
the discharge, for similar depths of the sea in front ; and this
consideration affects the choice of the outlet, and is the only
one of the four conditions which is subject to modification.
Bhone Jetties. In 1852, previously to the commencement
of jetty works at the mouth of the Rhone, the Piémanson and
Roustan branches (Plate 5, Fig. 1) had depths at their
outlets of 91 feet a third of a mile inside the bar, and 5¿
feet over the bar ; and the former had a depth of 331 feet
a third of a mile beyond the bar, and the latter 63 feet at
the same distance out. The depths at the Eugène outlet, at
similar positions, were io£, 4!, and 58t feet ; and for the East
outlet, 5j, 3s, and 13! feet1. Evidently, therefore, the
Roustan outlet was the best for improvement, as it possessed
as good depths up to, and over the bar, as the Piémanson out¬
let, and a much better depth seawards, a most important con¬
dition for a river charged with dense detritus : whilst the East
outlet was in every respect much the worst as regarded depth.
Nevertheless, this eastern branch was selected, on account
of its more favourable direction for the entry and exit of
vessels, a matter of little value if the depth is inadequate.
1 'Cours de Navigation intérieure.' M. Malézieux, p. 179.
188
Jetties at East Mouth of Rhone. [chap.
Embankments were carried out along each side of the
East channel in 1852-57. from above the entrances of the
other branches to within half a mile of the bar (Plate 5,
Fig. i ) ; and the three southern channels were thus shut off,
as well as the two minor northern ones ; and the whole
discharge of the Rhone was concentrated in the East branch.
The increased scour thereby produced at the eastern outlet
removed the existing bar, and for a time materially deepened
the outlet channel, increasing the average available depth from
43 feet in 1852, up to 9f feet in 1856. The bar, however,
gradually formed again further out ; and the average depth
over it has never since attained the moderate increase in depth
of 1856, having ranged between 8f feet in 1873 and 3I feet
in 1890. The maximum depth since 1856, which results
from great floods, reached nf feet in 1859 and 1873; and
the minimum depth of slightly under 2 feet occurred in 1890,
during which year the depth never exceeded 61 feet. Three
states of the bar, in 1841, 1873, and 1891 respectively1, are
shown in Plate 5, Fig. 9, from which it appears that the bar
is a mile further out than fifty years ago, the main advance
being due to the embankments, though a distinct advance
since 1873 is also manifest. The three sections also show how
small has been the improvement effected by the parallel
jetty system as applied to the Rhone, even in an exception¬
ally favourable year such as 1873. The average depth,
indeed, over the bar during the last thirty years amounts
to only feet, a gain of 3 feet over the depth of the eastern
bar in 1852, but slightly less than the depth over the Eugène
bar in 1841. Moreover, the bar having been pushed further
out from the shore by the increased discharge, is in a more
exposed situation. The steep sea-slope of the bar is the
result of the large proportion of heavy sediment rolled along
the bed forming the bar, which comes to rest before the
1 Charts of the Rhone in 1S41 and 1S91 were furnished me by Monsieur
A. Guérard, the engineer in charge of the maritime part of the river.
vin.] Causes of Failure of Rhone Jetties. 189
current passes the crest of the bar, and is only gradually
pushed over on to the outer slope.
The failure of parallel jetties to effect an adequate deepen¬
ing at the Rhone outlet is due to the course adopted, rather
than to any inherent defect in the system, or specially
unfavourable physical conditions. The waters of the Rhone,
indeed, bring down a rather large proportion of sediment,
and a notably large percentage of this sediment is heavy ;
but this unfavourable circumstance is somewhat compensated
for by the existence of an oceanic littoral current extending
to a considerable depth, and by the declivity of the bottom
of the Mediterranean Sea along the southern front of the
Rhone delta. The selection, however, of the eastern mouth
as the navigable channel, and the closing of the other
branches, were fatal to the success of the works. The eastern
outlet faces the prevalent and strongest winds blowing from
the sea, and it discharges into the shallow sheltered Gulf of
Foz, away from the influence of the littoral current ; and
the limitation of the discharge to this single outlet, whilst
increasing the scour, has concentrated the advance of the
delta on to this part, the comparative shoalness of which
increases the rate of progression. The advance, indeed, of
the delta across the gulf already impedes the access of
vessels to the ports of Bouc and St. Louis during northerly
winds ; and the reopening of the southern mouths to check
this advance has been proposed. The southern shore of the
delta, on the contrary, has been eroded in places under the
action of the sea, since the cessation of deposit at this part
owing to the closing of the southern outlets.
If one of the southern mouths of the Rhone had been
selected for improvement by parallel jetties., preferably the
Roustan branch, there is every reason to believe that the
navigable channel might have been adequately deepened, in
spite of the density of the sediment, owing to the good depth
in front, and the eroding influence of the waves and littoral
i9° Jetties at Salina Mouth of Danube. [chap.
current. Moreover, by leaving the other entrances open, so
as to spread the alluvium discharged over a wide area, the
rate of advance of the delta would not have been accelerated,
and the entrance to the Gulf of Foz would not have been
imperilled 1.
Jetties at Sulina Mouth of Danube. The Kilia branch
of the Danube possesses the best navigable channel through
the delta ; but the number of its mouths, and the small depth
over their bars, the rapid advance of its delta owing to its
large discharge, and the flatter slope of the sea-bottom there
than further south, rendered it unsuitable for improvement
(Plate 5, Fig. a). The St. George's branch was preferred by
Sir Charles Hartley, the engineer to the Danube Commission,
owing to its having a much better navigable channel through
the delta than the Sulina branch, on account of the greater
depth of the sea in front of its outlet, and as its mouth is
20 miles nearer the Bosphorus, and more favourably situated
for navigation than the Sulina mouth2. The Sulina mouth
was, however, selected for provisional improvement in 1858,
on account of its being at that time the only navigable outlet,
and owing to its bar being only half the distance from the
shore of the St. George's bar, and the cost of the works there¬
fore much less. The jetties were carried out to a depth of
18 feet of water in 1858-61 ; and they were consolidated in
1866-71, owing to their success and the insufficiency of funds
for improving the St. George's mouth. They converge from
the shore to the edge of the outlet channel, and then extend
in a parallel direction, 600 feet apart, out from the coast3,
beyond the site of the bar in 1857, the crest of which was
1 ' Amélioration de la partie maritime des Fleuves, y compris leurs Embouchures.'
L. F. Vernon-Harcourt, Vm° Congrès International de Navigation intérieure,
Paris, 1892.
2 ' Description of the Delta of the Danube.' C. A. Hartley. Minutes of Pro¬
ceedings Institution C. E., vol. xxi, p. 283.
3 ' Description of the Delta of the Danube.' Sir Charles A. Hartley. Minutes
of Proceedings Institution C. E., 1873, vol. xxxvi, p. 201.
vin.] Construction and Cost of Sulina Jetties. 191
only I of a mile beyond the Sulina mouth (Plate 5, Figs.
3 and 4). The jetties were formed in the first instance of
a mound of rubble stone and pilework, and consolidated
subsequently by a concrete wall along the top, and random
concrete blocks, from 10 to 20 tons in weight, on the
sea-slope of the mound. The south jetty, which originally
was overlapped by the north jetty, was extended 457 feet
in 1869, and 204 feet more in 1876-77, so as to terminate at
the same distance out as the other jetty, and thus scour
away the sand which was heaped up against the inside of the
projection of the north jetty under the influence of south¬
westerly winds. The shore end also of the north jetty was
gradually extended landwards, between 1867 and 1883,
a total length of 1,100 feet, to maintain its connexion with
the shore which has been eroded by the waves and littoral
current to the north of the jetties, owing to their projecting,
like a groyne, about a mile out from the coast. These jetties
directed the scour of the river across the bar, and pushed
its crest out seawards beyond the pierheads, so that the
navigable depth over it was gradually increased from 9 feet
in 1857, up to 20^ feet in 1872, which depth has been since
maintained (Plate 5, Fig. 10). The total cost of the works
amounted to about ^197,000, and the maintenance during
the last twenty years has averaged about ^2,300 a year.
The shelter afforded by the jetties on their southern side
from the waves raised by northerly gales, and from the littoral
current flowing from the north, has led to the deposit of
alluvium brought in by south-westerly winds, producing an
advance of the shore and of the lines of soundings to the
south of the jetty channel. * These accumulations deflect the
deepest channel towards the north outside the pierheads,
so that sometimes the depth falls below 18 feet in the direct
line seawards of the jetty channel1.
1 ' The Sulina Mouth of the Danube.' C. H. L. Kiihl. Minutes of Proceedings
Institution C. E., 1888, vol. xci, p. 331, and plate 3, fig. 2.
192 Causes of Success of the Sulina fetties. [chap
The works were expected to procure a depth of 16 feet
at the outlet ; and therefore the results achieved were 4 feet
in excess of the depth anticipated. Moreover, the navigable
depth of 2o| feet has been maintained for twenty years
without any extension of the jetties, with the exception of
the small addition to the southern jetty in 1876-77 in com¬
pletion of the original design. The success attained is due
to the lightness of the alluvial matter brought down, and
its removal on being brought under the influence of the
littoral currents by the action of the jetties in conveying
the discharge seawards. The lightness of the alluvium is
indicated by the comparatively small proportion of material
deposited within 1 £ miles of the pierheads, amounting to less
than one-sixth of the material carried in suspension. More¬
over, a still larger proportion of the suspended matter must
be conveyed by the currents beyond this limit before being
deposited, as some of the deposit measured must consist
of heavier sediment rolled along the bottom, the volume of
which has not been ascertained. The form of the bar also,
and the great contrast it presents on its sea-slope to the
Rhone bar (compare Figs. 9 and xo in Plate 5), show that
a considerable portion of the sediment passes over the crest
of the bar before being deposited by a kind of sifting process
according to its density, and that, therefore, a large propor¬
tion of the sediment forming the bar is fairly light. The
moderate ratio, moreover, of the sediment to the discharge
as compared with that of the Rhone, and the smaller density
of the Black Sea than of the Mediterranean, owing to the
smaller proportion of salt in the former, are conditions favour¬
able to the maintenance of the depth at the Sulina mouth.
The comparatively small discharge also through the Sulina
branch, and the non-interference with the discharges through
the other outlets, retard the rate of shoaling in front of the
Sulina jetties, and postpone the necessity for their extension.
In fact the slope of the sea-bottom is the only condition which
vin.] Shoaling in front of Sulina Month. 193
is less favourable at the Sulina mouth than at the mouth
of the Rhone; whilst the errors committed in the selection
of the outlet, and in closing the other outlets at the Rhone
delta, have been avoided at the Danube delta.
Notwithstanding the favourable conditions under which
the jetty system has been carried out so successfully at
Sulina, these works must not be regarded as securing an
absolutely permanent maintenance of the existing navigable
depth. Though the depth of 20 feet attained in 1872 has
hitherto been maintained, the sections over the bar in 1873
and 1891 (Plate 5, Fig. 10) show that deposit is extending
on the outer slope of the bar ; and the 4- and 5-fathom lines
of soundings have progressed seawards a little over a quarter
of a mile, in the direct prolongation of the jetty channel,
between 1871 and 1891 (Plate 5, Fig. 3), amounting to an
average advance of 70 feet a year ; whilst the gradual advance
of the shore-line to the south of the jetties is clearly marked h
The state of the sea-bottom in front of the Sulina jetties may
present fluctuations resulting from variations in the volume
of the floods of the Danube, the amount of sediment brought
down, and the direction of the winds, occasionally reverting
to a more favourable condition, as in 1891 compared with
1886 ; but eventually, sooner or later, the accumulation of
deposit must gradually creep in front of the line of the jetty
channel, and the depth in the navigable channel be at last
reduced below 20 feet. A moderate prolongation, however,
of the jetties will suffice to scour the bar again, and driving it
into deeper water, secure the requisite navigable depth in the
outlet channel for another period of years.
The outlet channel having had its depth more than doubled
over the bar, afforded a much better depth than many parts
of the Sulina branch at a low stage of the river. Accordingly,
1 Sir Charles Hartley supplied me with a Chart of the Danube of 1891, from
which the lines of soundings for 1891 on the plan, and the section of the bar in
1891, have been obtained.
O
194 Regulation of Siilina Branch of Danube, [chap.
the available depth throughout this branch has been improved
by groynes to reduce excessive widths, training banks, easing
bends, dredging shoals, and making cuts across very bad
bends ; one of these cuts, a little over 5 miles in length in
place of nearly xo miles of winding channel, having been
executed in 1890-94 1 (Plate 5, Fig. 2). These works have
already increased the minimum depth in the Sulina branch,
at the lowest stage of the river, from 7 feet in 1857 to 13 feet
in 1871, and 16J feet in 1891.
The tonnage of vessels outward bound with cargoes through
the Sulina mouth, which averaged about 300,000 tons in
1857, rose directly after the improvement of the channel,
and reached nearly i-| million tons in 1889.
Mississippi South Pass Jetties. The South-West Pass
was proposed by Mr. Eads for improvement by parallel
jetties at its outlet, on account of its possessing a larger
channel through the delta than the South Pass, and owing
to the South Pass being impeded by a shoal at its head.
Eventually, however, the smaller South Pass was chosen for
improvement, as in the case of the Danube, owing to the
much shorter jetties needed for directing the currents out
to the bar at this pass than at the South-West Pass, whose
bar was twice as far off from the shore 2.
The works at the outlet of the South Pass consist oi two
parallel jetties about 1,000 feet apart, formed of long willow
mattresses consolidated with limestone, and with large concrete
blocks at their outer ends (Plate 5, Figs. 6 and 8), carried out,
in 1876-79, for lengths of 2 J and i| miles on the east and west
sides of the channel respectively, from the shore out to the
bar. They terminate, at the same distance out, in 30 feet
of water, and were slightly curved round towards the south at
their extremities, to direct the discharge of the river at their
1 ' The Sulina Branch of the Danube.' C. H. L. Kiihl ; Minutes of Proceedings
Institution C. E., 1891, vol. cvi, p. 239.
2 ' A History of the Jetties at the Mouth of the Mississippi River.' E. L. Corthell.
vin.] Improvement of Mississippi South Pass. 195
outlet at right angles to the littoral current flowing from east
to west across the ends of the jetties. The jetty channel has
been further regulated, and the scour along the central portion
increased, by the addition of an inner line of training banks
of willow mattresses on each side (Plate Fig- 6). The
concentrated current at the outlet produced by these works
soon scoured out the channel between the jetties, and lowered
the bar beyond ; so that by 1880, an outlet channel had been
formed with a minimum central depth of 31 feet throughout ;
whereas in 1875 there was a depth of only 8 feet over the
bar (Plate 5, Fig. 7).
The shoal at the head of the South Pass, where it leaves
the main river, which had a depth of water of only 15 feet
over it, was lowered, in 1876-77, by contracting the funnel-
shaped entrance to the South Pass, by mattress dykes on
each side, into a uniform channel 850 feet wide. These
dykes concentrated the current entering the pass, over the
shoal ; and after dredging a narrow channel through the
shoal to a depth of 18 feet to hasten the deepening, the floods
of 1877 scoured out a good channel; and by 1880 a depth
of 30 feet had been attained across the shoal. As these
dykes, by narrowing the entrance to the South Pass, and
therefore somewhat checking the influx from the river into
this pass, would have diverted a portion of the discharge into
the other passes, mattress sills, 70 feet in width, were sunk
across the beds of the entrances to the South-West Pass and
the Pass à l'Outre, so that by slightly reducing the sectional
areas of these channels, the influx into them might be similarly
checked, and the proportionate discharges into the three
passes might remain unaltered.
The concentration and prolongation of the current at the
outlet by means of the parallel jetties, not merely lower the
bar, carrying the material scoured away, and the sediment
constantly brought down, into deeper water, but also bring
the discharged alluvium more under the influence of the
o 2
ig6 Shoaling in front of Mississippi fetties. [chap.
stronger littoral current further out from the shore, till the
gradual advance of the shore-line at the back of the jetties
shall have pushed the littoral current further out into the gulf.
Though, however, the lighter alluvium may be carried away by
the littoral current, the heavier material is merely transported
into the deeper water beyond the ends of the jetties, where
shoaling is gradually taking place (Plate 5, Fig. 7). In fact,
within a fan-shaped area of i¿ square miles in front of the
outlet, the decrease in depth between 1876 and 1892 has
amounted, on the average, to over 13 feet over the whole area h
the shoaling having reached a maximum of 2X% feet in the year
1890-91. Altogether, the greatest amount of shoaling has
occurred in the outer zone of this area, from 4,000 to 6,000
feet beyond the ends of the jetties ; but in 1890-91 and
1891-92 the shoaling was greatest in the central zone. The
least advance of the lines of soundings beyond the ends of
the jetties, between 1877 and 1892, occurred at the 20-foot
line amounting to an average of 810 feet, or about 54 feet
annually, the advance of the 30-foot line averaging 63 feet
annually, but exhibiting the maximum advance of 240 feet in
1891-92. Further out the rate of advance increases, and
reaches a maximum at the 70-foot line, where the average
advance was 1,555 feet from 1877 to 1892, or 104 feet a year.
The rate of advance decreases in greater depths, beyond the
limits of the area of 1J square miles ; but even up to the limit
of the soundings, at the ioo-foot line, the annual progression
is large, attaining an average of 87 feet at the 8o-foot line,
83 feet at the 90-foot line, and 81 feet at the ioo-foot
line. A new bar, accordingly, is in process of formation
about a quarter of a mile beyond the ends of the jetties ;
so that in 1891, the depth in the prolongation of the central
line of the jetty channel had been reduced to a minimum of
15 feet. The deep outlet channel also, which still maintains
1 'Annual Report of the Chief of Engineers, U.S.A., for the year 1892,' part 2,
p. 1478.
vin.] Dredging at Outlet of Mississippi. 197
a minimum depth of 30 feet, had been deflected to the east by
the accumulation of deposit on the western side of the outlet ;
and on emerging from between the jetties, diverged about 500
feet to the east of the east jetty, and was very tortuous and
narrow in places. Dredging has consequently been resorted
to for forming a better and more direct channel (Plate 5,
Fig. 6). During eleven days, moreover, in 1891, and for
a month in 189a, the central depth in the jetty channel fell
below the statutory depth of 30 feet. It is evident, therefore,
that though the South Pass jetties have hitherto maintained
a channel of 30 feet in depth at the outlet, this channel can
hardly even now be described as suitable for navigation ;
dredging in front of the jetties will become increasingly
necessary; and there is every prospect that the accumula¬
tion of deposit taking place in front of the outlet, will
before long necessitate the prolongation of the jetties, to
make the deep channel more direct, and to maintain a navi¬
gable channel of 30 feet at the outlet for another period
of years.
Mississippi and Danube Jetty Works compared. The
jetty works at the mouth of the Mississippi have been
carried out on precisely the same principles as the Danube
jetty works ; though the greater size of the South Pass, the
longer distance of its bar from the shore, and the deeper
channel required - for the trade of the Mississippi, have
necessitated more extensive jetties than at Sulina (Plate 5,
Pigs- 3) 7, and 10). The outlet of the smallest branch
through the delta was selected for improvement in both
cases; parallel jetties were carried out to the bar; and the
discharge through the other branches was not interfered
with. The results also of the works have been very similar :
the bar in both cases has been pushed out seawards by the
concentrated current, and the depth over it was at once
increased, and has since been fairly maintained ; whilst the
influence of the littoral current on the alluvium has, in both
198 Danube and Mississippi Outlets. [chap.
instances, caused a diversion of the deepest channel away to
one side from the direct line of the jetty channel ; and the
advance of the lines of soundings seawards, owing to
the gradual deposit of the alluvium of the river in front
of the outlet, has been most marked at some distance beyond
the ends of the jetties. The advantage of the sea-slope in front
of the South Pass being four times steeper than in front of
the Sulina mouth, has been neutralised by the sediment
brought down by the Mississippi being much greater in
proportion to the discharge, and also much denser than the
alluvium of the Danube. Moreover, the absence of the erosion
by the sea along the shore of the South Pass, which has
produced an increase in depth to the north of the Sulina
jetties, combined with the unfavourable character of the
alluvium, and the necessity of maintaining a minimum
navigable depth of 30 feet at the mouth of the Mississippi,
render the prospect of a prolongation of the jetties more
imminent at the South Pass than at Sulina.
Ship-Canal instead of a Delta Channel.
Owing to the natural impediments to navigation existing
at the outlets of a delta, it has sometimes been proposed to
connect the river, above its delta, with the sea by means of
a ship-canal. This plan resembles in principle the lateral
canals which are resorted to for forming a connecting link
between the two navigable portions of a river, in consequence
of the intervention of rocky rapids, or the existence of an
insurmountable obstacle such as the falls of Niagara. The
new outlet must be situated beyond the influence of the allu¬
vium of the river, otherwise it would become silted up. Thus
the deep-water entrance to the Tiber, provided in the reign
of the Emperor Claudius, by the construction of the port
of Ostia, and connecting it with the river by a canal, was
within the range of the alluvial deposits at the mouth of the
vin.] Ship-Canals to avoid Delta Outlets. 199
Tiber ; and consequently the port gradually silted up, and is
now two miles from the coast.
In a few instances, a ship-canal has been formed, diverging
sufficiently from the delta to provide a deeper permanent
outlet for the navigation of a tideless river than the channels
through the delta can afford. Thus the San Carlos Canal,
starting from the river Ebro at Amposta and emerging into
the Bay of Alfaques to the south of the delta of the Ebro,
provides a navigable channel for the trade of the river, which
is cut off from its natural outlet by the impediments of the
delta projecting into the sea, and barred by accumulations
of sand. The Mahmoudieh Canal leaves the Rosetta branch
of the Nile at Atfeh, about 30 miles from the Rosetta mouth ;
and after a westerly course of 50 miles, it enters the harbour
of Alexandria. This canal, which cost ,£300,000, and was
opened in 1830, is only 100 feet broad and of a very moderate
depth ; but it affords a deeper and much more sheltered
outlet for the traffic of the Nile than either of the shallow
and exposed mouths provides. In this case, the great
variation in the discharge of the river at high and low Nile,
and the paramount importance of irrigation works for Egypt,
preclude any attempt to improve one of the delta channels.
The construction of a ship-canal for avoiding the obstacles
of the Volga delta has been considered, but has been
abandoned, as the canal would have to be given a length of
135 miles in order to provide Astrakhan with an outlet to
the Caspian Sea beyond the influence of the delta ; and the
cost of such a work would be prohibitive.
A ship-canal was one of the proposals submitted for
improving the navigable outlet of the Danube, diverging to
the south of the St. George's branch, so as to avoid the shoals
in front of its mouths ; but the improvement of the Sulina
branch by parallel jetties was fortunately preferred, as it has
provided an unimpeded and deeper channel formed by
natural scour.
2oo Dredging, and Jetties at Delta Outlets, [chap.
The formation of a ship-canal for providing a deep-water
outlet for the Mississippi was advocated by several persons
in opposition to the jetty system ; but eventually the ad¬
vocates of the jetty system prevailed, with the result that an
unimpeded outlet with a greater depth than proposed for the
canal has been obtained.
Conclusions. Harrowing on the bars of the Danube and
the Mississippi failed to produce any material or permanent
improvement ; and owing to the inability of the issuing
current to transport its own burden of sediment, the increase
of its charge by stirring up the deposit offers no prospect of
securing any adequate increase in depth.
Dredging proved inefficacious to secure an adequate
deepening of the outlet channels of the Danube and the
Mississippi ; and it is evident that this method of improve¬
ment would have to be carried out continuously on a very
large scale to cope with the great volume of alluvium brought
down by such rivers. It is proposed, indeed, to adopt this
system of improvement in the case of the Volga ; but the
small depth in front, and the absence of any littoral current
preclude the use of the jetty system in this instance ; and the
increase in depth anticipated is small. Dredging alone, accord¬
ingly, appears chiefly applicable to the improvement of the
outlets of tideless rivers when, owing to unfavourable local
conditions, the jetty system cannot be resorted to, or in cases
where the volume of alluvium brought down is comparatively
small. Dredging is to be commenced very soon at Sulina, with
a bucket-ladder hopper dredger of large dimensions provided
with a sand pump, in order to endeavour to increase the
depth of the entrance channel of the Danube across the Sulina
bar to 23 or 24 feet. These operations will be of considerable
interest, in showing to what extent dredging can be employed
for increasing the improvement effected at an outlet by jetties,
and also, eventually, in proving whether dredging is economi¬
cally capable of dispensing with, or at any rate of deferring
vin.] Jetty System preferable to Ship-Canal. 201
for several years, the prolongation of the jetties for maintaining
the depth. Dredging is likely to prove more applicable at
the Sulina mouth than at the South Pass outlet of the Missis¬
sippi, owing to the smaller volume of alluvium brought down by
the Sulina branch, and the smaller depth of the sea in front.
The jetty system has produced most successful results at
the Sulina mouth of the Danube and the South Pass of the
Mississippi ; and the absence of improvement at the Rhone
outlet is fully accounted for by the unfavourable direction of
the outlet selected, the concentration of the discharge into
a single mouth by closing the other mouths, the absence of
a littoral current in front of the trained outlet, and the density
of the sand brought down. Parallel jetties may, indeed, be
regarded as furnishing a satisfactory method of improving the
outlet channel of tideless rivers, provided the mouth selected
is suitably situated, with deep water in front and a littoral cur¬
rent sweeping across it, and the other mouths are not closed.
The volume of material discharged, and its density, affect the
results obtained ; and the cost is reduced if one of the smaller
channels, just adequate for navigation, is selected, where the
progress of the delta is slower and the distance out of the bar
is less. Though the system does not constitute an absolutely
permanent improvement, the jetties generally suffice to
maintain the depth for several years, till a bar has had time
to form further out ; and eventually an extension of the
jetties would ensure the maintenance of the depth for another
period of years.
The evasion of the difficulty at the outlet, by the con¬
struction of a ship-canal with its outlet beyond the influence
of the delta, should not be resorted to unless the jetty system
is inapplicable, as it generally would constitute a work of con¬
siderable magnitude to secure an adequate depth, its passage
is necessarily impeded by a lock to shut out the turbid river
current, and any increase in depth at its outlet has to be
procured by dredging instead of by the natural scour.
CHAPTER IX.
JETTIES AND BREAKWATERS AT THE MOUTHS
OF RIVERS.
Bars : fluvial, marine ; origin of; favourable and unfavourable conditions. Jetties
at Outlets of Tideless Riversfree from Silt : Swinemünde, Vindau, Dwina, Pernow,
Chicago, Buffalo, and Oswego. Converging Breakwaters at Mouths of Tidal Rivers :
objects ; Liffey, Merrimac, Tees, Tyne, Wear, Nerv ion ; Remarks. TrainingJetties
at Mouths of Tidal Rivers : objects and instances ; Yare, Adur, Adour ; com¬
bined with training works on Maas up to Rotterdam, and on Nervion up to Bilbao.
Remarks on woiks at Outlets : deepening aided by dredging; materials for jetties ;
alteration of outlet.
Bars. The rivers described in the preceding chapter them¬
selves form the bars at the outlets of their delta channels, by
the alluvium which they bring down from inland. The bars,
however, found at the mouths of tideless rivers which are not
in the process of forming deltas, owing to the small amount of
material they carry into the sea, and at the mouths of tidal
rivers in which the tidal volume is much larger than the fresh¬
water discharge, cannot be attributed to this cause. In the
case of these latter rivers, the bar is of marine, and not of
fluvial origin, and is the result of the heaping-up action of the
waves along the coast, tending to form a continuous beach
across the mouth of the river, which, in most instances, is only
partially prevented by the scour of the tidal ebb and flow,
and by the fresh-water discharge of the river. The river
Neva is an instance of a tideless river which has a marine bar
at its mouth ; for though it brings down a large quantity of
alluvium in its course, it deposits it all in Lake Ladoga, thirty-
six miles above its outlet into the Gulf of Finland, an arm of
Bars at the Mouths of Rivers. 203
the Baltic Sea. The same is the case with the outlets of the
Oder into the Baltic, for this river, like the Neva, though
deltaic in appearance, does not at the present time discharge
much alluvium into the sea, as its burden of sediment is de¬
posited in some lakes and the Stettiner Haff before reaching
the sea. The outlet channels, also, of the Rhine and the Maas,
which, discharging a large volume of water charged with
alluvium into a part of the North Sea where the rise of tide is
small, have formed a regular delta in ancient times, must now
be classed as tidal rivers, since the sea tends to encroach upon
the land along that coast, and prevents any advance of the
foreshore.
At the mouths of rivers with a large tidal flow, the marine
origin of the bar is evident, both from the nature of the
materials of which it is composed, and also on account of
the inadequacy of a moderate fresh-water discharge to supply
the large volumes of detritus encumbering many large tidal
estuaries.
The term bar is generally applied to the ridge or kind of
submerged beach which intervenes between the mouth of the
river and deep water, stretching out seawards in front of the
outlet in the line of the greatest scour of the river, and curving
round to form a junction with the foreshore on each side, as
very clearly defined in the chart of the Mersey estuary (Plate 7,
Fig. xi). The outer, lowest part of the ridge, in the line of
the deepest channel, is actually called the bar, for it is this
portion, which being higher than the navigable channel on
either side, impedes the passage of vessels of large draught
between the river and the sea. A distinct hollow is found in
most rivers just inside the bar, which must be due to the
action of waves and of the flood tide in crossing the bar
(Plates 6, 7, 8, and 9).
Shoals often _ exist across the navigable channel in an
unimproved river, over which the depth of water is as
little, or even less than over the bar outside ; but being fairly
204 Conditions affecting Height of Bars. [chap.
sheltered, they can generally be lowered without great
difficulty by dredging or training works. The bar, on the
contrary, being fully exposed, and having the available depth
over it reduced by the wave oscillations during storms, presents
a more serious danger to navigation, and is far more difficult
to improve than inner shoals. The bars, however, at the
mouths of tideless rivers free from alluvium, and of tidal
rivers, when once adequately trained or protected by jetties or
breakwaters, are not liable to form again beyond the extremi¬
ties of the works after a certain lapse of time, as in the case of
delta-forming rivers ; but the principles on which the works at
the mouths of tidal rivers should be designed are more com¬
plex than in tideless silt-bearing rivers, on account of the
variability in the physical conditions, due to differences in
tidal influence.
The height and size of the bar are proportionate to the
amount of drift brought along the coast, and to the force and
duration of the prevailing winds and littoral currents which
occasion the accumulation of a portion of this drift across the
mouth of a river. An exposed site, and deep water near the
shore favour the formation of the submarine beach, by in¬
creasing the force of the waves ; but a sheltered position and
outlying sandbanks, whilst promoting deposit, at the same
time reduce the transporting power of the waves. A large
tidal ebb and flow in a river, and a large fresh-water discharge
serve, on the contrary, to lower the bar, and to drive the drift
into deep water. When the influence of these latter favour¬
able conditions is small, and the force of waves beating
obliquely on the coast and of littoral currents is considerable,
the natural mouth of a river is sometimes closed up by drift ;
and the river is obliged to find an outlet at another part of the
shore. Instances of this occurrence are afforded by the river
Yare on the Norfolk coast, the river Adur on the Sussex
coast, and the river Adour on the south-west coast of France
near Bayonne.
ix.] River Mouths improved by Jetties. 205
Jetties at Outlets of Tideless Rivers free
from Silt.
In the absence of any tidal rise, the only means available
for lowering the bar at the mouth of a river consist in the
scour of the current of the river, directed by jetties across the
bar, or in dredging a channel through the bar. The improve¬
ment by jetties, under favourable conditions, is fairly per¬
manent, for it extends the scouring power of the discharge
of the river across the foreshore into deep water, and thus
forms a channel through the submarine beach, and prevents
its filling up again with drift under the action of the sea. The
system, indeed, is similar to that adopted at delta outlets,
with the advantage that, in the absence of alluvium discharged
by the river, the bar does not gradually form again further
out. The deepening of the outlet channel by dredging is not
so permanent as that produced by the regulated scour of the
river ; but it is simple, and can be gradually extended as the
needs of any ports on the river, or the increase of trade may
render expedient.
Swinemtinde Jetties. The Swine mouth of the Oder, tra¬
versing the neck of land separating Stettiner Haff from the
Baltic Sea, forms the navigable outlet of the Oder, and the
entrance to the port of Swinemtinde. This mouth was formerly
obstructed by a bar, over which the available depth did not
exceed 7 feet. In 1818-28 jetties were constructed curving
round to a northerly direction, and prolonging the channel
for about a mile across the foreshore, in the line followed
by the issuing current of the river (Fig. 13, p. 206). These
works, by concentrating the scour across the bar, deepened
the channel over it to 28 feet ; but this depth was eventually
reduced, owing to the drift brought into the channel from the
west, round the shorter west pier, by the prevailing north¬
westerly winds and the littoral current flowing from west to
east. Dredging has been resorted to for maintaining the
2o6 Jcttics at Sivtfic Mouth of the Odev. [chap.
depth, which, moreover, is favoured by the concave form of
the east jetty towards the channel, so that in spite of the
exposure of the outer part of the jetty channel to the drift
from the west, owing to the shortness of the west jetty, the
channel at the present time has an available depth of 22-| feet.
* SCALE '¡73,000.
Fig- 13-
Jetties at Swine Mouth of River Oder.
In addition to the dredging, the east jetty has been prolonged
about 500 feet, to extend its guiding influence on the navi¬
gable chaiinel which keeps close alongside it ; whereas the
channel between the jetties shoals from 18 feet in mid-channel,
to under 6 feet against the west jetty. In this case, it is only the
peculiar concave form of the east jetty, following the natural
course of the outlet, which has enabled a prolongation of the
ix.] Jetties at Months of Vindau and Pernow. 207
western jetty, to correspond with the east jetty, to be dispensed
with.
Bussian Biver Outlets into the Baltic. The entrances to
some Russian ports near the mouths of rivers flowing into the
Baltic Sea, have also been improved by jetties, in prolongation
Fig. 14.
Jetties at Mouth of River Pernow.
of the banks of the river, concentrating the current across the
bar. Thus the mouth of the river Vindau, giving access to
the port of Vindau, has been deepened to 18 feet, by carrying
out jetties on each side into the sea, the southern jetty in pro¬
longation of the concave bank of the river having been extended
2o8 Outlet Jetties in Russia and America, [chap.
beyond the northern jetty, as in the case of the Swine ; though,
in this instance, the southern jetty not merely guides the issu¬
ing current which follows the concave shore, but also protects
the channel from the prevalent south-westerly winds and
the main drift along the coast h
Slightly converging jetties, terminating at the same distance
out, have been placed at the mouth of the river Dwina which
forms the approach to Riga, by which a central channel of
about 2i feet has been obtained.
Jetties have been carried out about miles from the shore
at the mouth of the river Pernow (Fig. 14, p. 207), which have
secured a minimum depth of 13Í feet in the jetty channel,
falling in one place to 12 feet a short distance beyond
the jetties. The available depth, however, in the river up
to the port of Pernow does not exceed 13! feet, and the
general depth of the sea near the extremities of the jetties
amounts only to between 13 and 15 feet, so that it would be
impossible to provide a channel more than from x to 2 feet
deeper than that obtained by the jetties, except by very exten¬
sive dredging operations carried out to deep water.
River Outlets into North American Lakes. The outlets of
the river Chicago into Lake Michigan, the river Buffalo into
Lake Erie, the river Oswego into Lake Ontario, and of other
rivers flowing into these inland lakes, have been trained and
deepened by parallel cribwork jetties filled with rubble stone,
the increase in depth being aided by dredging. The windward
jetty has generally been made to overlap the other, in order to
keep the drift along the coast from being driven into the outlet
channel; and sometimes the foreshore has advanced beyond
the jetties, necessitating their extension. Some years ago,
however, cribwork breakwaters were commenced at the mouths
of the above-mentioned rivers, enclosing a considerable space
to afford increased accommodation for vessels, which also serve
1 ' Atlas des Ports de Commerce de la Russie,' part I.
ix.] Converging Breakwaters at River Mouths. 209
to shelter the outlet of the river, and place the entrance to the
port in deep water h
Converging Breakwaters at the Mouths of
Tidal Rivers.
The convergence of jetties, to increase the scour, is not admis¬
sible in the case of tidal rivers ; for by narrowing the outlet,
the influx of the flood tide would be checked, and the scour¬
ing power of the ebb correspondingly reduced. Converging
breakwaters, however, starting from the shore at some distance
apart, and leaving a sufficiently wide opening at their extremi¬
ties to allow of the free admission of the flood tide into a river,
have been adopted with advantage ; for the area enclosed by
these breakwaters forms a kind of sluicing basin which, being
filled by each flood tide, produces a tidal scour on the ebb
through the narrowed outlet between the pierheads ; and the
depth is thereby increased, as exemplified by the works at the
mouth of the river Liffey (Plate 6, Figs. 1 and 2). Generally,
the converging breakwaters serve also to shelter the outlet of
the river, and to protect it from the inroad of drift brought
along the coast by waves and littoral currents, as, for instance,
the breakwaters at the mouth of the river Tees (Plate 8, Fig. 4).
Sometimes, moreover, the breakwaters enclose a large enough
area of sufficient depth to provide a sheltered approach to the
river, and a harbour of refuge, in which dredgers can easily
remove the bar and deepen the navigable channel, of which
the works at the mouth of the river Nervion, and of the river
Tyne, afford examples (Plate 6, Fig. 12, and Plate 8, Fig. 1).
Breakwaters at Mouth of Hiver Liffey. The river Liffey,
constituting the entrance to the port of Dublin, flows into
Dublin Bay across a sandy foreshore, about miles wide,
formed by the gradual accumulation of sand brought in from
the sea, which dries at low water. Formerly the outlet channel
1 ' Harbours and Docks.' l. F. Vernon-Harcourt, plate 5, figs. 5, 11, and 12.
p
2io Converging Breakwaters, Month of Liffey. [chap.
was constantly shifting, and was obstructed by a bar in front
of its mouth1. In 1711 the improvement of the river below
Dublin was commenced, by rectifying its channel, and lining
it with quays ; and during the progress of these works, the
Great South Wall was begun (Plate 6, Fig. 1), with the object
of straightening the outlet channel, and protecting it from
south-westerly gales and the inroad of sand from the southern
strand. This breakwater, 3! miles long, was only completed
in 1796 ; but though it fulfilled the purposes for which it was
designed, and also somewhat deepened the channel at the
mouth of the river by the tidal scour produced round its head,
it had no effect in lowering the bar beyond. Accordingly,
a second breakwater, called the Great North Wall, formed of
rubble stone, and converging from the shore towards the
south breakwater, was carried out in 1820-25 (Plate 6, Figs. 1
and 3). This breakwater, if miles long, was only raised to
half-tide level along its outer portion, which facilitates the
influx of the flood tide into the enclosed area, and prevents
the occurrence of rapid currents through the contracted open¬
ing between the heads of the breakwaters, 1,000 feet in width,
near high water, which would be inconvenient for navigation.
The tidal scour, moreover, is concentrated between the pier¬
heads during the latter half of the ebb, when most efficacious
for lowering the bar, and when, owing to the peculiar set of
the tidal currents in the bay near low water, there is no danger
of the sand removed during the latter part of the ebb being
brought back again into the sheltered area by the flood tide.
The river Liffey has a small discharge, and therefore the
outlet channel below Dublin has had to be improved by
dredging, the outer portion, across the foreshore, being
protected by the breakwaters. The channel, however, across
the bar outside has been deepened by the tidal scour pro¬
duced by the emptying of the large area enclosed by the
breakwaters, covered by water at each tide, the rise of tide in
1 Minutes of Proceedings Institution C. E., vol. lviii, p. 119.
ix.] Lowering of Bar at Mouth of Liffey. 211
Dublin Bay at springs being 13 feet. The tidal capacity of
this area has been gradually increased, owing to the washing
of sand from the north strand by the flow during the ebb into
the river channel, which is periodically removed from the
channel by the dredgers, and also by dredging ballast from
the north strand ; whilst reclamation within this area has
been strictly prohibited, in order that the tidal scour at the
outlet may be fully maintained. This scour has gradually
lowered the bar in front of the outlet, from a depth of 6 feet
in 1819 to 16 feet in 1873, below low water of spring tides,
which depth has been since maintained (Plate 6, Fig. 2) ; and
this depth has been extended up to the port by dredging1.
The improvement of 10 feet in depth effected by tidal scour
will be maintained so long as the tidal capacity of the fore¬
shore within the breakwater remains undiminished, and till
the gradual advance of the foreshore outside the breakwaters
brings low-water mark down to the ends of the breakwaters,
when an extension of these works will become necessary.
Converging Jetties at Mouth of Merrimac River. The
mouth of the river Merrimac is contracted from a wide
expanse of flats inside, to a width of 1,000 feet with a depth
of 30 feet. The scour through this opening, from the outflow
of the tidal waters covering these flats, with a rise of tide of
7| feet, maintains the depth at the outlet ; but on entering
the Atlantic Ocean, the current soon loses its efficiency, and
a bar due to wave-action existed across the outlet channel,
with a depth of only 7 feet over its crest at mean low-water.
Rubble jetties starting from the shore on each side, 4,000 feet
apart, and converging to a width of 1,000 feet over the bar,
were commenced in 1881. These jetties are designed to be
4,000 and 2,400 feet long respectively. By the middle of
1892, the north jetty had been carried out 2,485 feet, and the
south jetty 1,077 feet, their extremities being only about
1 Mr. J. P. Griffith furnished me with a recent chart of Dublin Bay, and other
particulars.
P 2
212 Works at Mouths of Merrimac and Tees. [chap.
i ,300 feet apart ; so that little more than the parallel outer
portions, 1,000 feet long, remained to be constructed1. The
concentrated scour across the bar, produced by these works,
has already increased the depth over the bar to 13 feet at
mean low water ; and it is expected that, on the completion
of the works, a channel across the bar, 17 feet deep, will be
obtained by the tidal scour. In this case, such extensive
works as those carried out at the mouth of the Liffey were
not required, for a natural sluicing basin already existed
inside the mouth, supplying the tidal reservoir for scour,
which had to be created artificially in Dublin Bay.
Breakwaters at Mouth of River Tees. The South and
North Gare breakwaters converging together from the oppo¬
site shores at the outer extremity of the wide Tees estuary,
shelter the estuary and the training walls, protect the en¬
closed area and the trained channel from sand brought
along the coast, and when completed will aid in maintaining
the outlet channel across the bar (Plate 8, Fig. 4). The
South Gare breakwater, commenced in 1863, formed of
a mound of the refuse slag from the neighbouring ironworks,
protected on its sea face and round its head by a concrete
wall, has been completed2; it stretches across the Bran Sand
to the bar, and has a length of nearly i\ miles (Plate 8,
Fig. 6). The North Gare breakwater is in progress, also
formed of a mound of slag, which is to be carried out for
a length of rather over a mile, across the sands to the bar,
leaving an opening between its extremity and the end of the
other breakwater of 2.400 feet. The rise of spring tides at the
mouth of the Tees is 15 feet, so that, on the completion of
the North Gare breakwater, there will be a considerable flow
between the pierheads at certain states of the tide, owing to
1 ' Annual Report of the Chief of Engineers, U.S.A., for the year 1892,' part 1,
p. 551; and atlas, plate 2.
2 ' River Tees Improvements.' John Fowler : Minutes of Proceedings Institution
C. E., 1887, vol. xc, p. 344.
ix.] Effect of Breakwaters at Mouth of Tees. 213
the wide tidal area enclosed. This current will serve to
maintain the depth at the outlet ; but if sand is introduced
by the flood tide during storms, and if reclamations are con¬
tinued inside the estuary, the tidal capacity will be reduced,
and the scour at the outlet correspondingly diminished.
The increase in depth of about 10 feet since 1852, already
obtained over the bar at the mouth of the Tees, has been
effected by the scour resulting from the training works
through the estuary, and dredging, aided of late years by
the guidance and protection of the South Gare breakwater
(Plate 8, Fig. 5). The converging breakwaters will always
shelter the trained channel, will protect the channel and
estuary from the inroad of drifting sand, and, by placing the
outlet in comparatively deep water, will reduce considerably
the amount of sand liable to be brought into the sheltered
area by waves, where it would readily deposit ; but the
efficiency of the breakwaters in aiding the maintenance of the
outlet channel by scour, will depend on the relinquishment of
further reclamations in the estuary, and the absence of deposit
in the enclosed tidal area.
Breakwaters at Mouth of River Tyne. The breakwaters
starting from the sea-shore on each side of the mouth of the
river Tyne, have been extended out beyond the bar into
a depth of 5 fathoms at low water, and converge so as to
leave an opening of 1,300 feet between the pierheads1
(Plate 8, Figs. 1 and 3). The rise of spring tides at the
mouth of the river is 14I feet ; and the harbour entrance has
been left adequately wide to admit the flood tide freely up
the deepened river. The increased depth over the bar, of
about 18 feet (Plate 8, Fig. 2), has not in this case been
produced by any scouring effect of the converging break¬
waters, but is due to dredging and the improved tidal scour
resulting from the great deepening of the river above. The
breakwaters, however, have enabled the bar to be dredged
1 1 River Tyne Improvement.' P. J. Messent, 1888.
214 Tynemouth Harbour; River Wear Jetties. [CHAP.
away under the shelter they have afforded ; and they have
also prevented the drift brought along the coast by- the
prevalent north-easterly winds, and by the flood tide, from
encumbering again the mouth of the river, and have stopped
the waves from re-forming the bar across the outlet. The
north and south piers, as the breakwaters are called, were
commenced in 1856, and, being extended as funds permitted,
have only been quite recently completed, with total lengths
of about 3,200 and 5,400 feet respectively. As the river flows
directly into the North Sea without any intervening estuary,
and the bar was only a short distance beyond the mouth of
the river, smaller works would have sufficed for merely
improving the outlet channel; but shortly after the commence¬
ment of the piers, it was decided to extend them sufficiently
to form a harbour of refuge, in view of the unprotected state
of the east coast of Great Britain between the H umber and
the Firth of Forth.
Breakwaters at Mouth of River Wear. Till the south
entrance to the Sunderland docks was built, the river Wear
afforded the only approach to the port. During the eighteenth
century, jetties were built on each side of the mouth of the
river, to direct the out-flowing current across the bar which
was raised by drift from the north during north-easterly
storms, reaching nearly up to low-water spring tides1 (Plate
6, Figs. 6 and 7). Though the drift naturally produced an
advance of the foreshore against the northern jetty, the outlet
channel was deepened to 4 feet at low water by the scour
over the bar occasioned by the trained outlet, aided by
dredging and the removal of shoals ; and the jetties, which
had originally been built of timber, were rebuilt with stone
about 1843. The rise of tide at the mouth of the Wear is
14! feet at springs ; and the scour at the outlet is aided by
tidal recesses which have been retained near the mouth, and
by the discharge from a hilly drainage-area of 456 square
1 Minutes of Proceedings Institution C. E., vol. vi, p. 256.
ix.] Breakwaters at Mouths of Wear and Nervion. 215
miles. Nevertheless, owing to the tendency of waves on that
exposed coast to obstruct the mouth of the river with drift,
the minimum depth in the outlet channel remained at about
4 feet below low water up to 1888. Since that time, however,
dredging has been actively carried on, increasing the available
depth to about 11 feet at low tide in 189a (Plate 6, Fig. 7).
Two converging breakwaters, moreover, are now in course
of construction for sheltering the approach to the river ; and
the northern breakwater, which is approaching completion,
will effectually prevent the drift from reaching the outlet of
the river1; and it will be possible to dredge the outlet
channel to any requisite depth within the sheltered area
(Plate 6, Figs. 6, 8, and 9).
Breakwaters at Mouth of River Nervion. The break¬
waters in progress across the small inlet on the southern
shore of the Bay of Biscay, into which the river Nervion
flows, are intended to form a harbour of refuge on that
stormy coast, and to protect the entrance of the river which
affords access to the port of Bilbao2 (Plate 6, F"igs. 12
and 15). They have been designed to extend out into depths
of from 40 to 46 feet at low water, and will be considerably
seawards of the site of the bar which formerly encumbered the
mouth of the river, but has been removed by the training
of the river (Plate 6, Fig. 13). These breakwaters, however,
by protecting the beach near the mouth of the river from the
violent waves which beat upon that coast, will prevent the
bar from forming again, and will also protect the outlet from
the easterly drift produced by north-westerly storms.
Remarks on Converging Breakwaters at River Mouths.
Converging breakwaters are necessarily larger works than
1 A plan and sections of the new breakwaters, and a longitudinal section of the
present outlet channel, were sent me by Mr. H. H. Wake, engineer to the River
Wear Commission
2 ' Memoria que manifiesta el Estado y Progreso de las Obras de Mejora de
la Ria de Bilbao, 1885-1889.' Evaristo de Churruca.
2x6 Effects of Breakwaters at Months of Rivers, [chap.
training jetties in prolongation of the river channel, and they
become very costly when extended into deep water in exposed
situations. This extension, though ensuring a deep outlet,
and enabling any desired increase in depth to be provided by
dredging within the sheltered area, is not expedient except at
the approach to an important seaport, and where a refuge har¬
bour is much needed, as at Tynemouth and the mouth of the
Nervion. Breakwaters, however, merely converging towards
the bar, may be essential for securing and maintaining a good
outlet channel, especially where, as at the mouth of the
Liffey, the tidal capacity of the river and the fresh-water
discharge are small ; but the works, though carried out
cheaply in Dublin Bay across a shallow foreshore, extended
over a long period. The long breakwaters sheltering the
Tees estuary were only rendered practicable by the refuse
slag composing them being deposited free of cost.
The permanence of the tidal scour obtained by converging
breakwaters depends on the absence of sediment in the
inflowing flood tide, or in the maintenance of the tidal
capacity of the area enclosed, by the periodical removal of
any deposit that may take place. Owing to the great cost
of large breakwaters, the system is only expedient where
a small tidal capacity in the river, combined with a small
fresh-water discharge, or the exposure of the mouth to heavy
seas, renders training jetties inadequate for providing the
requisite depth across the bar.
Training Jetties at the Mouths of Tidal Rivers.
Tidal rivers may be trained, like non-tidal rivers, by cross
jetties or longitudinal training banks, so as to obtain uniformity
in depth by regulating the width of the channel (see Chapter
III). The width, however, of the trained channel in tidal
rivers must be gradually enlarged as it descends towards the
sea to ensure the free admission of the flood tide, the rate
Training Jetties at Mouths of Rivers. 217
of enlargement being increased on approaching the mouth
of the river. Thus the first improvement works on the Clyde
consisted in the regulation of the river below Glasgow, first
by cross jetties, and subsequently by longitudinal training
walls, aided by the removal of hard shoals (Plate 8, Fig. 7).
The Tees has been regulated in a similar manner, and also
by the formation of straight cuts between Stockton and
Middlesbrough (Plate 8, Fig. 4) ; and the Tyne has been
improved by some straight cuts above Newcastle, by easing
some bends below, and by shutting off some wide recesses
near Jarrow (Plate 8, Fig. 1). The Scheur branch of the
Maas, moreover, has been systematically trained by longitu¬
dinal banks of fascine mattresses, where requisite for regulating
the width, from Rotterdam to its outlet in the North Sea
(Plate 6, Fig. 18) ; and the river Nervion has been similarly
regulated by rubble training walls and quays, between Bilbao
and its mouth (Plate 6, Fig. 12). The mouth of a river,
however, when opening directly on to the sea-coast, and
exposed to the obstruction of sand or shingle drifting along
the shore, generally presents the most difficult problem in
any scheme of improvement. Sometimes the action of the
waves and currents upon the drift is so powerful that the out¬
let frequently changes its position, until it is fixed by jetties ;
and often works are required at the outlet to render a river
accessible for vessels, or to enable ships to take advantage
of the improvement in depth obtained by regulation works
above. Jetties training the outlet channel across the bar are
simpler and more economical than the converging break¬
waters already referred to ; and they often suffice, either to
fix a wandering outlet, or to form a prolongation of the
training works into deep water.
The jetties at the mouths of the Yare, the Adur, and the
Adour, furnish examples of the fixing of shifting outlets
(Plate 6, Figs. 4 and n o) ; whilst the jetties at the mouths of
the Maas and the Nervion are instances of the prolongation
2i8 Changes of Month of Yare, and Jetties. [chap.
of training works along rivers out to deep water, beyond the
foreshore at the outlet (Plate 6, Figs. 12 and 16).
Jetties at Mouth of River Yare. Yarmouth Haven is
now formed by the junction of the three rivers, the Yare,
the Waveney, and the Bure, into a single channel above Great
Yarmouth, which after running past the town, nearly parallel
to the coast, for about 2f miles in a southerly direction,
makes a sharp bend, and flows straight into the North Sea
(Plate 6, Figs. 4 and 5). In Roman times, however, the
river Bure appears to have had an independent outlet about
4 miles north of the town ; whilst the mouth of the Yare was
only a mile south of the town. The Bure was eventually
forced to join the Yare by the blocking up of its outlet by
sand and shingle. Later on, however, the mouth of the Yare
was driven about 4 miles further south by the drift occasioned
by north-east storms; and though a new channel was cut
on several occasions between the fourteenth and sixteenth
centuries, to bring the mouth of the river nearer Great Yar¬
mouth, the new outlet was always sooner or later blocked up.
At last, in 1559, a cut was made at the site of the present
outlet, and was fixed by two nearly parallel jetties across
the foreshore, about 200 feet apart, which has since been
maintained. The northern jetty has been extended in
advance of the other, and has been made solid to protect
the mouth from the drift coming along the coast from the
north. The jetties maintain the outlet channel over the bar,
which the drift tends to form in front of the mouth, by
directing the fresh-water discharge and the tidal ebb across
the foreshore. The rise of spring tides at that part of the
coast is only 6 feet ; but owing to the flatness of the country
through which the rivers flow, the tidal influence extends
many miles inland ; and the tidal scour is promoted by the
wide expanse of Breydon Water, into which the Yare and the
Waveney flow above the town. Dredging also is employed
to increase the tidal capacity of the river, and to deepen the
Changes of Month of A dur, and Jetties. 219
channel. The minimum central depth in the jetty channel,
at low-water spring tides, is now jo feet, as compared with
7 feet in 1881 ; the minimum available depth over the bar
is about 9 feet ; and the available depth in the river up to
the port is about the same, so that in calm weather the bar
offers little impediment to the entrance of vessels.
Jetties at Mouth of River Adur. When the advance
of the shore cut Old Shoreham off from the sea, the port
was shifted to New Shoreham, situated on the river Adur
which gives it access to the sea. The mouth, however, of
the river was liable to be driven towards the east by the
eastward drift of shingle along the coast; and in 1750 the
mouth was 3^ miles to the east of New Shoreham. Ten
years later a new outlet was cut across the beach, a little
more than a mile to the east of the town, and was protected
at its extremity by two wooden piers ; but when these piers
were destroyed, the outlet travelled eastwards again, and at
the close of the eighteenth century it was 3} miles east of
the town. In 1819 the present entrance was cut through the
beach, in almost the same position as the previous cut, and
was again protected by more durable jetties, placing the
entrance once more nearer New Shoreham.
As compared with the Yare, the Adur outlet possesses the
advantages of a rise of tide of about 19 feet, and a somewhat
more sheltered position ; but, on the other hand, its drainage
area is only 160 square miles, whilst that of the Yare is 1,228
square miles, the tidal flow up the Adur is much more restricted,
and the foreshore on that part of the coast tends to advance.
Jetties at Mouth of River Adour. The mouth of the
river Adour being situated near the south-eastern angle of
the Bay of Biscay, where deep water approaches the coast,
is exposed to the full force of the surf which beats upon
that shore during north-westerly gales ; and formerly its
position was constantly shifted by the inroad of drift. Though
the travel of the sand and shingle is mainly from north to
22o Changes of A dour Outlet; Works at Mouth. [CHAP.
south, owing to the prevalence of north-westerly winds, the
mouth of the river, in 1450, had been driven 18 miles north
of its present outlet, which deviation, by obstructing its
outflow, threatened Bayonne with inundation during floods.
Accordingly, in 1579, a straight cut was formed to the
sea, and the old channel was barred, so that the river was
diverted into this new channel ; and the depth over the bar
at the outlet was for a time improved. The drift from the
north, however, soon drove the channel towards the south,
and reduced its depth, till at last masonry jetties were con¬
structed in 1733-41, which fixed the outlet in its present
position. The depth obtained at first by these works over
the bar was 20 feet at high water ; but the bar soon rose
again further out. This result being attributed to the width
between the jetties, of 900 feet, being too great, it was reduced
eventually by low inner jetties to 600 feet ; but the depth
over the bar was not thereby increased ; and the outlet channel
shifted in position, exhibiting a tendency to travel south
again. Between 1808 and 1838, accordingly, the jetties
were prolonged, mainly on the south side, which prevented
the deviation of the channel to the south ; and the width at
the outlet was reduced to 500 feet. These works, however,
whilst securing the stability of the outlet, only pushed the
bar further out, without reducing its height1. Prolongations
of the south jetty, with 656 feet of solid work and 656 feet
of open timber-work, on a rubble mound raised only to 6|
feet below the lowest low-water, and of the north jetty with
2,180 feet of open timber-work, on a low rubble mound, were
carried out in 1857-61. This open work was adopted for
the extensions to avoid a progression of the foreshore. The
timber-work having been attacked by the sea-worm, has been
replaced by iron cylinders, 65 feet in diameter, with open
spaces, jo feet wide, between them2; and the southern jetty
1 ' Ports Maritimes de la France,' vol. vi, p. 931.
2 Minutes of Proceedings Institution C. E., vol. Ixx, plate 3, fig. 5.
ix.] Results and Deficiencies of A dour Jetties. 221
is being prolonged, having been extended 200 feet since 1881
(Plate 6, Fig. 10). The intervals between the cylinders were
intended to be temporarily closed by sliding panels whenever
it might be expedient to concentrate the outgoing current ;
but this design has been abandoned, owing to difficulties in
working, and its cost.
These extensions of the jetties have increased the depth
over the bar from a minimum of I5f feet *n t^57> to 17 feet
in 1892, at high water of spring tides, the rise of tide at
springs being iof feet (Plate 6, Fig. xi). The improvement
in depth is, accordingly, very moderate as compared with the
extent of the works ; and both the foreshore and the bar
have advanced seawards since the extension of the jetties;
so that the stability of the mouth of the river is the main
benefit realised by the jetties.
As the Adour drains only a small basin, and therefore has
a small discharge except during floods, whilst the rise of tide
at its mouth is fairly good, the narrowing of the channel,
from about 1,000 feet near Bayonne to 550 feet at its outlet,
was a mistake, considering that the maintenance of the
outlet depends largely upon the tidal volume of the ebb, which
has been seriously reduced by the contraction of the jetty
channel, for the tidal range is decidedly less in the river than
outside. The formation of an enlarging channel by diverging
jetties, carried out beyond the bar, would have ensured the free
admission of the flood tide ; and the bar could have been
lowered within the jetty channel by dredging. Solid jetties
would have produced an advance of the foreshore ; but
probably the line of foreshore would have reached an equili¬
brium before compromising the outlet, with jetties extending
beyond the bar. If adequate funds had been available,
a still better outlet channel could have been obtained, on
such a very exposed coast, by converging breakwaters ex¬
tending into deep water, and dredging the outlet channel
within the area thus sheltered.
222 Bar, Cut, and Jetties at Month of Maas. [chap.
Jetties at Mouth of River Maas. The general tendency of
the numerous channels of the Maas, which intersect South
Holland, is for the northern outlets to silt up, and the
southern outlets to become the only navigable channels,
owing partly to the large reclamations which have been carried
out along the northern channels, and partly to the larger rise
of tide to the south, which favours the maintenance of the
southern outlets. Accordingly, the growth of the bar at the
mouth of the most northern Scheur branch, which is the most
direct channel to Rotterdam, and at the mouth of the adjacent
•
Brielle branch, obliged sea-going vessels trading with Rotter¬
dam to follow more southern, and very circuitous channels.
A canal was cut in 1829 through the island of Voorne, to
enable vessels entering by the deeper Hellevoetsluis entrance
to reach Rotterdam by the more direct Brielle branch, and
thereby avoid the Brielle bar. In course of time, however,
the increasing size and draught of vessels rendered the canal
inadequate for the requirements of navigation ; and conse¬
quently it was decided in 1863 to form a direct channel
between Rotterdam and the sea, about 20 miles in length,
with a minimum depth of 21 i feet at low water, by the
improvement of the Scheur branch1.
The works which were carried out on the Scheur branch of
the Maas between 1863 and 1876, comprised the training
of the river in a gradually enlarging channel from Rotterdam
to the Hook of Holland ; the formation of a direct outlet into
the North Sea across this neck of land, 3 miles in width ; the
closing of the old circuitous outlet round the Hook ; and
the training of the new outlet into deep water by slightly
diverging fascine-mattress jetties, 6.560 and 7)55° feet l°ng
respectively, across the sandy foreshore (Plate 6, Figs. 16
and 19). The outer end of the north jetty was only raised
to half-tide level ; but the south jetty was raised to ordinary
1 ( Etude sur l'Effet des Marées dans la partie maritime des Fleuves.' P. Caland,
1861.
ix ] Regulation and Dredging of Scheur Branch. 223
high-water level, in order that the flood tide coming from the
south might enter the river straight up channel, and that the
channel might be protected from the littoral drift from the
south. The trained channel was made 738 feet wide at
Krimpen above Rotterdam, increasing to 1,476 feet at Vlaar-
dingen, and attaining a width of 2,953 ^eet at t'ie extremities
of the jetties. The new cut, three miles in length, was only
partially excavated previously to the diversion of the river into
it ; and the widening and deepening of this channel was left to
the scour of the fresh-water and tidal currents. The current,
however, whilst deepening the narrow channel down to 40 feet
below low water in some places, did not enlarge it much
in width ; and a great portion of the sand scoured out of the
cut was deposited in the jetty channel, and arrested further
scour. Accordingly, the minimum depth along the new
waterway was only 12 feet at low water in 1877, after the
completion of the works, the rise of tide at the mouth being
only 6\ feet at springs, so that vessels drawing more than
18 feet could not get up to Rotterdam, only a slight im¬
provement on the available depth of i6f feet by the Voorne
Canal. Supplementary works were consequently commenced
in 1882, for securing the required depth, which have now been
completed.
The new works have been designed to remedy the defects
in the original scheme, by enlarging the new cut to the
requisite width, by excavation and dredging, to correspond
with the channel above and below ; removing the accumula¬
tions of sand in the jetty channel ; reducing somewhat the
enlargement of the trained channel ; and preventing the loss of
tidal water on the ebb, by checking its outflow into the
Noordgeul channel leading to the Brielle outlet h The trained
channel has been reduced in width to 985 feet a little below
Rotterdam, increasing fairly uniformly towards the sea, and
1 ' Amélioration de la Voie Fluviale de Rotterdam à la Mer.' J. W. Welcker,
\me Congrès International de Navigation Intérieure, Paris, 1892.
224 Results of New Works in Scheur Branch, [chap.
through- the enlarged cut, to 2,300 feet at the extremity of
the jetty channel, which has been narrowed from its original
width by. a submerged fascine-mattress jetty, constructed
about 650 feet to the north of the south jetty in continuation
of the southern trained bank of the enlarged cut (Plate 6,
Fig. 16). Extensive dredging works have also been carried
out in the jetty channel, to remove the deposits and deepen
the outlet channel, and also beyond the ends of the jetties to
extend the navigable channel to deep water. As the Noord-
geul channel, branching off from the Scheur branch nearly
opposite Vlaardingen, could not be entirely closed, owing
to its affording access to the Voorne Canal, its entrance width
has been reduced from 1,066 feet to 230 feet, so that very
little of the ebb current is now diverted from the Scheur
branch. These works have increased the minimum navigable
depth at the outlet to about 24Í feet at low water (Plate 6,
Fig- I7) i and whereas, in 1882, vessels could only reach
Rotterdam with a maximum draught of 195 feet, in 1888
vessels drawing 25 feet could effect the passage. Before
any works were commenced, the circuitous journey between
Rotterdam and the sea often occupied 18 hours; whereas
now a vessel can traverse the 20 miles of channel between the
sea and Rotterdam in 2 hours. The great improvement in
the navigable access to Rotterdam, both in directness and
depth, has transformed it as a seaport; and the increase in
the navigable depth achieved by the recent supplementary
works, has resulted in the trebling of the tonnage of the
vessels trading with Rotterdam since 1879. A total sum
of £2,965,000 has been expended, within .the last thirty
years, in the formation of the new channel.
Jetties at Mouth of River Nervion. The works of regula¬
tion and training which have been carried out, from 1873
to 1891, on the river Nervion between Bilbao and the mouth
of the river on the south coast of the Bay of Biscay, a distance
of nearly 8 miles, were designed to provide the developing
ix.] Training and Jetties of River Nervion. 225
port of Bilbao with a better navigable access to the sea1.
The works, which have formed a channel enlarging in width
from 210 feet at Bilbao to 525 feet at the extremity of the
north-eastern jetty, resemble in principle the training works
along the Scheur branch of the Maas ; but though on a much
smaller scale, on account of the great difference in the width
and length of the waterways, the training walls and jetty
extension have been constructed with rubble-stone, concrete,
and masonry, owing to the abundance of suitable materials
in the neighbourhood (Plate 6, Figs. 12 and 14). A new cut
was formed for the river, in place of a very tortuous channel,
just above the confluence of the river Cadagua ; the channel
has been rectified at other bends by rubble training walls and
quays ; and excessive widths have been reduced by similar
works. As the river only drains a small basin, its fresh¬
water discharge is small ; but the tidal rise at its mouth
amounts to iqf feet at equinoctial springs, and is 9 feet at
average tides.
The bar in front of the outlet, formed by littoral drift from
the west under the influence of the prevalent north-westerly
winds, and by wave-action on that exposed coast, has been
lowered by the scour produced by the prolongation of the old
Portugalete jetty, on the south-west side of the outlet, in
a slightly concave line towards the channel, for a length
of about 2,640 feet, extending out beyond the bar into a depth
of 3 fathoms at the lowest low-water (Plate 6, Figs. 12
and 13). The jetty resembles a breakwater at its outer end
(Plate 6, Fig. 14); but for the inner 1,870 feet, it is only solid
up to high-water spring tides, and is surmounted by an open
iron viaduct affording access between the old jetty and the
new pierhead. The inclined direction of approach of the
river to its outlet, causes the main outflowing current to keep
1 ' Memoria que manifiesta el Estado y Progreso de las Obras de Mejora de la
Ria de Bilbao,' 1879-1891 ; E. de Churruca; a.,d 'Anales de Obras Publicas,'
vol. xi, 1883.
Q
226 Increase in Depth of River Nervion. [chap.
close along the south-western jetty, which is further assisted
by the concave form of the jetty, so a prolongation of the old
jetty on the opposite side of the channel was unnecessary.
The scour and deepening produced by the extension of the
jetty have been aided by dredging in the river. The improve¬
ment in the channel produced by the jetty, the training and
regulation works, and dredging, is indicated by comparing
the longitudinal sections of 1878 and 1889 (Plate 6, Fig. 13);
for, with the exception of the river diversion which was com¬
menced in 1873, the works were only begun in 1878. The
minimum depth in the navigable channel of the river up to
Bilbao, at the lowest low-water, has been increased from 5 feet
in 1878, to about 13 feet in 1889. The depth over the crest
of the bar has been increased within the same period from
3f feet to 15! feet at the lowest tides ; and though its crest,
which has been pushed out seawards as far as the end of
the new jetty, is still apparent, it presents no impediment to
navigation, for it is about 3 feet lower than the bed of the
river a little higher up. The exposure of the outlet, with
deep water approaching the shore, and the liability of the
mouth of the river to be obstructed by drift during storms,
will be remedied by the shelter which the outlying break¬
waters will eventually afford (Plate 6, Fig. 12). The expen¬
diture on the training works and dredging since 1878 has
amounted to ,£408,000.
Remarks on "Works at River Outlets. With the exception
of the Maas, the bars at the outlets of the rivers referred to in
this chapter are of marine origin, and have been lowered by
the scour across them, resulting from the works carried out at
their mouths, aided by dredging. The scour, moreover, in
the case of tidal rivers, has been augmented by the larger
admission of tidal water up the regulated river, and the freer
discharge of the ebb, or by the enclosure of a large area of
tide-covered foreshore by converging breakwaters. The
assistance of dredging is very useful in providing a greater
ix.] Materials used for Jetties at Outlets. 227
depth than could have been attained by scour alone, and
in maintaining the tidal capacity of a river when liable to
be reduced by the deposit of silt ; and frequently scour is
able to maintain a depth which it could not have created
unaided. Though the Maas brings down from inland most
of the material which encumbers its outlets, the tendency
of the sea to encroach along that coast prevents a progression
of the foreshore ; and the alluvium is carried northward by
the flood tide.
The construction of the jetties depends upon the available
materials. Where stone is easily procured, as at Tynemouth,
Dublin Bay, and the mouth of the Nervion, the jetties are
naturally formed with it. Where, however, as at the mouth of
the Maas, and at the delta of the Mississippi, there is no stone
in the vicinity, but osiers are cultivated in abundance, fascine
work constitutes the most suitable method of construction,
especially on a soft sandy or silty foundation (Plate 6,
Fig. 19). The employment of fascine work for structures
in the sea, first adopted at the mouth of the Maas, might
have appeared rather too bold an expedient ; but it has been
justified by experience, for the fragile material is soon pro¬
tected from decay by a coating of sand or silt, whilst the
yielding nature of the fascine work enables it to withstand
the shocks of the waves without injury. Fascine mattresses
have since been resorted to in the more extensive jetties
for improving the entrances to Charleston and Galveston
harbours *, as well as for the South Pass jetties at the mouth
of the Mississippi (Plate 5, Fig. 8).
The mouths of rivers are not generally altered in position,
except when a shifting outlet has to be fixed by jetties ; but
in the case of the Maas, it was possible to obtain a more
direct outlet, and better depths inshore than at the natural
1 'Harbours and Docks.' L. F. Vernon-Harcourt, plate 2, figs. I, 2, £2,
and 13.
Q 2
228 Value of Jetties at Mouths of Rivers.
outlet, by a cut of moderate length across a low neck of
land.
The great advantages for navigation that may be secured
by jetty works at the mouth of a river, have been amply
demonstrated by the instances described, especially if supple¬
mented by training works and dredging carried up to the
port for which improved access is required.
CHAPTER X.
TIDAL FLOW IN RIVERS ; AND FORMS OF
ESTUARIES.
Tidal Flow in a Creek ; its effects. Tidal Flow and Fresh-water Discharge in
a River. Alluvium in a Tidal River ; its origin ; instances. Influence of Fresh¬
water Discharge ; advantages of large discharge ; examples. Influence of Tidal
Flow ; importance of tidal volume ; contrast of state of tidal and tideless rivers.
Influences of Flood and Ebb Tides; periods of greatest velocity; divergence in
directions. Conditions affecting Tidal Flow ; large tidal rise ; small fall of river ;
examples. High-water line ; general uniformity ; instances of rise ; irregularities.
Lowering of Low-water Line ; advantages ; indication of improvement. Progress
of Tide up a River ; indicated by simultaneous tidal curves ; acceleration of flood
tide. Bore, cause, instances. Obstructions to Tidal Flow, effects, disadvantages.
Effect of Tidal Flow on Outlet. Forms of Channels and Estuaries ; best
theoretical form ; varieties in forms of estuaries. General Principles relating to
Tidal Rivers.
The physical conditions of tidal rivers are more varied
than those of the non-tidal portions of rivers, or of rivers
flowing into tideless seas, owing to the differences which exist
in the tidal rise at different parts of the sea-coast, in the
distance to which the tidal influence extends up a river,
and in the tidal capacity of a river.
Influence of Tidal Flow in a Creek. When the tide flows
up a sheltered creek, into which there is no discharge of
fresh water from the land, if the flood tide enters charged
with sand or silt, the creek will gradually silt up ; for, though
the ebb tide carries out again a large portion of the material
brought in by the flood, a portion of the sediment, deposited
during slack water at the turn of the flood tide, is left behind.
The flood tide, which is impelled up a creek by the difference
in head between the water in the sea and in the creek, has its
230 Tidal Flow in a Creek, and in a River, [chap.
force diminished by the inert mass of water previously in the
creek, and by friction against the sides of the creek, which
also reduce the force when reversed on the ebb. Accordingly,
the force of the ebb in the creek is less than the force of the
flood by the sum of the retarding forces acting during both
the flood and ebb tides. The silting up is also promoted by
the sediment, when once deposited, being less easily removed
than when held in suspension.
Tidal Flow, and Fresh-water Discharge in a River. The
tidal flow in a river differs in two respects from the tidal flow
in a creek. In the first place, a river affords a much longer
distance for the run of the flood tide than a mere creek, and
therefore the tidal currents are stronger and more continuous,
presenting consequently less opportunities for the deposit
of sediment transported by the water. Secondly, the ebb
tide is reinforced by the fresh-water discharge of the river,
which is penned back during the flood tide ; so that the
whole volume of fresh water brought down by a river during
a period of rather over twelve hours, has to flow away during
the ebb. Accordingly, a tidal river is much less subject
to accretion than a simple creek. The influence, indeed, of
the fresh-water discharge of a tidal river in maintaining the
depth of the river, is manifested by the bars at the mouths of
rivers being frequently raised when the discharge is small,
and being lowered again on the descent of floods. When
the tidal influence extends some distance up a river, the
fresh-water discharge during the tidal period does not reach
the sea in a single ebb tide, but it proceeds down the river in
successive stages, and is equally efficacious in reinforcing the
ebb. The tidal oscillations of the flood and ebb maintain
the water in almost constant motion, which hinders deposit ;
whilst the fresh-water discharge counterbalances the pre¬
ponderating energy of the flood tide. A kind of equilibrium
is, accordingly, generally maintained in tidal rivers, until
modifications are introduced into the existing conditions,
x.] Tidal and Fresh-Water Flow contrasted. 231
either by engineering works, or by the slow changes produced
by the irresistible operations of nature. If a state of equili¬
brium was not attained, a river would either continue deepen¬
ing its channel till its scouring power was sufficiently reduced,
by the increase in the size of its bed, to produce no further
effect, or its channel would gradually silt up till the concen¬
trated current was able to maintain the diminished channel.
The fresh-water discharge of a river is a definite volume,
proportionate to the area of the basin and the rainfall over
that area, which varies according to the seasons, and fluctuates
from year to year with the rainfall, but approximates to
a certain average when measured during a series of years.
This volume of water has to be discharged by the river ; it
cannot be augmented ; but it can be diminished in amount,either
by abstraction for water supply, or by diverting it into a new
channel. The tidal volume, on the contrary, is practically
unlimited ; it can be increased by facilitating its influx into
a river, and by removing obstacles to its upward flow ; or
it can be reduced to any extent, by restricting the width
of the entrance to a river, and by barring its progress above.
Accordingly, a tidal river possesses capabilities for improve¬
ment or injury, which are absent in a tideless river.
Alluvium in a Tidal River. Alluvium may be introduced
into a tidal river from inland by the fresh-water discharge,
as in the case of tideless rivers ; or it may be brought in
by the flood tide, having been stirred up by waves from
shoals, or eroded from the neighbouring sea-coast. In the
former case, the river is liable to deposit its burden in the
upper tidal portion, when arrested in its flow by the flood
tide. In the latter case, the incoming flood tide tends to
deposit the heavier sediment in any sheltered places near
the mouth of the river, out of the main run of the tide, and
the lighter alluvium higher up during slack water at the
turn of the tide, particularly when the fresh-water discharge
is small. The greater part of the alluvium from inland is
232 Origin and Deposit of River Alluvium, [chap.
brought down by floods, which convey it further down the
river, before it is arrested by the flood tide, in proportion
as the flood discharge overpowers the enfeebled tidal flow
near the tidal limit ; and it is thus brought under the
influences of a greater exposure and stronger tidal currents,
which prevent its being readily deposited. The flood tide
at springs, being larger in volume and more rapid in flow
than at neaps, brings in a greater quantity of material ; but
the ebb tide at springs, being for the same reasons more
powerful, counteracts this tendency to accretion, except in
sheltered creeks and near the limit of the tidal flow. The
flood tide comes in specially loaded with detritus after storms.
Accordingly, a river is most exposed to the accumulation of
sediment in its upper tidal portion when the fresh-water dis¬
charge is small, and near its outlet after storms. Nevertheless,
the occurrence of slack tide at different places all along the
tidal portion successively, obviates the continuous accumulation
of deposit in any particular spot, such as takes place beyond
the outlets of silt-bearing tideless rivers ; whilst the almost
constant tidal oscillations are unfavourable to silting-up.
The alluvium in most tidal rivers is of mixed origin, being
partly derived from inland, and partly from the sea, varying
in proportion according to the physical conditions of the
river, and of its adjacent seaboard. Thus the alluvium of
the Ganges comes wholly from inland ; and the fresh-water
discharge in this case is so large, and is so densely charged
with alluvium, that, in spite of the rise of tide of about
12 feet at the mouths of the Ganges in the Bay of Bengal,
the river has formed a delta like sediment-bearing rivers
flowing into tideless seas. The alluvium in the Seine estuary,
on the contrary, is almost entirely of marine origin, for the
fresh-water discharge of the Seine is remarkably free from
sediment ; whereas the flood tide sweeping along the coast
of Calvados before entering the Seine, brings large quantities
of detritus into the estuary from the eroded cliffs. This,
x.] Value of large Fresh-Water Discharge. 233
moreover, is indicated by the good depth which has always
existed in the upper half of the tidal Seine ; whilst the
lower half was encumbered by sands (Plate 9, Figs. 1 and 2).
The alluvium in the Loire estuary affords an example of
mixed origin ; for this torrential river brings down large
quantities of material in flood-time, which have accumulated
in sheltered parts of its upper estuary; whilst the sands of
its outer estuary are of marine origin (Plate 9, Fig. 4).
Those rivers are naturally most easily maintained and im¬
proved, which are least exposed to the introduction of
alluvium, either from inland or from the sea.
Influence of Fresh-water Discharge. It has been already
pointed out that the fresh-water discharge of a tidal river
reinforces the ebb tide, and consequently is favourable to
the maintenance of a tidal river. Provided the fresh-water
discharge is not densely charged with alluvium, and is not
so large as to overpower the flood tide and thus prevent
the tidal oscillations, making the river approximate to the
condition of a tideless river, the value of the discharge in
maintaining a tidal estuary is proportionate to its volume.
Thus the tidal Seine, with an average fresh-water discharge
into its estuary of 28,000,000 cubic yards of fairly clear
water every tide, is in this respect in a much more favour¬
able position than the Mersey estuary, which receives a
fresh-water discharge averaging only about 2,500,000 cubic
yards per tide of more turbid water. Though a large fresh¬
water flow reduces proportionately the volume of tidal water
that can enter a tidal river, this loss of tidal flow does not
signify, since the volume of the ebb is maintained, unless the
fresh water brings down more alluvium than can readily be
dispersed, in which case the zone of deposit is unfavourably
concentrated by the restriction of the limits of the tidal
oscillations.
The amount of alluvium, however, brought down by a
river can only be very slightly modified by protecting its
234 Influences of Fresh-Water and Tidal Flow. [chap.
banks from erosion ; and therefore the scope for the removal
of defects resulting from the condition of the fresh-water
discharge of a river, is very limited compared with the
alterations that can be effected in the tidal flow.
A small fresh-water discharge has the disadvantage of
rendering the transition from the tidal to the non-tidal
portion of a river very abrupt, where the river flows into
a large estuary, and the tidal rise at its mouth is considerable.
Thus the river Mersey, which affords access to the largest
class of vessels up to Liverpool, becomes unsuitable for
ordinary inland navigation before its tidal limit is reached,
about 46 miles from the bar at its outlet. The Thames
also, which is the waterway, below London Bridge, for the
largest port in the world, can only give accommodation to
small river craft near its tidal limit. A large fresh-water
discharge, on the contrary, enables vessels of good draught
to penetrate long distances inland. Thus the Rhine, with
a basin about fifteen times that of the Thames, provides an
open navigable waterway far inland up to the flourishing
ports of Mainz and Mannheim ; and Rouen, the fifth seaport
of France, is 75 miles from the mouth of the Seine.
Influence of Tidal Plow in Rivers. The most striking
characteristic of the influence of tidal flow is the navigable
importance with which it invests rivers with a small fresh¬
water dischai'ge, and which without its aid would be of no
value to ocean navigation. Great Britain, indeed, owes her
commercial pre-eminence to the tidal wave which travels
round her coasts ; for her rivers, with their small drainage
areas, could never, without tidal influences, have provided
access for the commerce of the world. The channels
of non-tidal rivers are merely proportional to the average
discharge ; and in dry weather the flow is liable to become
so small as to afford a very inadequate depth for navigation.
As pointed out in Chapter III, this variation in discharge
renders non-tidal rivers unsuitable for continuous navigation
x.] hnportance of Tidal Flow in a River. 235
throughout the year, unless, from the great size of their
basins, or from differences in the period of the low-water
flow of their tributaries, their discharge never falls very low.
Tidal flow, however, under favourable conditions, very largely
increases the volume of water passing through the channel,
and also furnishes the additional supply requisite for filling
an estuary when the fresh - water discharge is reduced.
Moreover, the volume of tidal water entering and leaving
a tidal river at each tide, though varying somewhat between
springs and neaps, fluctuates within far narrower limits than
the fresh-water discharge. Accordingly, a tidal river, though
rising and falling with every tide, is much more regular in
the volume of its flow than a non-tidal river ; and it is able
to maintain an outlet channel enormously larger than could
be maintained by its fresh-water discharge. The Mersey
furnishes an instance of extreme disparity between tidal and
fresh-water flow ; for at a high spring tide, the tidal volume
entering and leaving the Mersey estuary is seven hundred
and ten times the average volume of fresh water discharged
into the estuary each tide.
The value of tidal flow in a river is forcibly illustrated
by contrasting the size of the Thames above Teddington and
at Gravesend, and of the Mersey above Woolston weir and
at Liverpool. The influence also of tidal flow in placing
small rivers on a par, as regards navigability at their outlets,
with some of the largest rivers flowing into tideless seas,
is readily manifested by a comparison of their conditions.
Thus the Thames, having a basin only that of the Danube,
affords superior facilities for navigation at high water between
the sea and London than the Sulina mouth ; the Mersey,
with a basin only that of the Mississippi, is equally
accessible at high tide up to Liverpool as the South Pass ;
the Ribble, with a basin ¿V that of the Rhone, is of more
use for navigation up to Preston than the Rhone outlet ;
and the Usk, with a basin only -a±w that of the Volga, has
236 Flow of Flood and Ebb Tides. [chap.
nearly double the depth in its navigable channel at high
water of fair neap tides up to Newport, which it is hoped
may be obtained by dredging at the most favourable Volga
outlet, and about three times the present available depth
in that river from the Caspian Sea up to Astrachan.
Influences of Flood and Ebb Tides. The flood tide running
up a river, flows with a greater velocity, and has a more
powerful scouring effect during the earlier half of its flow,
when the channel is more restricted, and before the banks
in an estuary are covered. The greater density of sea-water
augments the scour of the flood tide, by causing the incoming
current to pass along the bottom under the lighter water
of the ebb, which is more or less mixed with fresh water.
The flood tide having thus at first to overcome the friction
against the bottom, and the opposing current of the last of
the ebb, does not attain its maximum velocity till a little
time after the turn of the tide. The scour of the ebb is
greatest during the latter half of its flow, when it runs in
a more restricted channel, and with a greater inclination
owing to the fall of the tide outside ; and the ebbing current
attains its greatest velocity at about half ebb, when the banks
in an estuary begin to uncover, and the waters begin to
converge towards the low-water channel.
A river is in the best condition when the flood and ebb tides
follow the same course, and thus both assist in maintaining
the depth ; but this is not generally the case in a wide
channel, or in an estuary, for the flood tide adopts as straight
a course as possible, whereas the ebb tide runs close along the
concave banks of a river, shifting from side to side, like the main
current in a non-tidal winding river. The flood and ebb currents
are, accordingly, conflicting in their actions, the one tending to
scour out where the other tends to deposit. The ebb current
forms the fairly continuous winding low-water channel found
in most wide tidal rivers and estuaries ; whilst the flood tide
forms the straight blind channels on the opposite side to the
x] Divergence of Flood and Ebb Currents. 237
ebb channel; which the action of the ebb prevents it from
completing. These blind channels are clearly defined in the
low-water charts of the estuaries of the Humber, Mersey,
Clyde, Seine, Weser, Loire, and Ribble, and in the plan of the
river Usk (Plates 7, 8, and 9). In the narrower upper por¬
tions of tidal rivers, the flood and ebb are necessarily confined
to the same channel ; but even under these conditions, the
reversal of the direction of the current prevents the action of
the flood and ebb in a winding channel being quite uniform.
The preponderance, however, of the ebb tide, due to the fresh¬
water discharge, increases with the distance from the mouth,
owing to the gradual diminution of the tidal influx in ascend¬
ing a river.
The introduction of the early flood tide under the ebb
current, and the divergence in the directions of the flood and
ebb, enable the ebb current to continue flowing down during
the first of the flood ; and therefore the sea water gets higher
up a river, and the fresh water lower down in each tide, than
if the flood and ebb currents were directly opposed. Accord¬
ingly, there is a less direct backing up of the fresh water, and
a greater intermixture of salt and fresh waters, with their
burdens of alluvium, than might have been anticipated, which
is further aided by eddies resulting from irregularities in the
banks and bed.
Conditions affecting Tidal Flow. A large rise of tide at
the mouth of a river affords a good navigable depth at high
water, facilitates the introduction of the flood tide, creates
energetic flood and ebb currents, and ensures the propagation
of the tidal influence some distance inland ; and it is conse¬
quently favourable to the maintenance of a tidal river. The
Dee, the Mersey, the Ribble, the Severn, and the Usk are
peculiarly favoured in this respect (Plate 7, Fig. 12 ; Plate 8,
Fig. 10, and Plate 9, Figs. 12 and 14); whereas the Nervion,
the Maas, the Loire, the Weser, and the Clyde are less fortun¬
ately circumstanced (Plate 6, Figs. 13 and 17 ; Plate 8, Fig. 8,
238 Distances of Tidal Flow up Rivers, [chap.
and Plate 9, Figs. 5 and 9). The distance up a river to which
the tide extends is also largely affected by the rise inland of
the river-bed. When the inclination of the river-bed is very
small, owing to the flatness of the country through which the
river flows, as in the case of the Maas, the Scheldt, the Seine,
and the Thames, the tide penetrates a considerable distance up
the river. Thus the tide flows up to Ghent on the Scheldt,
110 miles from the sea ; to Martot on the Seine, 90 miles from
Havre ; and to Teddington on the Thames, 67 miles from the
Nore. The length of the tidal Scheldt is due to the flatness
of the low-lying lands of Flanders through which it flows, for
the rise of tide at its mouth is less than at the mouth of the
Seine and of the Thames. The length of the tidal Mersey,
on the other hand, is less than that of the tidal Weser, owing
to the rise of the bed of the river above Eastham, though the
average rise of tide at the mouth of the Mersey is about twice
the rise at the mouth of the Weser (Plate 7, Figs. 3, 5, and 12,
and Plate 9, Fig. 9). The extent of the tidal influence,
accordingly, varies in proportion to the tidal rise and inversely
as the rise of the river-bed ; and the position of the tidal limit
changes with the differences in rise between springs and neaps,
and is, moreover, affected by the volume of the fresh-water
discharge, which in flood-time pushes the tidal limit seawards.
High-Water Line in a Tidal River. The high-water line
along a river is generally fairly level, and is not susceptible of
much alteration by improvement works carried out in a river.
When, however, an estuary faces the direction of the travel
of the tidal wave and is funnel-shaped, the level of high water
is raised in ascending a river by the concentration of the tidal
wave. The raising of the tidal wave in flowing up the British
Channel is a well-known instance of this influence, causing the
range of tide at the mouth of the Severn to become about
double the range in the open sea outside. The rise of the
high-water line is also noticeable in the Seine and the Weser,
resulting from the above causes (Plate 7, Figs. 1 to 4) ; but
x.] Variations in High-Water Line. 239
the rise in the case of the Mersey, though due to similar
causes at Liverpool, must be attributed to the rapid rise in
the river-bed in the upper part of the estuary from the accumu¬
lation of sand (Plate 7, Figs. 5, 6, and 12). When the outlet
of a river is restricted in width, without a corresponding deep¬
ening in the narrow part and shoaling in the wider channel
above, or when the channel of a river is abruptly narrowed,
the ingress of the flood tide is checked, and there is a fall in
the high-water line. Thus the contracted outlet of the Adour
reduces the rise of tide in the river (Plate 6, Figs, to and 11) ;
and the narrowness of the trained channel of the Seine causes
a fall in the high-water line above Quillebeuf (Plate 7, Fig. 1,
and Plate 9, Fig. i). Irregularities in the channel also produce
irregularities in the tidal flow. A rise in the high-water line
inland affords a river a surplus tidal capacity beyond the
normal high-water level at sea, and an extended tidal influence ;
whilst a depression of the high-water line causes a loss of tidal
capacity, and a reduction in tidal influence.
Lowering of the Low-Water Line. The low-water line
depends upon the depth and regularity of the channel ; and it
can be lowered considerably by the deepening of the channel,
and the removal of shoals which impede the efflux of the ebb.
The promotion of the outflow of the ebb increases its scouring
efficiency, by causing a larger body of water to flow down at
each tide, and augmenting its fall near low water. The more
thorough efflux of the ebb, moreover, brings more rapidly sea¬
wards any alluvium carried down by the fresh water ; and by
emptying the tidal channel more completely, the tidal capacity
of the river is increased and a greater volume of tidal water
flows and ebbs through the outlet at each tide, promoting the
increase of its depth. Provided the width of the entrance to
the river is adequate, the increased tidal capacity does not
affect the high-water line ; for whilst a larger volume of tidal
water has to enter at each tide, its influx is facilitated by the
deepened channel. The lowering of the low-water line by
240 Lowering of Low-Water Line in Rivers, [chap.
dredging and training works, is indicated on the longitudinal
sections of the Wear, the Tyne, the Tees, the Clyde, the Seine,
the Weser, and the Ribble (Plate 6, Fig. 7, and Plates 8 and 9).
The increase in tidal capacity, however, thus indicated, only
applies to the navigable channel ; and though correctly ex¬
hibiting an improved tidal flow in the channel itself, and in
narrow rivers like the Tyne, and the upper portions of the
Clyde and the Weser, it proves nothing as to the actual influ¬
ence of training works in wide estuaries, like the Tees, the
Seine, and the Ribble, on the tidal capacity as a whole.
Nevertheless, with this reservation, the lowering of the low-
water line indicates a distinct improvement in the condition
of a tidal river, especially in its navigable channel ; for the
greater freedom of discharge thereby exhibited, manifests
a deepening or rectification of the main channel.
Progress of the Tide up a Biver. The high- and low-water
lines on the longitudinal sections of rivers are merely conve¬
nient representations of the highest and lowest levels of the
water at various places on the river (Plates 6 to 9), for high and
low water never occur all along a river at the same time. It
is, indeed, sometimes high water at one part of a river when
it is low water at another ; and the actual water-lines, and the
progress of the tide up a river are shown by simultaneous tidal
curves, obtained by recording the height of the tide at a series
of tide gauges along a river simultaneously (Plate 7, Figs. 1, 3,
and 5). The tidal oscillations at each station are obtained
continuously by recording the changes in level of a float, by
means of a pencil marking a line on a sheet of paper wound
round a revolving cylinder, thereby producing a tidal diagram
of the heights of the river at the periods indicated (Plate 7,
Figs. 2,4, and 6). It will be noticed that on ascending a river
the duration of the flood tide constantly decreases, also that
the rise of the flood tide is very abrupt in some cases, owing
to the opposition offered to its progress by sandbanks, and that
the flood tide would go further up the Seine if not arrested by
x.] Flow of the Tide up a River. 241
a weir at Martot. The duration of high water for 2? hours
at the mouth of the Seine (Plate 7, Fig. 2) is due to the
entrance of a second tidal wave, travelling more slowly along
the coast of Normandy, after the primary tidal wave, coming
into the estuary from deep water in the English Channel, has
passed. This is a purely local condition, due to the configura¬
tion of the coast, similar to the double tide at Southampton,
but it materially facilitates the access of vessels to the port of
Havre. It must, however, somewhat conduce to the deposit
of material in the estuary, owing to the long period of fairly
slack water, only partially interrupted by a reverse current
flowing out of the estuary near Havre. The flood tide has
a shorter duration than the ebb, even near the outlet of a river;
and its period, as well as the extent of its influence, is reduced
by a large fresh-water discharge.
Owing to the differences in the time of high water in a river,
a tidal river is never completely filled, as, indeed, is evident
from an inspection of the simultaneous tidal curves of the
Seine, the Mersey, and the Weser (Plate 7, Figs. 1, 3, and 5) ;
and allowance must therefore be made for this deficiency in
estimating the tidal capacity of a river. It is possible, accord¬
ingly, to increase the tidal volume in a river by facilitating the
influx of the flood tide by the removal of shoals or other
obstructions, and thereby making the tide rise higher in the
inner parts of a river before the turn of the tide at the mouth.
This result is indicated by the occurrence of high water at
places along a river at shorter intervals after high water at the
mouth than previously, such, for instance, as has been effected
on the Tyne by dredging, where the interval between high
water at Newcastle and the mouth has been reduced from
one hour to twelve minutes. An advance, on the contrary,
of the time of high water at the mouth of an estuary, close
to the open sea, proves that the estuary becomes full earlier
than formerly, and therefore that its tidal capacity has been
reduced, which is exemplified by the case of the Seine, where
R
242 Bore on the Severn, Amazon, Seine, etc. [chap.
the training works in the estuary have been followed by very
extensive accretions, and the time of high water at Havre has
consequently been advanced 38 minutes.
Bore on Rivers. When the progress of the flood tide up
a river is impeded by extensive shoals, the tide has to rise at
the mouth till it acquires a sufficient head to overcome the
friction caused by the obstruction ; and then it forces its way
up as a rapidly-travelling crested wave, producing, as it passes
along, an instantaneous reversal of the current, and a sudden
rise of the water in the river. This phenomenon, which is
termed a bore, is developed mainly at spring tides, and
is increased by wind and by a funnel-shaped channel ; it is
a well-known occurrence on the Severn, the Seine, the
Hooghly, and the Amazon ; and it was observed by Capt.
Moore, R.N., on the Tsien-Tang Kiang1, in China, in 1888.
The bore on the Severn is about 6 feet high near Newnham ;
and on the Hooghly, it attains a height of over 7 feet. On the
Amazon, it appears like two or three waves in succession, from
12 to 15 feet high ; and on the Tsien-Tang Kiang, its maximum
height is said to vary between 10 and 15 feet. The bore on
the Seine reaches its maximum, of about 6 feet, at Caudebec,
as, indeed, is evident from the simultaneous flood-tide curve
between that place and Honfleur (Plate 7, Fig. 1), which
represents to a distorted scale the true form of the bore, with
an almost vertical breaking face in front, and a slight rise,
instead of a hollow like a wave, behind.
Effect of Obstructions placed across Tidal Rivers. It has
sometimes been assumed that, as the tide when flowing into
bays or creeks tends to fill them up gradually with sand or
silt, therefore the tidal flow and ebb in rivers is injurious, and
that rivers would be in a better condition if the tide was ex¬
cluded from them, as the sedimentary matter brought in by
the flood tide would be thus shut out, and not be deposited in
1 ' Report ou the Bore of the Tsien-Tang Kiang,' Commander Moore, R.N.,
London, 1888; and a Further Report in 1892.
x.] Effect of Sluice-Gates on Tidal Rivers. 243
the river channel. This opinion, however, is not borne out by
experience. For instance, the harbour of Rye, at the mouth
of the river Rother, was injured by reclamations, and by the
erection of Scott's sluice, which, stopping the flow of the tide
up and down the river, caused a rapid accumulation of silt in
the channel below the sluice ; and the subsequent destruction
of the sluice produced an improvement in the outlet. It is
related by Mr. Cordier that the harbours of Dunkirk and
Gravelines suffered injury from a similar cause1. Mr. Bouniceau
describes the deterioration of the channel of the river Vire, in
Vays Bay, by the erection of tidal gates across the river near
Isigny, and its restoration to its original condition, on giving
free admission again to the tide by the removal of the gates2.
Denver Sluice, placed across the Great Ouse, soon occasioned
a silting up of the river below ; and the Grand Sluice across
the Witham, just above Boston, restricting the tidal flow
which formerly extended up to Lincoln, thirty miles higher
up the river, seriously injured Boston as a port.
A deterioration of the channel below is, indeed, the inevit¬
able result of the stoppage of the tidal flow, when the flood
tide is charged with silt. When the tide flows up a river, it
possesses a definite velocity which enables it to hold in sus¬
pension the sand and silt which it carries up with it ; but on
encountering any barrier which prevents it from flowing
further up, it simply rises vertically against the barrier, and
deposits, during the long period of almost absolute slack
water thus produced, nearly all the sand and silt which it
brought up. Thus before the excavation of the new outlet
channel for the Witham, the silt was sometimes heaped up
as high as ten feet against the gates of the Grand Sluice
at Boston in the summer, and was only removed by the
winter floods. Moreover, the effect of such a barrier is not
confined to the portion of the river in its immediate vicinity;
1 ' De la navigation intérieure du Département du Nord,' J. Cordier, p. 105.
2 ' Etude sur la navigation des rivières à marées,' P. Bouniceau, p. 62.
R 2
244 Injuries due to Obstruction of Tidal Flow. [chap.
for though the tidal flow is only absolutely arrested at the
barrier, it is affected for a considerable distance down a river,
in proportion to the proximity of the barrier and the burden
of sediment brought up by the flood tide. For instance, the
limit of the tidal Yorkshire Ouse- has been restricted by
a lock and weir at Naburn, to increase the navigable depth
between that place and York; whilst the flood tide brings up
large quantities of silt with which the waters of the Humber
are always densely charged. In dry weather, when the fresh¬
water discharge of the Ouse is very small, this silt deposits
to such an extent during slack tide in the upper part of
the river, where the tidal scour is necessarily feeble, that
towards the close of a dry summer, the bed of the river is
3 or 4 feet higher than in the spring, for 12 miles below
the lock, and is only restored to its normal condition by the
winter floods.
The tidal flow on the Thames was greatly impeded by the
obstruction presented by Old London Bridge, which formed
a sort of weir across the river, with its numerous piers resting
upon wide foundations protected by piles, so that spring
tides rose somewhat higher, and fell about 4 feet lower below
the bridge than above ; and the checking of the flood tide
caused an accumulation of silt below the bridge. The
removal of Old London Bridge, on the completion of the new
bridge in 1831, and the substitution of bridges with fewer
piers and larger waterways for the old bridges at Blackfriars
and Westminster, aided by some dredging and the constant
removal of gravel and sand from the bed of the river for
building operations, have very materially improved the tidal
condition of the river, so that the range of spring tides has
been increased by ,5® feet at London Bridge, and the low-
water line has been lowered 4! feet at Teddington ; whilst
the silt has been scoured away below London Bridge.
Experience, accordingly, demonstrates that obstructions to
the tidal flow are injurious to the depth of a river for some
x.] Influence of Tidal Flow on River Outlets. 245
distance below them ; and the value of the flux and reflux of
the tide in maintaining a channel, is manifested by the
deterioration which results from a diminution in the tidal
volume passing up and down a river, and the improvement
which follows on its restoration.
Effect of Tidal Flow on the Outlet of a River. It has
been seen that the outlet of a river is the most critical point
of the whole channel, and that, owing to the diminution in
velocity which the current generally experiences at that part,
it is the place where a shoal, or bar, is most likely to form.
Slack water is the period when deposit mainly occurs, and
hence it is of great importance to reduce this period as much
as possible. It is evident that in a simple creek, the longer
and larger the creek is inside, the better will its outlet be
maintained ; for the period of slack water will be reduced
by the tide taking a longer period to flow up and down the
creek, and a stronger current will be produced at the outlet.
The deterioration of the outlet channels of the harbours of
Calais and Ostend, as the shallow tide-covered areas within
the harbours were reclaimed, furnishes examples of the great
injury that may be effected in an outlet by the simple
reduction of the tidal volume passing and repassing through
the entrance. As regards, therefore, the outlet channel, it
will be better maintained in proportion to the tidal capacity
of the river above it.
Forms of Tidal Channels and Estuaries. As it is impossible
to prevent the introduction of alluvium into a tidal river, either
from inland by the fresh-water discharge, or from the sea-
coast by the flood tide, the object to be aimed at is to prevent
its accumulation in the channel of a river, and a consequent
reduction in the navigable depth. Deposit occurs when any
considerable reduction takes place in the velocity of a current
charged with alluvium; and therefore it is important that
uniformity in the rate of the tidal ebb and flow should be
maintained as far as possible, combined with the free admission
246 Vavions Forms of Tidal Estuaries. [chap.
of the flood tide. To comply, accordingly, with theoretical
requirements, the mouth of a river should be large enough
to admit freely at every tide the full volume of tidal water
that the river is capable of receiving ; and the sectional area
of the river should gradually diminish inland, in proportion
as the tidal volume required for filling the receptacle above
it becomes less in ascending the river. A contraction at the
mouth of an estuary, by checking the influx of the flood tide,
and also increasing the velocity of the current through the
narrow outlet, reduces the volume of tidal water entering and
leaving the river, shortens the extent of the tidal influence
up the river, and promotes deposit by the variation in the
rate of flow. A tidal river, therefore, gradually enlarging
in width as it approaches the sea, so as to secure the
free admission of the flood tide and uniformity of flow,
affords the best prospects of maintenance and uniformity
in depth.
The actual forms of estuaries are very varied, depending
upon the configuration of the coast, the physical conditions
of the surrounding country, and the nature of the strata in
which the estuary is situated ; and they are all more or less
encumbered with sandbanks and shoals, between which the
low-water channel of the river meanders, and which are
mostly covered at high tide (Plates 7 to 9). Some rivers
enlarge fairly regularly on approaching the sea, as for instance
the Thames and the Clyde (Plate 7, Fig. 7, and Plate 8, Fig. 7);
whilst others widen out very rapidly, like the Ribble and the
Weser, or emerge abruptly into a large estuary like the Tees,
the Seine, and the Dee (Plate 8, Fig. 4, and Plate 9). Many
estuaries exhibit irregularities in width, such as the Humber
(Plate 7, Fig. 9), the Scheldt, and the Shannon1 ; whilst others
1 Plans and longitudinal sections of these and other estuaries are given in the
plates of my paper on ' The Physical Conditions affecting Tidal Rivers ; and the
Principles applicable to their Improvement.' Manchester Inland Navigation
Congress, 1890.
x.] Influence of Changes in Width on Depth. 247
are contracted in width at, or near their mouth, as for
example the Mersey (Plate 7, Fig. 11) the Loire (Plate 9,
Fig. 4), the Tay, and the Gironde. Some tidal rivers, on the
contrary, enter the sea direct without passing through any
estuary, like the Tyne (Plate 8, Fig. 1), the Wear, the Nervion,
and the Maas (Plate 6) ; but the outlets of such rivers have
generally to be protected by jetties or breakwaters. The
estuaries which are most regular in width, with a trumpet-
shaped outlet, are also most uniform in depth, like the Thames
and the Usk (Plate 7, Figs. 7 and 8, and Plate 8, Figs. 9 and
10); whilst estuaries which present great irregularities in
width, with a narrow outlet, exhibit also great variations in
depth, of which the Mersey and the Loire furnish examples.
Gradual accretion in recesses often reduces the irregularities
of an estuary in respect of the low-water channel, making it
adopt a more trumpet-shaped form and become more regular
in depth, as indicated in the estuaries of the Humber and the
Tay. These deposits, which are the natural consequences of
irregularity of flow due to variations in width, though reducing
the tidal flow and ebb through the outlet, produce some
compensation in the channel above by facilitating the influx
higher up of the tidal water which previously spread over the
recess, and are of no importance in estuaries, where, as in
the case of the Humber and the Tay, the depth at the mouth
and out to deep water is ample. In an estuary, however, like
the Mersey, where the depth over the bar is already inade¬
quate, and where the existing depth of the outlet channel
depends upon the tidal capacity of the upper estuary, which
acts like a sluicing basin, any material diminution in this
tidal capacity would be prejudicial to the depth over the bar,
and should be carefully avoided as far as possible.
General Principles relating to Tidal Rivers. The value
of the tidal ebb and flow in rivers has been demonstrated by
the commercial importance bestowed by the tide on rivers
possessing a small fresh-water discharge, and the contrast
248 Principles of Tidal River Improvement. [chap.
which many of them present at their outlets to tideless rivers
of great size. The importance also of promoting the influx
of the tide up a river to the fullest practicable extent, has
been proved by the injuries produced in the navigable
condition of tidal rivers by restricting the tidal influence.
Accordingly, the primary objects in all schemes of tidal river
improvement, should be the admission of the flood tide as
freely as possible into a river, and the removal of all obstruc¬
tions to its progress up to the furthest practicable limit. Not
only is the tidal volume passing up and down the channel
thereby increased, leading to an enlargement of the channel ;
but the period of slack tide is thus reduced to a minimum,
and the tendency to deposit greatly diminished.
As the fresh-water discharge gives a preponderating
influence to the ebb tide, and thus materially aids in the
removal of alluvium seawards and in the maintenance of
the channel, it should not be diverted from its beneficial
action at the head of a tidal river, except for the necessary
paramount claims of water-supply. This inevitable abstraction
of water from many rivers, however, should be compensated
for, in the case of a tidal navigation, by a corresponding
maintenance of the channel near the tidal limit by dredging,
as the depth of a tidal river at its upper extremity depends
upon the fresh-water discharge.
Nature and theory both indicate that the best form for
a tidal channel or estuary is that of a trumpet, gradually
enlarging in width as it approaches the sea, and having
a more rapid increase in width near its outlet. This form
ensures the freest admission of the flood tide, and the most
complete filling of the tidal receptacle ; and by promoting
uniformity of flow, it secures uniformity in depth. A narrow
outlet to a wide estuary produces a powerful scour, and
consequently a considerable depth in the contracted channel ;
but the scouring influence of the flood tide above the outlet,
and of the ebb tide below, only extends for a limited distance,
x.] Modification of Estuaries by Deposit. 249
depending upon the rise of tide and the extent of the estuary :
and nature tends gradually to shape the channel to an
expanding form by the deposit of alluvium in the wide
sheltered recesses of the estuary, where the current loses
its velocity.
CHAPTER XI.
DREDGING IN TIDAL RIVERS.
Advantages of Removal of Hard Shoals, and Dredging. River Clyde : Original
Condition ; Removal of Hard Shoals ; Jetties and Training Walls ; Deepening by
Dredging, and Blasting Reef ; Extent of Improvements effected, and Increase of
Trade of Port; Need of Maintenance of Depth; Cessation of Floods. River
Tyne : Original Condition; Jetties and Training Walls; State in i860; Piers at
Mouth; Dredging; Regulation Works; Improved State; Influence on Traffic;
Remarks. River Tees: Condition ; Cuts and Jetties; Training Walls; Dredging;
Blasting Reef ; Breakwaters at Mouth ; Improvements in Condition ; Remarks.
River Usk: Description; Defects in Channel up to Newport; Improvements
proposed. River Mersey : Description of Estuary ; Influence of Inner Estuary on
Outlet ; Objections to Training Works in Inner Estuary; Dredging on Bar. Con¬
cluding Remarks.
Many tidal rivers, in their natural condition, are impeded
by hard shoals of clay, gravel, boulders, or rock, which the
scour of the current is unable to remove, even when the river
is regulated by training works. Moreover, in the case of
softer materials, a current is capable of maintaining a depth
in a channel, which its scour is unable to obtain ; whilst in
some rivers leading to important centres of trade, a depth
is required for the accommodation of sea-going vessels of
large draught, which the current of the river would be wholly
incapable of providing, or even of maintaining when artificially
procured. The removal of hard shoals often effects a con¬
siderable improvement in a tidal river, for it facilitates the
influx of the flood tide, and promotes the scour of the ebb,
thus both increasing the tidal volume, and augmenting the
depth over the softer shoals by the improved scour resulting
from the greater velocity of the current. The dredging of
soft shoals, moreover, not merely increases the depth of the
Tidal Rivers improved by Dredging. 251
channel, but also increases the tidal capacity of a river when
the shoals are above low-water level, and even when they are
below this level, in facilitating the flow by the enlargement
of the channel, and by lowering the low-water line. The
deepening of a channel, however, beyond certain limits,
reduces so much the rate of flow, that the current deposits
some of its sediment ; and the depth has then to be main¬
tained by dredging.
The improvement of most tidal rivers involves regulation
and training works, as well as dredging ; but in some cases
the increase in depth has been principally effected by dredging,
of which the Tyne, the Tees, and the Clyde are instances;
whilst in other rivers, training works have been supplemented
by large dredging operations, examples of which are furnished
by the Ribble and the Weser. The reduction in cost of
dredging, effected within recent years, by great improvements
in dredging appliances, has led to a large extension of the
system ; and the certainty of its effects, and the facility with
which the depth of a river can be increased by dredging in
proportion to the development of a port to which the river
gives access, have rendered dredgers almost indispensable
machines for the improvement of rivers.
The success which has attended the efforts to improve
the approach channels to the ports of Dunkirk, Calais,
and Ostend, across the sandy foreshore encumbering their
entrances, and also the deepening of Gedney's channel leading
to New York harbour, by means of sand-pump dredgers, have
resulted in the endeavour to lower the sandy bar of the
Mersey, in Liverpool Bay, by a similar method. Accordingly,
the employment of dredging for the improvement of tidal
rivers, which has rendered such services in deepening their
comparatively sheltered outlet channels, is now being extended
beyond their mouths into the open sea.
252 Training of the River Clyde below Glasgow, [chap.
River Clyde.
The Clyde, draining an area of 945 square miles, was
formerly a small stream encumbered by shoals ; but it flows
into a deep, wide, and long estuary, called the Firth of
Clyde, which extends from Greenock to the sea. Its outlet
is therefore most effectually protected, so that no difficulty
is experienced in maintaining it. Considering that the
improvement of the outlet of a river, across a shallow beach
and exposed to waves and littoral currents, is generally the
most difficult problem in river improvement, the Clyde is
peculiarly favoured in this respect (Plate 8, Fig. 7).
Training Jetties, and Removal of Hard Shoals. Till 1773,
the Clyde was such an insignificant river that it was fordable
on foot at Dumbuck Ford, more than twelve miles below
Glasgow ; and there were numerous shoals higher up h In
that year the system of narrowing the river by jetties was com¬
menced, so that it might scour itself out a deeper channel ; and
hard gravel shoals, which the current was unable to act upon,
were removed by dredging or ploughing. In the two follow¬
ing years, the channel across Dumbuck Ford was made 7 feet
deep and 300 feet wide, at low water, by dredging ; and in
1781, the natural scour of the river had lowered it to 14 feet.
About twenty years later, more than two hundred addi¬
tional jetties were erected ; and low rubble training walls
were constructed connecting the ends of the jetties, to render
the channel uniform in width, and to prevent the slipping
in of the accumulated deposit on the banks, and the conse¬
quent formation of shoals. Also, as the original jetties had
been constructed to no uniform lines, they were regulated
by shortening several, and lengthening others, so that the
river might enlarge uniformly as it approached the sea.
1 1 The River Clyde,' James Deas, Minutes of Proceedings Institution C. E., 1873»
vol. xxxvi, p. 124; and Transactions of the Institution of Naval Architects, July,
1888.
xi.] Regulation and Dredging of the River Clyde. 253
The widths then fixed for the river were, 180 feet just below
the harbour of Glasgow, enlarging gradually to 696 feet at
Dumbarton Castle. The increase, however, in the size and
number of the vessels frequenting the port became subse¬
quently so great, that the river has been widened to 370 feet
at the lower end of Glasgow harbour, enlarging by degrees
to 1,000 feet at Dumbarton Castle. The low training walls
were gradually raised, as the ground behind them became
higher, either by natural silting-up, or by the deposit of
material dredged from the river ; and the land was reclaimed.
Quays were gradually extended along the banks of the river ;
and, from 1807, the channel was continuously deepened,
widened, and straightened, for about thirty years.
Deepening the River Clyde by Dredging. In 1840, a
scheme of improvement was authorized, in accordance with
which the subsequent widening and deepening of the river
have been carried out (Plate 8, Figs. 7 and 8)1. The present
navigable condition of the river is, indeed, the result of the
systematic dredging operations inaugurated at that period,
and since then carried on without intermission with gradually
improved and more powerful plant. In 1844-5, five bucket-
ladder dredgers were at work in the river, capable of dredging
to a depth of from 13 to 20 feet ; and 234,000 cubic yards of
material were removed in the year. The work proceeded
at about the same rate during the next six years, till in 1852
a dredger working to a depth of 225 feet took the place of
the smallest of the original dredgers ; and the amount raised
during the five following years averaged about 350,000 cubic
yards annually. A dredger capable of working to a depth
of 284 feet was added in 1856, so that in 1856-7 the quantity
dredged reached 506,000 cubic yards, and the next year
630,000 cubic yards. The dredged material, up to 1862, was
discharged into punts of 8 cubic yards capacity, and de-
1 A recent plan and longitudinal section were furnished me by Mr. James Deas,
the engineer to the Clyde Navigation Trustees.
254 Volitme of Materials Dredged from Clyde. [chap.
posited on the foreshores and low-lying land adjoining the
river ; but since then the material has been conveyed in
steam hopper barges, holding from 240 to 320 cubic yards,
and deposited in deep water in Loch Long, an inlet from
the Firth of Clyde. This method of disposing of the dredged
materials has enabled much larger quantities to be extracted
from the river ; and by the gradual substitution of more
powerful dredgers for the earlier machines, capable of dredging
to depths of 28£ to 34 feet, necessitated by the increase in
depth of the river, from one to over one and a half million cubic
yards have been removed annually since 1874-5, a maximum
of 1,841,300 cubic yards having been dredged in the year
1892-3, which was accomplished by five bucket - ladder
dredgers, and a io-ton steam digger. The total quantity
of material dredged from the river and harbour in the last
forty-nine years amounts to 41,352,800 cubic yards. The
continuance of the deposit of the dredgings in Loch Long
having been objected to as a nuisance, steam hopper barges
of 1200 tons capacity, and having a speed of io| knots an
hour, have been constructed for conveying the dredged
materials further down the Firth of Clyde, to the new
depositing ground three miles below Garroch Head.
A reef of hard whinstone rock was discovered in 1854,
cropping up in the bed of the river, 925 feet long and 320
feet broad ; and it had to be bored and blasted, which was
effected from a barge by diamond drills and dynamite,
involving a cost of ,£70,000 for the removal of 110,000 tons
of whinstone and boulder clay, so as to provide a depth of
20 feet at low-water spring tides over the reef.
Results of Dredging in River Clyde. By means of dredging,
aided in the earlier stages of the improvement works by
training and regulation, the Clyde has been converted from
an insignificant stream into a deep navigable river, capable
of giving access to ocean-going vessels of large draught up
to Glasgow, where extensive basins and large graving docks
xi.] Changes effected by Dredging in the Clyde. 255
have been constructed for their accommodation. A com¬
parison of the longitudinal sections of the bed of the river
in 1758 and 1891 (Plate 8, Fig. 8), indicates the great increase
in depth that has been effected in the Clyde between Glasgow
and Port Glasgow, along a distance of about 19 miles, amount¬
ing on the average to about 22 feet between Glasgow and
Dumbarton. In 1758, the Clyde was only 15 inches deep
at Glasgow at low water, 18 inches about 5 miles below
Glasgow bridge, and 2 feet at Erskine and Dumbuck ; whilst
the rise of tide at Glasgow was only %\ feet. At the present
time, the bed of the river is practically level from Glasgow
to Port Glasgow, affording a depth of from 17 to 20 feet
at low water along the whole distance ; and the rise of spring
tides at Glasgow is feet, an increase of 8f feet, and 2 feet
more than at Greenock. The high-water line has been
raised about 3 inches at Glasgow since 1758, owing to the
improved form and depth of the channel more than com¬
pensating for the increase in the volume of tidal water
required for filling the river at each tide ; and the actual rise
of the high-water line between Greenock and Glasgow is
nearly 2 feet. The great increase in tidal range at Glasgow
is, however, due to the fall of the low-water line, resulting
from the deepening of the river, which amounts to 8| feet
as compared with low water in 1758. This has naturally
been accompanied by an advance in the time of high water
at Glasgow, and an acceleration of the ebb tide. Thus
whereas, in 1800, high water at Glasgow was two hours later
than at Port Glasgow, it is now only one hour later ; and
whilst, in 1858, a float put into the river at Glasgow bridge
just after high water, only reached Port Glasgow, a distance
of 195 miles, after being carried up and down by the tidal
currents for 43Í tides, in 1881 floats under similar conditions
travelled down 231 miles in 10 tides on the average1. The
1 ' Report on Tidal Velocities in River Clyde between Glasgow and Greenock,'
James Deas, 1881.
256 Development of Trade of Glasgow. [chap.
removal, in 1880, of a weir which stretched across the river
a short distance above Glasgow bridge, has extended the
tidal influence higher up the river.
Till 1818, no sea-going vessels came up the river beyond
Greenock or Port Glasgow, their cargoes being discharged
into lighters for conveyance to Glasgow; in 1821 vessels
drawing 13^ feet came up the river ; in 1850 vessels of 19 feet
draught could get up; and now steamers with a maximum
draught of 251 feet can go up to Glasgow. The improve¬
ment of the Clyde has placed Glasgow in a foremost position
amongst British ports, as the tonnage of its vessels is only
exceeded by London, Liverpool, the Tyne ports, and Cardiff;
and the importance of Glasgow as a seaport has been largely
instrumental in making it the second city, as regards popu¬
lation, in the United Kingdom. The deepening of the river
has also led to the development of shipbuilding on a very
large scale along its banks ; and the recently built, gigantic
Atlantic liners, the Campania and the Lucania, 620 feet
long, 65 feet beam, and 26 feet draught, were launched on
the Clyde.
Remarks on River Clyde Works. The works on the Clyde
have extended over a long period ; they have been carried
out under very favourable conditions in respect of the outlet ;
and the deepening of the channel up to Glasgow has been
gradually increased, to keep pace with the growing require¬
ments of the port, and the augmented draught of vessels.
The hard shoals which were lowered by dredging, and the
rocky reef which was removed by blasting, were natural
obstructions which prevented a deepening of the channel by
scour, and impeded the tidal flow ; and their removal consti¬
tutes a permanent improvement in the river, facilitating its
maintenance as well as increasing its depth. The deepening,
however, of the bed of the river right up to Glasgow, to
the extent to which it has been carried, gives an artificial
character to the river, with a depth considerably in excess
xi.] Dredging for Maintenance of Depth in Clyde. 257
of the power of the currents to maintain, and consequently
necessitates its maintenance by artificial means. As the
mouth of the river is situated on a deep and well-sheltered
estuary, it is not exposed to the importation of sand stirred
up by the waves from the bottom during storms, or drifting
along the coast. Accordingly, the only deposit that can
accumulate in the river comes from the alluvium brought
down by the river, from the sewage and other refuse of
Glasgow, and from the sandbanks opposite Port Glasgow
and Greenock, which afford but a limited supply when com¬
pared with that of an open sea-coast. The depth, however,
of the channel has been long ago carried beyond its natural
limits, and it is constantly being increased to meet the
requirements of navigation ; so that each year the channel is
becoming more artificial in its character, and its maintenance
more costly. Consequently, a considerable portion of the
materials which are being dredged from the river every year
consists of deposit, which, if not periodically removed, would
soon reduce the available depth of the river. The amount
of deposit dredged from the river, which constitutes the
work of maintenance, varies from year to year. In 1883-4,
out of 1,040,700 cubic yards dredged, 703,000 cubic yards
consisted of deposit; whilst in 1891-2, as much as 824,100
cubic yards of deposit were removed, and only 491,800 cubic
yards in 1892-3. Probably from 600,000 to 700,000 cubic
yards would have to be dredged annually from the Clyde
merely to maintain its depth ; and the dredging required for
maintenance will increase with every augmentation of depth,
unless the volume of sediment is reduced by diverting the
sewage of Glasgow from the Clyde, an improvement which
is urgently needed on sanitary grounds as well as for the
sake of the river.
One incidental result of the straightening and deepening
of the Clyde, has been the cessation of inundations from river
floods in the low-lying parts of Glasgow ; for the last time
S
258 Original Condition of River Tyne. [chap.
that the river rose above the level of the quays of Glasgow
harbour, was in 1856. This immunity from inundations is
due to the facility with which the flood discharge passes
down the greatly enlarged bed of the river.
River Tyne.
The river Tyne has been frequently compared to the Clyde,
because both the system of jetties and training walls, for
contracting and regulating the channel and concentrating its
scour, and also dredging, have been resorted to for its improve¬
ment. It differs, however, from the Clyde in some important
particulars. It was originally in a better condition than the
Clyde ; and it is a somewhat larger river, having a basin of
1,130 square miles. Its course was not impeded by hard
shoals ; but, instead of opening into a deep sheltered estuary,
it flows direct into the North Sea (Plate 8, Fig. 1). New¬
castle, the port for which the river had to be improved, is
ioJ miles from the sea-coast. The river between Newcastle
and the sea was very irregular in width, its navigable channel
was winding, shoals rose in some parts above low water in
the centre of the river, and a bar existed at its mouth.
Up to 1842 no improvements had been effected; and at
that period there was a depth, at the shallowest place in the
main channel, of only 2 feet at low water, and 143 feet at
high-water spring tides; and the high-water level at Newcastle
was sometimes fropi 7 to 10 inches lower than at the sea.
Jetties and Training Walls. Between 1843 and 1858 the
river was systematically regulated by jetties, whose extre¬
mities were eventually connected by low training walls of
rubble and slag, after the spaces between them had been
partially filled up by the deposit of silt in the slack water h
These works were aided by a small amount of dredging
annually, commenced in 1838 with one dredger on a very
1 Minutes of Proceedings Institution c. e., vol. xxvi, p. 406.
xi.] Training of Tyne, and Piers at Mouth. 259
small scale, the amount raised being gradually increased,
first by a more powerful dredger, and then by the employ¬
ment of a second dredger in 1855, and a third dredger in
1857, whereby the yearly dredging was augmented from a
minimum of 9,700 cubic yards in 1841, to 449,000 cubic
yards in i860. The minimum depth in the channel up to
Newcastle, at high-water spring tides, was thereby increased
to 18 feet, a gain of 3! feet ; high-water mark at Newcastle
attained the same level as at sea ; and the low-water line
was lowered about 16 inches on the average. The amount
of improvement effected by these works, up to i860, was small,
and had been accomplished slowly, but, being produced by
natural scour, the river was sure to maintain its depth ; and
the cost of the works was moderate. The depth of the river,
however, was far from satisfactory, as indicated by the longi¬
tudinal section of the river in i860 (Plate 8, Fig. 2) h The
depth over the bar at the mouth was only 6 feet at low-water
spring tides, and 21 feet at high water ; the river also above
was still greatly impeded by shoals ; and the navigable channel
was circuitous.
Tynemouth Piers. The converging breakwaters protect¬
ing the entrance to the Tyne, and facilitating the removal of
the bar by dredging, have been referred to in Chapter IX
(p. 213). They were commenced in 1856, with the object of
deepening the approach to the river across the bar, being
necessitated by the exposed state of the bar projecting beyond
the coast-line in front of the mouth of the river. The break¬
waters consist of a rubble mound, surmounted by a super¬
structure, with masonry facework and concrete hearting, and
founded from 7 to 21 feet below low water2 (Plate 8, Figs. 1
and 3). The entrance of 1,300 feet between the pierheads
allows the flood tide to pass freely up the river. These piers,
1 'River Tyne Improvement,'P. J. Messent, 1888.
2 'Harbours and Docks,' L. F. Vernon-Harcourt, pp. 317-321, and plate 7,
fig. 9.
S 2
26o Dredging Operations in River Tyne. [chap.
though constituting an essential portion of the improvement
works, have involved a greater cost than actually required for
the navigation of the river, owing to their original limits having
been extended for forming a refuge harbour.
Dredging in the River Tyne. The comprehensive scheme
for the improvement of the river authorized in 1861, consisted
mainly in deepening the river by very large dredging opera¬
tions from its mouth right up to Hedwin Streams, about 8f
miles above Newcastle, a total distance of 19^ miles. In 1861
two new very powerful dredgers were set to work, in conjunc¬
tion with two of the former dredgers ; and two more were
started in 1863, since which period there have been always
five, and in most years six dredgers at work in the river1. At
the commencement of these operations, the amount dredged
in a year rapidly rose from 1,243,000 cubic yards in 1861, up
to a maximum of 3,515,700 cubic yards in 1866 ; since which
time the volume of material raised from the river has been
gradually diminished with occasional fluctuations, the smallest
amount since 1861 having been raised in 1893, when 780,100
cubic yards were dredged from the river. The total volume
of materials dredged from the Tyne and two docks, between
1838 and 1893, amounted to 62,309,000 cubic yards, out of
which only 1,866,000 cubic yards were dredged in the twenty-
three years preceding 1861 ; whilst in the last thirty-three
years, 60,443,000 cubic yards have been removed, giving
a yearly average of 1,831,600 cubic yards, almost equal to the
total output during the earlier period. The dredging has been
effected by bucket-ladder dredgers, loading into hopper barges
which convey the dredged material out to sea between two
and three miles beyond the mouth of the river, and deposit it
in a depth of water of 20 fathoms at low tide. Three of the
dredgers have been at work every year since they began work
in 1843, 1855, and 1861 respectively, having each been only
1 A statement of the annual quantities of dredging was furnished me by Mr.
"Messent, the engineer to the River Tyne Commission.
xi.] Dredging mid Regulating River Tyne. 261
laid up once for repairs during periods of from four months to
a year ; and the maximum amount raised by one dredger in
a year was 974,400 cubic yards, accomplished by the newest
dredger in 1866, which, though not working in 1893, has
raised the greatest amount of material, reaching a total of
14,553,ooo cubic yards in its thirty years of work, including
laying up on three occasions for repairs.
Dredging is being carried on in the Tyne at the present
time, in order to maintain the depth already attained from
the sea up to Newcastle and two or three miles above, and
also for extending the increase in depth higher up the river
to the limit proposed (Plate 8, Fig. 1).
Auxiliary Improvement Works in the Tyne. An abrupt
contraction in the channel rather less than a mile above the
bar, known as the ' Narrows,' which was an obstruction to
navigation and impeded the flow of the tide, has been widened
from 400 feet to 700 feet. The channel, moreover, has been
rendered more uniform in width ; and sharp bends, inconvenient
for navigation, have been eased by the removal of several
projecting points, of which Bill Point and Whitehill Point were
the most obstructive (Plate 8, Fig. 1). By the removal of
Whitehill Point, the river has been widened 300 feet ; and the
rocky projection of Bill Point has been cut back 400 feet,
involving the blasting of 380,000 tons of rock under water by
means of dynamite, and the removal altogether of two million
tons of rock and soil. This latter obstacle not merely pro¬
duced a very sharp bend in the river difficult to steer round,
but also, by its height of 72 feet, dangerously obstructed the
view of any approaching vessel. The old stone bridge at
Newcastle has been replaced by a large swing-bridge affording
an enlarged waterway, thereby greatly facilitating naviga¬
tion and the flow of the tide. Above Newcastle, the river has
been regulated by training works, the channel having been
widened in some places and narrowed in others, and points
removed ; and a straight cut has been substituted for a very
2ó2 Changes effected by Works in the Tyne. [chap.
circuitous channel near Blaydon, reducing the length by three
quarters of a mile, and promoting the tidal influx.
Results of Works in River Tyne. The great improvement
in the navigable condition of the Tyne between Newcastle and
the sea, exhibited by a comparison of the longitudinal sections
of the river in i860 and 1888 (Plate 8, Fig. 2), amounting to
an increase in depth of about 20 feet, has been almost entirely
accomplished by the unprecedented dredging operations
carried out in the river during the last thirty-three years. In
i860, in spite of a rise of nearly 15 feet at springs, vessels
drawing more than 20 feet could not enter the Tyne at high-
water spring tides ; and vessels of 17 to 18 feet draught were
detained sometimes for two or three months in the river during
easterly winds, owing to the reduction in the available depth
over the bar by wave oscillations. Moreover, only vessels
with a draught not exceeding 15 feet could reach Newcastle,
even at high-water spring tides, along a tortuous channel
between sandbanks ; whilst above Newcastle, small craft alone
could navigate the river, and only during high water. At the
present time, the depth at the outlet is about 24 feet at low
water of springs ; there is a deep piece of channel, 1 \ miles long,
near the mouth of the river, in which vessels can moor in 30 feet
of water at low-water spring tides, waiting for an opportunity
of going out to sea ; and between Shields and Newcastle,
where formerly steamers of only 3 to 4 feet draught used to
ground for hours, there is now a depth of 20 feet throughout
at the lowest tides h The river has also been deepened for
some three or four miles above Newcastle, to 18 feet below
low-water spring tides ; and this deepening will eventually be
extended, at a reduced depth of 12 feet below the same level
above Scotswood, to the proposed limit of the improvements
at Hedwin Streams, about miles above Scotswood, and
actually three-quarters of a mile beyond the limit of the tidal
influence at spring tides in i860.
1 ' River Tyne Improvement,' P. J. Messent, 1888, p. 4.
xi.] Alteration of Tidal Flow in the Tyne. 263
The improvement of the Tyne has raised high-water mark
at Newcastle about a foot, which is about the present rise of
the high-water line between the sea and Newcastle, the line
having been level in i860 ; whereas it had a fall inland in 1842.
The low-water line has been lowered about 3 feet at Newcastle
since 1860 ; and this lowering had already reached a maximum
of about 7 feet at Scots wood in 1885, which will be still
further increased as the deepening progresses in the upper
part of the river. It has been estimated that when the pro¬
posed deepening has been completed, the tidal capacity of the
river will have been increased by 20 million cubic yards, which
will very materially assist in maintaining the channel. The
tidal range at Newcastle is now about 9 inches greater than
at the mouth of the river.
High water at Newcastle, in i860, occurred 1 hour 3
minutes later than high water at Shields ; in 1872, the differ¬
ence in time had decreased to 23 minutes ; and now high
water occurs at Newcastle only 12 minutes after high water at
Shields. The tidal interval also between Newcastle and
Newburn, which is about 7 miles higher up the river, has been
reduced from 29 minutes to 8 minutes since 1860.
The deepening of the river has so greatly facilitated the
passage of land floods down the river that the height attained
by the floods has been reduced, so that the lands and quays
bordering the river between Wylam and Newcastle are no
longer subject to periodical inundation.
The great increase in the depth of the river has naturally
exercised a powerful influence on the trade of the Tyne ports,
which is manifested by the rapid increase in the yearly ton¬
nage of the vessels entering and leaving the river, so that the
Tyne ports come next in this respect to London and Liver¬
pool amongst British ports. The average tonnage of the
vessels has also gradually increased from 149 tons in 1854, up
to 445 tons in 1886 ; whilst many vessels of between 2,000 and
4,000 tons enter the river now, which could not possibly have
264 Maintenance of Depth in River Tyne. [chap.
gained access to it in i860. Shipbuilding also has been
largely developed in consequence of the improved condition
of the river.
Remarks on River Tyne Works. As in the case of the
Clyde, the deepening of the Tyne, which has been mainly
effected by dredging, has been carried considerably beyond
the limits which the scour of the currents could maintain, not¬
withstanding the increase in the tidal capacity of the river.
The depth, indeed, in which the outlet between the pierheads
is situated, will prevent the introduction of much sand from
the sea, except during violent storms ; and whatever sand
may be brought by waves into the harbour will deposit in
its sheltered area, and be readily removed. The alluvium,
however, brought down by the river from inland, and the
refuse thrown in from the banks, will deposit to a large
extent in the deepened channel ; and therefore dredging will
always be required for maintaining the depth attained, the
average annual amount of which can only be determined
after the completion of the deepening of the upper part of
the river.
The river Tyne possesses the advantage of a well-defined
channel, restricting the necessary works of improvement
within narrow limits, and free from the uncertainties attending
the improvement of the outlet channel of a river through a wide
sandy estuary in an exposed situation. It furnishes an in¬
stance of a river in which dredging operations have been
carried out to the largest extent within a distance of about
14 miles ; and though the deepening effected up to Newcastle
is not greater than that on the Clyde near Glasgow, the
amount of material removed is half as much again as in the
Clyde, and has been mainly accomplished in two-thirds of
the time. The large volumes of material, however, raised from
the bed of the Tyne, vast as they appear, sink into insignifi¬
cance when compared with the operations of nature ; for the
Mississippi carries into the Gulf of Mexico in flood-time more
xi.] Regulation of River Tees below Stockton. 265
material in ten hours than the maximum amount dredged out
of the Tyne in a year.
River Tees.
The river Tees, which has a drainage area of 735 square
miles, was formerly very irregular and tortuous between
Stockton and Middlesbrough. A short distance below the
latter town it enters a wide sandy estuary, through which it
flowed in a winding shifting channel for about 7 miles before
reaching the North Sea (Plate 8, Fig. 4).
Straight Cuts and. Jetties. The first improvement in the
tidal portion of the river was made in 1810, when a cut, 220
yards long, was opened near Stockton, which reduced the
length of the river by about %\ miles, and increased the depth
at Stockton from 9 feet to 11 feet '. Another cut was made
in 1830, having a length of 1,100 yards; but though removing
a very bad bend, it did not shorten the river very much, whilst
it cost twice as much as the first cut, and reduced the tidal
capacity. About the same period the system of regulating
and contracting the river by timber jetties, projecting out at
right angles to the banks, was introduced, and was carried on
till about 1852. By these means a deeper and more direct
channel was formed between Stockton and Cargo Fleet below
Middlesbrough, a distance of about 7 miles ; the increase in
depth varying from about 2 to 5 feet. The jetties, however,
produced irregularities in the depth, shoaling the channel
in the intervals between them, and merely deepening it in
front of them. The works, moreover, by cutting off a large
area of tide-covered strand between the jetties, which was
raised by deposit, had a somewhat injurious effect on the
channel just below. There were also shoals in the channel
through the estuary, especially where it divided into two
branches on each side of the Middle Sand, and also over the
bar at the mouth of the estuary (Plate 8, Fig. 5).
1 Minutes of Proceedings Institution C. E., vol. xxiv, p. 64.
266 Training and Dredging in River Tees. [chaf.
Training Walls in the Tees. Owing to the deficiency in
depth of the channel below Stockton, and the growing import¬
ance of Middlesbrough as the centre of a rapidly developing
iron trade, the improvement of the river between Stockton
and the sea by training walls and dredging was decided upon
in 1853. Longitudinal training walls, composed of mounds of
slag from the iron-works, were gradually formed on each side of
the channel between Stockton and Middlesbrough, connecting
the ends of the jetties and placed a suitable width apart.
The channel was at the same time straightened and widened
where expedient ; two projecting points were cut off ; and the
foreshores at the back of the training walls have been reclaimed
The training walls, raised from 4 to 7 feet above low water,
were by degrees extended through the estuary, as funds per¬
mitted, along the more direct southern channel ; and the
northern channel was shut off, thereby directing the currents
into a single fixed channel1. These training walls regulated
the bed of the channel, and rendered the channel stable, with¬
out materially increasing its depth except in soft places.
Deepening the River Tees by Dredging. Clay, with stones
and boulders, underlies sand and silt in the greater part of the
bed of the river, being met with at from 7 to 9 feet below
low water ; and a ledge of lias rock extended across the
channel at the eighth buoy, so that two-thirds of the channel
was laid bare at low water, leaving only a narrow gorge on
the west side, through which the fall of the low-water line was
considerably increased. Dredging, accordingly, and blasting
the reef were essential for obtaining an adequate improvement
of the river, for scour could not have removed the boulder
clay ; and the rocky ledge, besides presenting a dangerous
shoal, impeded the influx and efflux of the tide.
The dredging has been effected by bucket-ladder dredgers,
loading into hopper barges towed to sea by steam-tugs, which
1 * River Tees Improvements,' John Fowler, Minutes of Proceedings Institution
C. K., 1887, vol. xc, p. 346 ; and ' The River Tees,' W. Fallows.
xi.] Dredging, and Blasting Rock in the Tees. 267
deposit the dredged materials 3 miles beyond the barl.
Dredging operations were commenced in 1854 with one
dredger; and a second dredger was set to work in 1866, a
third in 1870, a fourth in 1878, and a fifth in 1884. The
total amount dredged from the commencement up to the
close of 1880, was 6,380,000 cubic yards ; and in the next five
years, nearly the same amount was removed from the river
by means of four double-ladder dredgers and a Priestman
grab, the total reaching 12,371,700 cubic yards at the close
of 1885, giving an average of nearly 1,200,000 cubic yards
of dredging per annum during the five years. The total
volume of materials raised from the Tees up to the close of
1893, reached about 19,784,000 cubic yards, an average annual
amount of dredging of 926,000 cubic yards having been
carried out during the last eight years ; and the maximum
quantity dredged in one year was accomplished in 1885,
when 1,431,100 cubic yards were raised from the river
The rocky ledge was removed, over an area of 13 acres, by
means of diamond rock-boring drills and blasting with dyna¬
mite, 124,000 cubic yards being thus shattered, and the débris
raised, procuring a depth of 14 feet at low water, at a cost of
£26,780. The depth over the rock has now been increased
to 20 feet below low water, by using cast-steel claws in a
bucket dredger, alternating with two buckets on the ladder.
North and South Gare Breakwaters. The two break¬
waters converging across the wide exposed outlet of the Tees
estuary, have been already referred to in a previous chapter
(p. 212). Their main object is to protect the estuary and the
training walls, for the deepening of the outlet channel over the
bar has already been effected by the tidal scour in the
deepened channel directed by the training walls. Being
1 ' Dredgers and Dredging on the Tees/ John Fowler, Minutes of Proceedings
Institution C. E., 1884, v°l* lxxv, p. 239.
2 Recent particulars relating to the dredging were supplied by Mr. G. J. Clarke,
engineer to the Tees Conservancy Commission.
268 Reclamations in Estnary of the Tees. [chap.
chiefly composed of refuse slag from the neighbouring iron¬
works, they have been constructed at an unusually small
cost (Plate 8, Figs. 4 and 6). Reclamations have already
been made in the shallower portions of the estuary; and
further reclamations are contemplated alongside the training
walls, at the upper end near Cargo Fleet. These reclamations
are favoured by the protection afforded by the breakwaters ;
but if they are gradually extended, they will proportionately
reduce the value of the estuary, as a sort of sluicing basin
enclosed by the breakwaters, for maintaining the depth at
the outlet. Gradual accretion of the estuary outside the
trained channel will, indeed, result from alluvium brought
down by the river, or sand driven in by storms ; but any
acceleration of this process should be avoided, for it will
entail considerable increase in dredging for maintaining the
depth near the mouth and over the bar outside.
Results of Works in River Tees. The great increase in
depth obtained in the Tees since 1851, from the sea up to
Stockton, as exhibited on the longitudinal section of the
river (Plate 8, Fig. 5) \ is the result of dredging, aided by
the fixity of channel afforded by the half-tide training walls
through the estuary. There was a depth of only 8 feet over
the bar outside the mouth of the river in 1851, at low-water
spring tides. Though no dredging has been carried out on
the bar beyond the mouth, the improved scour into and
out of the deepened channel has lowered the bar to 19 feet
below low water ; and a similar depth has been obtained
by dredging for about 3 miles above the mouth of the river,
except at two or three places where the depth is reduced
to about 15 feet. As, however, the rise of spring tides is
15 feet, there is an available depth of 30 feet at high-water
spring tides for over 3! miles from the bar, being a gain
1 A recent plan of the Tees was sent me by Mr. J. H. Amos, secretary to the
Tees Conservancy Commission, and also a longitudinal section prepared by Mr.
G. J. Clarke, the engineer to the Commission.
xi.] Improvements effected by Tees Works. 269
of about 10 feet as compared with the shallowest place in
1851. Proceeding up the river, there is a minimum depth
in mid-channel at low water, of 10 feet near Cargo Fleet,
8 feet opposite Middlesbrough, and 5 feet near Stockton,
the dredging having effected an average increase in the depth
of the bed since 1851, of about 10 feet up to Newport, and
about 5 feet from Newport to Stockton. It is proposed to
continue deepening the upper part of the river by dredging,
so as to obtain an available depth of 18 feet at low water
up to Middlesbrough, and of 12 feet up to Stockton.
High-water mark at Stockton is about a foot higher than at
the sea ; and the rise of spring tides at Stockton is 15 feet, the
same as at the mouth of the river. Since 1851, the low-water
line has been lowered 4 feet at Stockton by the improvement
of the river below ; and as the flood tide must have been
considerably impeded in flowing up the river in its former
condition, it is probable that high water did not rise as
high at Stockton formerly as it does now, so that the actual
increase in the range of spring tides there probably amounts
to 5 feet or more. A further lowering of the low-water
line may be anticipated as the deepening of the channel
up to Stockton progresses. The tidal flow and ebb have,
accordingly, been considerably facilitated in the channel up
to Stockton by the deepening of the river ; and the tidal
capacity of the channel has been increased.
Since 1878, vessels of from 1,000 to 3,000 tons have been
able to leave Middlesbrough fully laden and proceed to
sea drawing 21 feet of water. Eventually, when the proposed
extension of the depth of the navigable channel by dredging
shall have been accomplished, vessels of moderate draught
will be able to navigate the river between Middlesbrough
and the sea at any state of the tide ; and vessels of large
draught will be able to get up to Stockton at high tide.
Remarks on River Tees Works. The deepening of the
Tees has been effected mainly by dredging, as in the case
270 Prospects of Maintenance of Depth in Tees. [chap.
of the Tyne and the Clyde ; but the dredging operations
have hitherto been on a more moderate scale, and the increase
in depth considerably less. The proposed deepening, how¬
ever, has not reached such an advanced stage as on the other
two rivers ; and the distance over which it extends is only
12 miles, as compared with about 19 miles on the Tyne
and the Clyde. The removal of the eighth buoy scarp con¬
stitutes an important permanent improvement in the river ;
but in spite of the improved tidal condition and increased
tidal capacity of the channel, the deepening is being carried
beyond the limits of natural scour, and every increase in
depth will involve additional dredging for maintenance.
The lowering of the bar beyond the mouth of the Tees
having been effected by the scour resulting from the regulated
and deepened river, the maintenance of the depth over it is
dependent on the permanence of this scour. The break¬
waters at the mouth of the estuary, by narrowing the outlet,
tend, in the first instance, to intensify the tidal scour over the
bar; but they also, by the shelter they afford to the wide
estuary, and the changes in velocity they necessarily produce
in the tidal currents through the restricted outlet and over
the wide estuary, favour accretion in the wide sheltered
expanse, leading to eventual reclamation, and diminution in
the tidal capacity of the natural sluicing basin. The reclama¬
tion of 2,600 acres has, indeed, been already effected along
the shallow borders of the estuary (Plate 8, Fig. 4) ; but these
portions added but little to the tidal capacity, owing to their
insignificant depth at high tide ; whereas the sale of the
reclaimed lands provided funds for the extension of the
improvements. Further reclamations, however, of 1,000 acres
are now proposed, situated on each side of the channel below
Cargo Fleet ; and it is probable that these will be followed
later on by other reclamations, especially when the completion
of the North Gare breakwater increases the shelter in the
estuary. Accordingly, the continued maintenance of the bar
xi.] State of River Usk below Newport. . 271
by tidal scour cannot be regarded as assured ; but it is fairly
certain that the income derived from the sale of the reclaimed
lands will more than provide for the maintenance of the depth
over the bar by dredging. Otherwise it would be expedient
to abandon further reclamations, and to maintain the tidal
capacity of the estuary by periodically removing, by dredging
in the sheltered area, a volume of material equivalent to
the amount of accretion.
River Usk.
The river Usk, having a drainage area of only 634 square
miles, and flowing into the Bristol Channel about 4 miles
below the port of Newport to which it gives access, has been
far more favourably endowed by nature than the rivers
just described. Opening at right angles into a sheltered
portion of the Bristol Channel a long distance from the open
sea, its outlet needs no protection ; and though the flood tide
comes into the Usk densely charged with the yellowish mud
which the waters of the Severn estuary hold in suspension,
and which readily deposits in the creeks, foreshores, and
sheltered places of the river, the main low-water channel
is generally fairly free from deposit, being scoured by the
rapid tidal currents aided by the fresh-water discharge. The
range of tide, moreover, in the lower part of the river is
exceptionally large, attaining 40 feet at a high spring tide
at the mouth, and 34 feet at Newport bridge 5 miles above,
which constitutes the upper limit of the seaport (Plate 8,
Figs. 9 and 10). The trumpet shape of the high-water
channel of the river raises the high-water line of spring tides
one foot higher at Newport bridge than at the mouth, in spite
of the tortuous course of the river. As the river rises in
a mountainous and rainy district, its fresh-water discharge
is large in proportion to the size of its basin. Notwith¬
standing however, these favourable conditions which have
enabled a large seagoing trade to grow up at Newport, the
272 Defects in Outlet Channel of River Usk. [chap.
state of the river is not free from defects, which demand
attention in view of the constantly increasing draught of
vessels and requirements of shipowners, and the keen com¬
petition of the neighbouring South Wales ports.
Defects in Approach Channel to Newport. Though there
is an ample depth in the river up to Newport at high-water
spring tides for vessels of the largest draught, the depth
in mid-channel at high water of the lowest neaps is reduced
by some shoals in the middle of the river below the require¬
ments of navigation, so that occasionally vessels of large
draught have been delayed for two or three days from in¬
sufficiency of depth in the channel. The highest shoal is
a sandbank which has accumulated below the old dock
entrance, driving the low-water channel against the left bank
of the river, over which the depth at the highest part, in the
centre of the river, at high water of the lowest neaps, is
only 17 feet (Plate 8, Figs. 9 and 10). This is the most
defective portion of the river within the limits of the port ;
but as it is above the part traversed by the vessels of largest
draught, which enter the Alexandra docks, a hard shoal
composed of stone, compact clay, and boulders, known as the
• gravel patch,' lying in the centre of the river below the
entrances to the principal docks, is a more important obstacle,
though the depth over it at the lowest high tides is 2a feet.
The other defects in the river consist in the marked divergence
of the flood and ebb channels in three or four places, occasioned
by the bends, and resulting in sandbanks in the middle of the
river ; the tendency of the ebbing current to cut into the
concave bends, increasing the irregularities in the width of
the river ; and the hard shoal formed at the mouth of the
river Ebbw, by the gravel brought down by this tributary and
the refuse carried down in flood-time from the works along
its banks.
Improvements proposed for the Usk. It was proposed
in 1883 that the 'gravel patch' shoal should be removed, as
xi.] Conservancy and Dredging of River Usk. 273
not merely liable occasionally to impede the passage of
vessels of very large draught, but also as preventing the
natural scour of the river producing its full effect both above
and below the shoal ; for its removal consequently affords
a prospect of improving the condition of the river beyond the
mere deepening of the channel at the shoal h It was also
urged that the injuries to the Usk caused by the accumulation
in its channel of hard refuse brought down by its tributaries,
the Ebbw, and the Afon Llwyd which flows in above Newport,
should be arrested at their source, by preventing the manu¬
facturers using these rivers as receptacles for their refuse.
Stringent provisions for dealing effectually with this nuisance
were, accordingly, obtained by the Newport Harbour Com¬
missioners in their Act of 1890. Another recommendation
made in 1883 was also carried out, namely, a series of cross-
sections of the river between Newport bridge and the mouth,
to enable future changes in the river to be readily ascertained.
By aid of the cross-sections taken in 1884, and some new
ones over the same lines, the state of the river was thoroughly
investigated in 1:890 ; and as an improved depth over the
shoals appeared more urgent than in 1883, a scheme was
proposed for deepening the river by dredging2, so as to
provide a minimum depth in the main channel, at high water
of the lowest neaps, of 24 feet from the bridge down to near
the upper entrance to the Alexandra docks, and 26 feet
for the remainder of the channel down to the mouth of the
river (Plate 8, Fig. 10). This improvement of the depth
was approved by the Commissioners ; and the proposed
dredging is being carried out by a dredger with hopper barges
built specially for the purpose. Training works also for
regulating some wide reaches of the river, thereby securing
1 ' Report on the River Usk and its Tributaries,' L. F. Vernon-Harcourt, 1883.
2 'Report on the Improvement of the River Usk,' L. F. Vernon-Harcourt, 1890 ;
and 'The River Usk and Harbour of Newport,' L. F. Vernon-Harcourt, Brit. Assoc.
Cardiff Meeting, 1891.
T
274 Dredging for Improvement of River Usk. [chap.
a better maintenance of the deepened channel and preventing
a further increase of the sinuous course of the river and of
the variations in width, constituted a portion of the proposed
works. The dredging required for obtaining the proposed
depth is only about half a million cubic yards, which is
a very small amount compared with the quantities dredged
from the Tyne, the Clyde, and the Tees ; but it is scattered,
of small thickness, and some of it very stiff material ; and
it is probable that a greater depth will be considered ex¬
pedient after the proposed depth has been reached. The
lowering of the shoals will facilitate the influx of the tide ;
and by also promoting the more thorough efflux of the ebb,
the low-water line, which at present has an irregular fall, will
be lowered ; and the difference of 6 feet between the range
of tide at the mouth and at the bridge, will be materially
diminished.
The river Usk is so well endowed by nature, that its
improvement was not so urgent as on less favoured rivers ;
but a river with such natural capabilities cannot be allowed
to fall behind other rivers in accommodation for navigation.
Though the great range of its tides renders it inexpedient
to attempt to make the Usk accessible at all states of the
tide, like the three rivers previously considered, the works
in progress will provide a channel of ample depth for some
time before and after high water, of even the lowest tides, for
vessels of the largest draught up to the principal docks.
River Mersey.
The river Mersey, draining an area, in conjunction with
the Weaver, of 1,722 square miles, flows through an extremely
irregular estuary into the sea. The river, which is narrow
and winding near Warrington, gradually expands below,
but is contracted at the rocky gap of Runcorn, below which
the estuary again widens out, increasing rapidly in width
below Hale Head, opposite which the Weaver flows in ;
xi.] Form and Channels of Mersey Estuary. 275
and attaining a maximum width of about 3 miles in front
of Ellesmere Port, it contracts again, and flows in a com¬
paratively narrow channel, with a minimum width of about
three-quarters of a mile, between Liverpool and Birkenhead
(Plate 7, Fig. 11). The river emerges abruptly at New
Brighton into Liverpool Bay, and flows through sandbanks
and over a bar into the Irish Sea. The depth of the low-
water channel exhibits similar variations, for it is shallow
and variable down to Garston, below which it deepens to
a minimum depth of about 50 feet in the narrow channel
past Liverpool, and gradually shoals through Liverpool Bay,
till in passing the crest of the bar, the depth has sometimes
been reduced to about 9 feet below the lowest low water
(Plate 7, Fig. 12). The flood tide enters the estuary across
the bar, and also to a smaller extent from the west along the
narrow Rock channel, which is prevented from becoming
deeper by a ledge of rock across its entrance to the main
channel. Above the narrow neck, the flood tide pursuing
a straight course, maintains a deep blind channel in the
direction of Eastham along the Cheshire shore. The ebb
channel, on the contrary, whilst constantly shifting its course
in the upper part of the inner estuary, always emerges near
Garston on the Lancashire shore, and following along that
shore, only joins the flood-tide channel between Dingle Point
and Pluckington Bank.
Influence of Inner Estuary on Mersey Outlet Channel.
The wide inner estuary, together with the neck, has an area
of 22,500 acres, of which 17,300 acres become bare at low-
water spring tides ; and it forms a vast natural sluicing basin
which, when filled during an equinoctial spring tide rising
29 feet at the bar, has a tidal volume of 710 million cubic
yards flowing out past New Brighton on the ebb, and 281
million cubic yards at a low neap tide. This large volume
of tidal water, being directed and concentrated by the narrow
neck in flowing into Liverpool Bay, maintains the outlet
T 2
276 Condition of Mersey Estuary and Outlet. [chap.
channel through the sandbanks, which gradually shoals as
the outflowing current spreads out with a reduction in
velocity ; and the sands are only prevented from forming
a continuous strand across the bay by the tidal ebb and flow,
for the fresh-water discharge of the river only amounts on
the average to about 2,500,000 cubic yards per tide, which
could only maintain a comparatively insignificant channel.
The tidal capacity of the inner estuary is to some extent
preserved by the slight preponderance the fresh-water dis¬
charge gives to the ebb. but mainly by the changes in the
channel which, by fretting the sandbanks periodically, prevent
their consolidation and growth, and by stirring up the sand
facilitate its removal by the current. The inexhaustible
supply of sand in the bay, and the prevalence of north¬
westerly winds blowing into the estuary, together with the
difference in the velocity of the current in the neck and above,
render it probable that most of the sand which encumbers
the inner estuary has come from outside ; and as in this
instance the maintenance of the outlet channel depends on
the maintenance of the tidal capacity of the inner estuary, the
possibility of a continuance of the accumulation of sand inside
is not devoid of interest. If the process is still in operation,
it is a slow one, for the decennial surveys of the estuary,
though exhibiting fluctuations, do not indicate any marked
change ; but they are not exact enough, nor have they been
carried on long enough, to decide the question. The depth
over the bar has remained fairly uniform for several years,
exhibiting an average of about 10 feet below low water of
the lowest tides, with slight variations which may be accounted
for by changes in the position and in the cross-section of
the channel ; and the area of the cross-section of the channel,
rather than the maximum depth, would be the true measure
of the scouring capacity of the ebb current.
The bar appears to have gradually shifted seawards
between 1840 and 1880, indicating a strengthening of the
xi.] Objections to Training Works in Mersey. 277
ebb current, which must have been due to the regulation of
the narrow channel by the formation of quays on each side,
and also to the deterioration of the Rock channel, thereby
checking the diversion of a portion of the ebb in that
direction. Any increase in the tidal capacity of the inner
estuary would have had a similar effect ; but beyond the
temporary influence of a large fret, further accretion is more
probable than erosion. Since 1880, however, a retrograde
movement of the bar seems to have commenced, for the bar
in 1890 was more than half a mile inside of its position in
1880, and approaching the place which it occupied in 1870
(Plate 7, Fig. i a).
Objections to Training Works in Inner Mersey Estuary.
Above Garston, the Mersey is quite unsuitable for vessels of
large draught, on account of the wanderings of the channel
and the inadequacy of its depth. The channel might be
fixed by training walls, which would increase the depth across
the sands by concentrating the scour ; and it could be
deepened to any extent by dredging. This was the scheme
proposed for providing an outlet for the Manchester Ship-
Canal below Runcorn, in 1883 and 1884. This is the system
of improvement that has been adopted with success in the
estuaries of the Clyde and the Tees ; and it might have been
adopted in the Mersey estuary if Liverpool and the bar
outside had not existed. There was no bar, and little
prospect of accretion at the mouth of the Clyde. On the
Tees, the training works have been carried nearly out to
the bar, and the outlet has been protected by breakwaters ;
but in this case, if sand comes into the estuary, in spite of
the projection given to the outlet, accretion will take place,
and the bar will be raised. The training works, however,
in the inner Mersey estuary would have left the sands free
to be carried in from outside as before by the rapid flood
tide through the neck ; and some of this sand, being brought
into the sheltered spaces on each side of the trained channel,
278 Lowering Mersey Bar by Dredging, [chap.
would have been deposited beyond the scouring influence of
the ebb, which would have been mainly confined within the
fixed channel. The sandbanks would, accordingly, have been
gradually raised on each side of the estuary, without chance
of disturbance by any wandering of the channel, thereby
greatly reducing the tidal capacity of the inner estuary.
The consequent reduction of the volume and velocity of
the outflowing current would have inevitably led to the
shoaling of the outlet channel, the raising and widening of
the bar, and the shifting of its position somewhat landwards.
Accordingly, independently of the paramount interests of
Liverpool, the scheme was much less favourable for Man¬
chester than the canal skirting the estuary down to Eastham,
which has been carried out.
Dredging the Mersey Bar. The bar of the Mersey differs
from the bars which existed in front of the Tyne and the Tees,
in being much further out from the coast, being about xi
miles beyond the actual mouth of the river at New Brighton,
owing partly to the large mass of sand encumbering Liver¬
pool Bay, and also to the powerful ebbing current from the
inner estuary. This distance, and the situation of the bar in
relation to the coast-line preclude the possibility of resorting
to converging breakwaters for lowering the bar. A training
wall across the sandbanks, in continuation of the left bank of
the river beyond the Rock lighthouse, would undoubtedly
concentrate the flood and ebb tides in the main channel by
closing the Rock channel and other subsidiary outlets, and
consequently lower the bar ; but such a work would be costly
on shifting sandbanks and in an exposed situation. The bar,
however, prevents Liverpool being accessible at all states of
the tide ; and Atlantic liners, which have crossed the ocean
at a high speed, are liable to be delayed outside the bar if
arriving near low water.
Since 1876, the approach channels to Dunkirk and Calais,
across the sandy foreshore in the open sea, have been greatly
xi.] Sand-Pump Dredgers at Mersey Bar. 279
improved by dredging with sand-pump dredgers ; and more
recently a similar system has been adopted with success at
Ostend. The results achieved in dredging sand in the sea
at these places and elsewhere, suggested the expediency of
giving the system a trial on the Mersey bar1 ; and in
September, 1890, a sand-pump dredger was set to work on
the bar. Two sand pumps were at first fitted up on two
steam hopper barges, capable of filling the hoppers of 500 tons
capacity in an hour, and dredging to a depth of 36 feet2.
These two dredgers succeeded in augmenting the depth over
the crest of the bar, in the centre line of the navigable channel,
from 12 feet to i8¿ feet below low water of the lowest tides, by
the beginning of 189a, by the removal of 657,000 tons of sand,
equivalent to about 438,000 cubic yards. During the next year
and a half, however, though the dredgers removed nearly three
times this quantity, the depth remained stationary, partly
owing doubtless to sands from the sides being drawn down
the steepened slopes, and partly to deposit in the deepened
channel, which will always require periodical removal. Accord¬
ingly, a sand-pump hopper dredger was specially built for the
work, capable of conveying 3,000 tons of sand in its eight
hoppers, which can be filled in three-quarters of an hour3. The
vessel can convey its load to the depositing ground at the rate
of 10 knots per hour ; and it commenced working in the latter
part of 1893. The greatly increased dredging power thus
obtained, aided by the greater stability of the large vessel,
should enable the desired depth, of 26 feet over the bar at low
water of the lowest tides, to be attained, and the deep channel
enlarged to an adequate width, so as to give access to vessels
of large draught at all states of the tide. Till the deep channel
across the bar has been formed, and maintained for two or three
1 Minutes of Proceedings Institution C. E., vol. lxxxix, p. 35.
2 ' Dredging Operations on the Mersey Bar,' A. G. Lyster, London Maritime
Congress, 1893, Sect. I, p. 50.
3 ' The Sand Dredger " Brancker," ' A. Blechynden, London Maritime Congress,
1893, Sect. Ill, p. 52.
28o Value of Dredging for Tidal Rivers.
years, it will be impossible to tell how much sand will have to
be annually raised to preserve the depth, and consequently
what the cost of maintenance will be. Besides, however, the
important deepening of the channel over the bar, which is sure
to be effected by the dredging plant at present in operation
in Liverpool Bay, the improved channel will facilitate the
influx of the flood tide up the navigable channel, and will
check the changes in the position of the main channel.
Concluding Remarks. The foregoing examples of the
results of dredging operations demonstrate the great improve¬
ments that can be effected by dredging in the navigable
condition of tidal rivers, far exceeding the limits of depth
obtainable by natural scour directed by training walls. Whilst,
however, natural scour is sure to maintain the depth which
it has formed, and may be promoted by the removal of hard
shoals and other obstacles which impede its action, and by
an extension of the tidal limit, dredging soon carries the
depth beyond the limits of natural maintenance, in spite of
the increase of the tidal volume produced by the lowering
of the low-water line, and therefore involves a continuance
of the dredging for maintaining the depth. In seaports on
tidal rivers, which form centres of trade, the increase in traffic,
which generally results from an increased depth in the river,
and greater facility and safety of access, usually suffices to
pay for the expenditure on the deepening and the costs of
maintenance. Moreover, in the present day, the continued
prosperity of a port depends upon its providing the increased
depth and accessibility demanded by shipowners ; and many
ports are indebted to dredging for enabling them to maintain
their position and to extend their trade.
CHAPTER XII.
TRAINING WORKS IN ESTUARIES.
Simplicity of Regulation Works compared with Training Works in Estuaries.
Influences of Training Works in Estuaries ; Improvement of Trained Channel ;
Concentration of Scour ; Accretion promoted outside Channel. Training Works
in the Wash: Condition of Rivers draining Fens; River Witham, Training
Works, New Outlet Channel, Effects of Works ; River Weiland, Training Works,
Necessity for Prolongation ; River Nene, Straight Cut, its Prolongation, Training
Works, Need of Extension ; River Ouse, Eau Brink Cut, Straight Cut below
King's Lynn, Training Works in Estuary ; Remarks on Improvement of Outfall
of these Rivers. River Dee Training Works: Original Condition of Estuary:
Training Works in Estuary, Reclamation the Main Object, Unfavourable Line
adopted, Effects of Works ; Remarks on Works, their Injurious Effect on Estuary.
Ribble Estuary Training Works: Former Condition, Changes in Channels and
Outlet ; Training Works, their Extension, and Improvements in Channel; Works
in Progress and Proposed ; Effects of Works on Condition of Estuary; Remarks.
Weser Training Works: Description of River and Estuary; Training Works in
Tidal River and Estuary ; Results ; Remarks on Improvements and Prospects.
Where the channel of a tidal river is fairly well defined,
as in the upper portion of most tidal rivers, and in rivers
which flow direct into the sea, such as the Tyne and the
Nervion, training works are necessarily limited to a regulation
of the existing channel as regards its width, so as to secure
a greater uniformity in depth, aided sometimes by a rectifica¬
tion of its course at sharp bends. The training works carried
out on the Nervion below Bilbao, on the Maas below Rotter¬
dam (Plate 6, Figs. 12 and 18), and on the Tees between
Stockton and Middlesbrough (Plate 8, Fig. 4), regulated
more or less defined, though irregular channels, giving them
a gradually expanding width seawards, and a somewhat
straightened course, so far as the existing conditions rendered
282 Effects of Training Works in Estuaries, [chap.
reasonably practicable. A slight reduction in the tidal area
of a river may be occasioned by such works, owing to the
narrowing of the channel at unduly wide places ; but this
diminution in tidal capacity is generally compensated for
by the progress upwards of the flood tide being facilitated,
and by the lowering of the low-water line from the increased
depth resulting from the training works. Moreover, an im¬
provement in depth and an increase in tidal capacity can
readily be obtained by dredging in the trained channel.
When, however, a river emerges abruptly into a wide estuary
encumbered with sandbanks, through which it flows in a
winding, shallow, and shifting channel to the sea, the improve¬
ment of its outlet channel by training works is a far more
complicated problem.
Influence of Training Works in an Estuary. Longitudinal
training walls formed of mounds of rubble, slag, or fascines,
guiding the channel of a river through a sandy estuary, fix
the formerly wandering channel, and by concentrating the
fresh-water discharge and the main flow of the flood and
ebb tides into the trained channel, increase the scour and
thereby improve the depth. Accordingly, training works
fix, straighten, and deepen the channel of a river through
a sandy estuary; and if their influence ended there, they
would be unquestionably of great value in guiding rivers
through sandy estuaries. In concentrating, however, the main
currents within the trained channel, they withdraw the scour¬
ing influences of these currents from the rest of the estuary;
and by stopping the wanderings of the channel, they prevent
the periodical erosion of the sandbanks, which formerly pre¬
vented these banks from becoming consolidated and increasing
in size. Any materials brought down by the river from
inland, or stirred up by the waves and introduced by the
flood tide, tend to deposit in the comparatively slack water at
the sides of the estuary behind the training walls ; and when
once deposited, they are no longer liable to be disturbed
xii.] Accretion produced by Training Works. 283
by the shifting of the channel. Training works, consequently,
promote accretion in the portions of an estuary outside the
trained channel, wherever the waters coming into the estuary
are charged with alluvium ; and this influence extends to
the sides of an estuary below the ends of the training walls,
where the works terminate within the estuary, owing to the
reduction in the tidal capacity of the upper part of the estuary
by the accretion produced behind the training walls. Ex¬
perience, also, proves that, whilst the accreting influence of
training walls may extend a considerable distance beyond
their extremities, their influence in deepening the channel
by the concentrated scour ceases a short distance below
their termination. The rate of accretion in an estuary
through which a river is trained is dependent upon the volume
of alluvium introduced into the estuary, and the height of
the training walls ; but even low training walls, in so far as
they fix and deepen the channel and concentrate the scour,
promote accretion at the sides of an estuary.
Wherever, therefore, alluvium is brought into an estuary,
the benefits of training works to navigation, in fixing and
deepening the outlet channel, are necessarily accompanied
by a reduction in the tidal capacity of the estuary from the
resulting accretion, which generally is only very partially com¬
pensated for by the increased tidal capacity of the deepened
channel. In some cases, the diminished tidal capacity of the
estuary is immaterial compared with the advantages of an
improved channel ; but in other cases, the loss of tidal
capacity imperils the maintenance of the depth at the outlet,
or the accretion may prejudice the access to ports on the
estuary. The local benefits of training works have often been
exclusively kept in view in designing these works, whilst
their effects on the estuary have been entirely overlooked ;
so that training works have in some instances been attended
by totally unexpected results. Accordingly, before com¬
mencing training works in an estuary, the probable accretion
284 Effects of Training in Clyde and Tees. [chap.
and its consequences should be weighed against the advan¬
tages of an improved channel.
The Clyde is so protected from the introduction of alluvium
from the sea, and the materials brought down by the river
are so effectually removed by dredging, that the training
works in its estuary have been wholly beneficial ; and any
accretion that may have occurred behind the walls has been
far more than compensated for by the improved tidal capacity
of the dredged channel. The Tees estuary is not so secure
from accretion, in spite of the protection afforded by the
breakwaters at its outlet ; but any loss of tidal scour that
may gradually result from the training works, can be made
up for by dredging. In some other estuaries, however, the
balance of advantages is not so manifest ; and under any
circumstances, the consequences of the accretion resulting
eventually from training works, as well as the immediate
benefits to the channel, should be fully considered at the
outset.
Training Works in the Wash.
Four rivers draining the extensive Fen districts, flow into
the Wash, which is a sort of wide estuary or inlet, encumbered
with sandbanks, enclosed by the Lincolnshire and Norfolk
coasts. These rivers, the Witham, the Welland, the Nene,
and the Ouse, draining areas of 1,079, 760, 1,077, ant^
square miles respectively, traverse very flat low-lying districts
for several miles before reaching the Wash, and then pass
through tortuous channels between the sandbanks into deep
water in Boston and Lynn deeps. The small fall of the land
throughout the Fen country necessitates the supplementing
of the straightened and embanked rivers by numerous straight
drains, for the discharge of the inland waters ; and the low
level of the lands has led to the closing of the outlets of some
of the main drains against the tide by gates at sluices
xii.] Sluice-Gates and Training on Fen Rivers. 285
to prevent the inroad of the sea, as for instance at the Black
sluice, Maud Foster sluice, and Hobhole sluice on the Witham,
the North Level sluice on the Nene, and the Middle Level
sluice on the Ouse. Protection, moreover, was further pro¬
vided for this district against the sea by the simple expedient
of arresting the flow of the tide at the Grand sluice just
above Boston on the Witham, and at Denver sluice on the
Ouse, by gates pointing seawards, without due consideration
of the influence of this exclusion of the tide on the outfalls of
these rivers. The restriction of the flow of the tide by these
sluice-gates, together with the diversion to some extent of
portions of the fresh-water discharge from parts of the main
channels by the supplementary drains, produced a deteriora¬
tion in the tidal portion of these rivers. Training works,
however, have been carried out on these rivers for deepening
their outlet channels, and improving the condition of the
ports to which they give access. The average rise of spring
tides in the Wash is 23^ feet.
River Witham. The outfall channel of the Witham be¬
tween Boston and Hobhole, a distance of three miles, was
trained and straightened nearly seventy years ago by banks
of fascines, stones, and earth, in a uniformly increasing section
seawards. This work improved the channel so much, that
the flood tide reached Boston 15 hours earlier; and vessels of
larger draught could get up to the haven More recently
the training works were continued towards the junction of
the channel of the Witham with the Weiland. The direction,
however, of the Witham outfall till it joined the Weiland, was
at right angles to its subsequent course, which followed the
line of the Welland outfall into Boston deeps. This circuitous
channel, occasioned by the hardness of the strata of a project¬
ing piece of foreshore known as the Scalp, was shallow and
very inconvenient for navigation, and not only checked the
progress of the flood tide, but also impeded the discharge
of the land waters ; and the improvement of the outfall, by
286 New Outlet Channel formed for Witham. [chap.
opening a new channel for the river through the Scalp, was
recommended at various times.
In 1880 the proposed outlet channel for the Witham across
the Scalp was at length commenced ; and it was completed
in 1884. This curved channel, 3 miles long, 100 feet wide
at the bottom at Hobhole, and increasing to 130 feet at its
outlet at Clayhole at the head of Boston deeps, was excavated
through silt, clay, peat, and boulder clay 1, reducing the length
of the channel between Hobhole and Clayhole by nearly one
half. The ends of the new channel were guided and secured
by training walls of fascines with clay, weighted at the top by
rubble stone. The river between the Grand sluice and the
new channel was at the same time improved by training and
dredging, affording an increase in depth of 8 feet, and a
navigable depth of 22 feet at high-water spring tides up to
Boston dock. The new channel has caused the flood tide
to commence flowing up at Boston an hour earlier at spring
tides than formerly ; and as the tide no longer rushes up with
a bore charging the river with sand in passing over the
shallows, the deposit at the Grand sluice, which used to
attain a height of over 11 feet at the close of a dry summer,
is now insignificant. The facilitating also of the efflux of the
ebb by the improved outfall has produced a lowering of the
low-water line, amounting to 5£ feet at Hobhole and 4 feet at
the Grand sluice, thereby materially increasing the tidal ebb
and flow, for spring tides formerly rose only 12^ feet at the
Grand sluice ; but unfortunately for the outfall, the tidal
flow of the Witham has been permanently restricted to the
nine miles between deep water and the Grand sluice. The
old outfall channel was only suitable for vessels of 300 tons ;
whereas the new channel gives access to vessels of 2,000 tons.
River Weiland. The outfall of the Weiland was much
improved by training the river with fascine work for 3! miles
1 ' The Witham New Outfall Channel,' J. E. Williams, Minutes of Proceedings,
Institution C. E., 1889, vol. xcv, p. 81.
xii.] State of Weiland ; Straight Cuts on Nene. 287
below Fossdyke Bridge, which lowered the low-water level
7 feet at the bridge ; but want of funds prevented the continua¬
tion of this work down to its former junction with the
Witham. The new cut for the Witham across the Scalp has
diverted the outfall of this river away from the Welland,
so that the two rivers only form an uncertain junction now at
Clayhole ; and the Welland has lost the scour of the Witham
along two or three miles of its outfall channel through the
estuary. Accordingly, in order to improve the outfall of the
Welland, it would be necessary to extend the training works
in a fairly straight line to join the new outlet of the Witham
at Clayhole. Besides deepening and fixing the Welland out¬
fall and facilitating its tidal flow, the prolongation of the train¬
ing walls would also be advantageous to the Witham in
directing the flow of the Welland into Clayhole, thereby
assisting in maintaining the outfall of both rivers into Boston
deeps ; whereas at present the Welland is liable to be diverted
into a more southern channel which at one time it followed.
River Nene. In 1721, Kinderley proposed that a straight
cut should be made below Wisbeach, towards the sea, for
improving the outfall of the Nene. This cut was only com¬
menced in 1770, about five miles below Wisbeach, and was
carried out for a length of ij miles. Though this cut was
considerably shorter than proposed, its effects sufficed to show
that the principle was correct, and that important benefits
might result from an extension of the system further down
the river. The improvement effected by this cut was
gradually lost ; but in 1827-30, the cut was prolonged for
five miles down to Crabhole, being only partially excavated,
and was scoured out to the requisite size on letting in the
river. The scour, in fact, of the fresh and tidal waters com¬
bined, was so effectual, that the sides of the channel had to be
protected by stonework. The tidal flow along the river was
much improved by the cut, so that its effects extended as far
up as Wisbeach, five miles above the cut, to such a degree
288 Improvement of Outfall of Great Ouse. [chap.
that buildings had to be protected from being undermined ;
and the low-water line was lowered more than 10 feet.
The Nene has been trained for a short distance into the
estuary ; but beyond the end of the training works, the channel
is shallow and tortuous ; and it would be necessary to extend
the training walls for three or four miles to reach deep water
in the Wisbeach channel leading to Lynn deeps.
River Ouse. This river was considerably improved above
King's Lynn by the formation of the straight Eau Brink cut,
in substitution for a very tortuous channel, which was com¬
pleted in 1821. This cut lowered the low-water level about
61 feet at its upper end ; and it admitted a sufficient flow
of tide to scour the river up to Denver sluice, enabling the
sill of that sluice to be lowered 6 feet, which greatly benefited
the drainage of the district.
The river has since been improved below King's Lynn by
a straight cut two miles long, and by training the river for an
additional distance of i| miles through the sandy shoals, with
training walls of fascines, which works have effected a further
reduction of 3 feet in the low-water level. The channel,
however, beyond the training works is circuitous and some¬
what shallow ; and an extension of the training walls for about
three miles would materially improve the outfall channel.
Remarks on Outfalls of Fen Rivers. There are only two
methods of improving the tidal portions of the Fen rivers
flowing into the Wash, namely, straight cuts and training
works ; for the removal of the barriers formed by the sluices
is precluded by the interests of drainage. Straight cuts
increase the fall of these rivers, which is very defective ; they
consequently afford a freer discharge to the drainage waters
which, accordingly, have a greater scouring force ; and they
facilitate the tidal flow up and down, which assists in the
maintenance of the outfall.
Training works through the sandbanks in the estuary have
also a beneficial effect both on navigation and inland drainage,
xii.] Outfalls of Fen Rivers: Dee Estuary. 289
for they straighten, fix, and deepen the outlet channel ; and by
facilitating the influx of the tide, they promote the mainten¬
ance of the outlet, and by lowering the low-water line they
facilitate the discharge of the inland waters. The extension,
however, of the training works promotes accretion at the sides,
and facilitates reclamation ; and as the estuary exhibits indica¬
tions of a gradual silting-up along its shores1, and reclamations
of the foreshores are made from time to time, attention will
always have to be directed to the maintenance of the depth at
the outlets of these rivers ; and the training walls will eventu¬
ally have to be prolonged as accretion progresses.
River Dee Training Works.
Original Condition of Estuary. The river Dee, draining an
area of 813 square miles, flows into the Irish Sea through
a wide open estuary which formerly extended up to Chester.
As the estuary faces the north-west, it is exposed to the intro¬
duction of the sands which border the adjacent coasts during
storms from that quarter ; and it is, consequently, encumbered
by extensive sandbanks. Accretion appears to have been
taking place in the upper part of the estuary, along the Flint¬
shire shore, more than two centuries ago, which was subse¬
quently assisted by the enclosure of the marsh lands. There
was, however, a good wide and direct channel near the
Cheshire shore in 1684 (Plate 9, Figs. 13 and 14), which afforded
a depth of 6 feet at low-water spring tides opposite Burton
Point, increasing to 12 feet opposite Heswell, and reaching 60
feet off Air Point at the mouth of the estuary, which with
a rise of tide at springs of 29 feet at the outlet, and about half
that height at Burton Point, provided a good depth at high
water for vessels of that period up to this part of the estuary2.
The channel, nevertheless, was naturally unstable in the sandy
1 Minutes of Proceedings Institution C. E., vol. xxviii, p. 109.
2 'Admiralty Inquiry, Chester, Dee Navigation Improvement,'1849, Drawings
A and b.
u
290 Deterioration of the Dee Estuary.
estuary; and by 1732, the' progress of accretion along the
Welsh shore, above the site of Connah's Quay, had driven the
channel close to the Cheshire shore above Burton Point, re¬
sulting in the formation of a tortuous channel round that rocky
point. Whilst the channel below retained its ample depth and
width, the channel between Chester and Burton Point showed
signs of deterioration ; and the available depth up to Chester
at high-water spring tides was only 9 feet. Accordingly, it
had become evident that works were needed in the upper part
of the estuary to improve the channel, in order for Chester to
retain its position as a seaport.
Training Works in the Dee Estuary. The flood tide enter¬
ing the estuary mainly from the west, formed a deep channel
over towards the Cheshire shore; and so long as the main
channel remained near the Cheshire shore, the ebb tide
followed the flood-tide channel, and a deep wide outlet was
maintained. Any training works in the estuary should mani¬
festly have been designed to preserve this favourable condition,
following approximately the channel of 1684, and thereby
avoiding the diversion of the channel at Burton Point produced
in 1732, and forming a straighter course for the river near
Chester. At this period, however, a company was constituted,
called The River Dee Company, which, in consideration of
providing a channel 15 feet in depth at high-water spring tides
up to Chester, was authorized to reclaim the estuary down to
Burton Point about 7,000 acres in extent. The upper part
of the estuary was thus practically handed over to a land
reclamation company, without due consideration on the part of
the persons interested in the port of Chester, that the outlet
channel was liable to be compromised by the proposed works.
Though ostensibly proposing to improve the navigation up to
Chester, the works were carried out with a view to the most
profitable arrangement for the reclamation of the land, the real
object of the company. Accordingly, instead of following the
course indicated by nature near the Cheshire shore, and
xii.] Training and Reclamation in Dee Estuary. 291
improving the tortuous course just below Chester, the river
was trained between parallel banks like a canal ; the defective
course near Chester was made worse, by training the river at
a very acute angle to secure more land ; and then after
another bend at Saltney directing the channel over to the
Flintshire shore, the training works were terminated close
against the shore at Connah's Quay (Plate 9, Fig. 13). This
arrangement left the whole area of the upper estuary between
the trained channel and the Cheshire shore available for
reclamation, so as to form an undivided estate ; whilst the river
provided a well-defined boundary between the Dee Company's
land and the marshes on the Flintshire side. The training
works appear to have been commenced about 1737, and to
have been completed by 1771, at which period 2.400 acres of
land had been reclaimed from the estuary by the company ;
by 1826 the area reclaimed reached 4.000 acres; and the re¬
maining 3,000 acres are in process of reclamation. Altogether,
including the reclamations by the landowners on the Flintshire
side, the tidal area lost to the estuary on the completion of
the last reclamation, will amount to over 10,000 acres.
The trained channel, besides being diverted from its natural
course, directed the issuing current towards the projecting
Pentre rock near Flint, which turned the channel across the
estuary, as indicated in the charts of 1771 and 1849. The
prolongation of the outer training wall for about i| miles
towards Flint, in a direction diverging from the shore, has
made the channel pass outside the Pentre rock ; but it still
tends to go across the estuary lower down, though its con¬
nexion with the flood channel on the Cheshire shore has now
been cut off at low water1 (Plate 9, Fig. 13). The training
walls, formed of embankments protected by stone, improved
the depth of the channel between Chester and Connah's
Quay, but did not secure the stipulated depth of 15 feet at
1 A chart of the estuary of the Dee in 1892 was furnished me by Mr. Enfield
Taylor, engineer to the River Dee Commission.
U 2
292 Injurious Effects of Dee Training Works, [chap.
high-water spring tides ; and occasionally at a low stage of
the river, the depth was reduced to such an extent that, even
at springs, vessels drawing only 8 J feet could not get up the
river1. The depth, however, of the trained channel has now
been increased, by a reconstruction of portions of the south¬
west training wall, and by dredging, to a minimum depth of
15 feet. The channel, however, in the estuary beyond is
very tortuous and shifting ; and in places between the end
of the training works and Flint, it is only 16J feet deep at
high-water spring tides (Plate 9, Fig. 14).
Remarks on the Dee Works. In the Dee estuary, exposed
to the inroad of sand from outside, and already showing signs
of silting-up in the seventeenth century, reclamations in the
estuary should have been sedulously avoided ; and when
training works became necessary to maintain the depth of the
navigable channel to Chester, they should have been kept as
low as possible, and directed, with an expanding width sea¬
wards, in as straight a course as practicable near the Cheshire
shore. The flood and the ebb tides would thus have been
kept in the same channel, tending to preserve the good depth
below Burton Point ; and any necessary increase in depth in
the trained channel could have been obtained by dredging.
The direction, and almost uniform width actually adopted for
the trained channel, were most unfavourable for navigation,
and for the maintenance of the channel by the aid of tidal
scour ; whilst the wholesale system of reclamation undertaken,
by annihilating the tidal capacity of the upper estuary, has
occasioned a great growth of sandbanks in the lower estuary
from the loss of tidal scour, and a considerable reduction in
the depths of the channels (Plate 9, Figs. 13 and 14). In
fact, if it had been desired to destroy the navigable capabili¬
ties of the estuary of the Dee, it could hardly have been
accomplished in a more effectual manner than by the works
actually carried out by the River Dee Company. The history
1 ' Second Report of the Tidal Harbours Commission,' 1846, p. 10.
Early Condition of Ribble Estuary. 293
of the Dee estuary serves as a warning of the fatal consequences
that may result from mixing up schemes of land reclamation
with the improvement of navigation in a sandy estuary.
Ribble Estuary Training Works.
The Ribble estuary formerly commenced a very short dis¬
tance below Preston ; and about three miles further down, it
commenced expanding rapidly to its existing wide outlet into
the Irish Sea. The river at its mouth drains an area of 8x5
square miles, and used to flow in a shallow winding channel
below Preston down to the wide estuary, through which it
wandered in a constantly shifting channel, amongst extensive
sandbanks, generally branching off into two outlets before
reaching the Irish Sea.
Former Condition of the Ribble Estuary. From a series of
charts of the estuary prepared at intervals between 1736 and
1836, it appears that there were three outlet channels through
which the Ribble in its wanderings flowed into the sea, namely,
the north channel following the northern shore past St. Anne's ;
a middle channel in the centre of the estuary, but varying
considerably in position ; and the South, or Boghole Channel
passing along near the southern shore opposite Southport,
which has remained the most stable in position, owing to the
main flood tide coming in from the south-west1 (Plate 9,
Fig. 11). With the exception of the Boghole Channel, the
position, size, and even the existence of a defined outlet
depended upon the course taken by the channel of the river
through the estuary. In 1736, the main channel of the Ribble
went into the South Channel, with only two small shallow
channels branching off to a central and a northern outlet ; and
a similar condition of things existed in 1761. At the date of
the next survey, taken in 1820-24, the Ribble followed a fairly
1 1 Report on the Extension of the Navigable Channel of the Ribble to the Sea,'
L. F. Vernon-H arcourt, Southport, 1891, pp. 7 to 12; and 'Amélioration de la
partie maritime des Fleuves,' L. F. Vernon-Harcourt, Vme Congrès International
de Navigation intérieure, Paris, 1892, plate 2, figs. 5 and 6.
294 Changes, and Works in Ribble Estuary, [chap.
direct channel to a central outlet; the northern outlet had
entirely disappeared : and the South Channel, though cut
off from the Ribble, still extended nearly 3 miles above
Southport, and was the outlet for the Crossens stream draining
an area of 55 square miles. The South Channel had become
connected with the Ribble again in 1836 ; but the river pos¬
sessed rather a better outlet by the northern channel, of which
no trace existed twelve years previously. Accordingly, the
South Channel had proved the most stable and deepest outlet
for the Ribble ; and in the event of training works being
resorted to for fixing and deepening the channel of the river
through the estuary, it was clearly indicated by nature as the
most favourable outlet for the navigable channel.
Training Works in Ribble Estuary. Up to 1839, the only
works which appear to have been carried out on the river were
some cross jetties, or groynes, which the landowners placed in
front of their property near Preston to prevent the encroach¬
ment of the river in its wanderings. About half a mile below
Preston, a reef of red sandstone extended across the river, with
only a foot of water over it at low water ; and spring tides only
rose 6 feet at Preston, whilst neap tides did not get up to the
quay. Vessels had to unload their cargoes at Lytham into
lighters, drawing only 6 feet of water, for conveyance up to
Preston h
In 1839 the improvement of the river below Preston was
commenced, by blasting and removing the rocky reef ; and
the dredging of gravel from the channel was carried out, the
materials being deposited at the sides to form training walls
on each side of the river. By 1845 the training walls had been
extended for 4 miles down the river, and together with the
dredging, had increased the range of spring tides to 10 feet
at Preston by lowering the low-water line, and had accelerated
the flow of the tide up to Preston at springs by one hour ; and
vessels of 200 tons, drawing 11 feet of water, were able to get
' ' Second Report of the Tidal Harbours Commission,' 1846, p. 10.
xii.] Prolongation of Ribble Training Works. 295
up to the quay. The double training walls were subsequently
continued down to the Naze ; and in 1853 powers were obtained
for prolonging the south training wall 4 miles further, and for
guiding the river Douglas into the trained channel by training
it for two miles down to its confluence with the Ribble (Plate 9,
Fig. 11). These works increased the available depth to a
maximum of 14 feet at high-water spring tides, and enabled
vessels of 300 tons to reach Preston at the highest tides.
The channel, however, below the training works continued
shallow and unstable ; and the outer part of the south training
wall settled, and in one place was cut through by the current,
producing a diversion of the channel.
In 1883 the raising of the existing training walls and
a further extension of the training works were authorized
together with the construction of a dock near Preston and
dredging. It was then considered by the engineers of the
scheme, that a prolongation of the south training wall for
3^ miles, extending a mile below Lytham, in a slightly con¬
cave line towards the channel, and the extension of the
north training wall for 2 miles, terminating less than half
a mile above Lytham, together with the dredging of the
river down to 7 miles from the dock, or 2 miles above
Lytham, would secure a navigable channel out to sea, with
a minimum depth of 6 feet at low-water spring tides, which,
with a rise of tide of 26 feet at springs, would give access to
the largest class of vessels. It was, however, stated by me in
a report to the Corporation of Southport, in March, 1883,
that in order to improve effectually the navigation of the
Ribble, it would be necessary to prolong the training walls
further seawards. The carrying out of the south training
wall to the full extent, with an addition of about a mile in
length of a mound of dredged material, together with some
dredging, had produced by 1890 a general improvement in
depth down to below Lytham, and a lowering of three or four
feet in the low-water line near Preston ; but a shoal extended
296 Dredging and Training in Ribble Estuary. [chap.
across the channel just above Lytham ; and from a mile below
Lytham, the channel had become worse than in 1883. The
depth of channel proposed in 1883 had not nearly been
obtained in the trained channel ; and in view of the expenditure
on the dock and improvement works having considerably ex¬
ceeded the estimated cost, a Commission appointed by the
Board of Trade advised, in 1891, that the dredging should be
limited to 5 feet less than contemplated in 1883, which is
being carried out1. This dredging, when completed, will
afford a depth of 23^ feet at high-water spring tides at
Preston, and 25! feet at Lytham, enabling vessels drawing
17 feet to navigate the river between Preston and Lytham at
spring tides. In order, however, to obtain a stable channel,
and the requisite depth of 26^ feet out to five miles beyond
Lytham, it is proposed that the training walls should be
extended out on each side of the channel to this distance,
which is about two miles short of the bar ; but the bar in
this instance marks the outer line of the sandbanks rather
than the chief shoal at the mouth of the Ribble, which is
found about i\ miles further in. It is, however, admitted
that changes in the estuary, resulting from the training works,
may necessitate the prolongation of the training works out to
the bar, to ensure a stable channel of the requisite depth.
At present the works are confined to the execution of the
north training wall above Lytham, authorized in 1883, which
has produced a marked improvement in the portion of the
channel along which it extends, and the deepening by dredging
indicated on the section2 (Plate 9, P"ig. 12). The dredging
contemplated in 1883 was commenced at the close of 1886,
since which period about 4,173,000 cubic yards have been
removed from the channel. The improvement in depth is
shown on the section ; and the low-water line has been
1 ' Final Report of the Ribble Navigation Commission,' 1891, p. 19.
2 Some particulars about the present state of the works in the Ribble were given
me by Mr. A. F. Fowler, resident engineer to the Ribble Navigation.
xii.] Results of Works in Ribble Estuary. 297
lowered about 6 feet at the entrance to Preston dock,
augmenting the tidal range at Preston by this amount, and
increasing the tidal volume in the channel ; whilst the arrival
of the flood tide at Preston has been advanced by an hour
since 1883.
Effects of the Training Works in Ribble Estuary. The
training walls in the estuary of the Ribble have formed a
permanent channel in place of a wandering one, and have
to some extent increased the depth in the trained channel
by scour, which has been further improved by dredging,
leading to a considerable improvement in the tidal condition
at Preston. The works have, however, led to the reclama¬
tion of the whole of the upper part of the estuary, down to
the Naze on the northern side, and down to the confluence
of the Douglas on the southern side, so that the estuary now
commences at the Naze instead of at Preston (Plate 9, Fig. 11 ).
The training of the river, moreover, has enabled the landowners
to carry out large reclamations in the southern part of the
estuary below the Douglas, which formerly were prevented
by the wanderings of the channel ; accretion is progressing
all along the southern side of the estuary ; and the banks
to the south of the training works are gradually rising. The
completion of the northern training wall also, between the
Naze and Lytham, will undoubtedly lead to the accretion of
the recess between these places, behind the wall.
Till the commencement of training works on the Ribble,
the South, or Boghole Channel was generally the main
outlet ; and even when cut off for a time from this outlet
by the changes in the channel through the estuary, the river
soon reverted to this southern outfall, which afforded the best
depth, and was the most favourable for the admission of the
flood tide. The direction, however, given to the trained
channel, as close as practicable to the northern shore, tended
to divert the river from its southern outlet; but even in 1850,
a narrow subsidiary channel of the river communicated with
298 Southport Channel injured by Works, [chap.
the South Channel. Soon after, however, the prolongation
of the training walls near the northern shore, the training of
the Douglas into this northern channel, and the resulting
accretion and reclamation on the southern side of the estuary,
diverted the river permanently from the Boghole. The in¬
fluence of these works was already apparent in the chart of
i860, which exhibits the channel of the river close along the
northern shore past Lytham. and then dividing into a northern
and central outlet, a continuous unbroken stretch of sand
between the southern shore and the channel in the estuary
above Crossens, and the South Channel cut off from the
main channel except at half tide, and connected only with
Crossens stream1. By degrees the effect of the training works
on the channel and the resulting accretion and reclamations
produced a raising of the intervening sandbanks, and caused
a considerable narrowing and curtailment of the South
Channel, and by 1883 had cut it off from Crossens stream
which afforded its sole remaining means of maintenance2.
The raising of the old training walls since 1883 and the
prolongation of the south training wall have still further
promoted accretion in the southern portion of the estuary,
and the raising of the banks between the trained channel and
the Boghole. Eventually, unless Crossens stream is brought
back into the Boghole, this South Channel, which afforded so
good an outlet for the Ribble in former times, must disappear,
and thus remove from Southport the only appearance of sea
which it possesses at low water.
Three effects of training works are clearly manifested by
the results of training the Ribble, namely, the improvement
and stability of the trained channel, the inefficacy of training
walls to improve the channel beyond their termination, and
1 ' Report on the Extension of the Navigable Channel of the Ribble to the Sea,'
L. F. Vernon-Harcourt, Southport, 1891, pp. 9 to 14.
- ' Report on the Crossens Channel New Cut,' L. F. \ ernon-Harcourt, South-
port, 18S4.
Defects of Ribble Training Works. 299
the occurrence of accretion in the portions of a sandy estuary
severed by the training walls from the main channel. The
South Channel has already been effectually cut off by the
training works from the Ribble, and is in course of silting up ;
and further extensions of the training works, needed for
providing a navigable outlet, will hasten this process. The
northern channel, which is far less stable than the South
Channel, has not as yet been affected by the training works ;
but it too will eventually be cut off permanently from the
Ribble by the extension of the north training wall past
Lytham, and will disappear.
Remark son Ribble Training Works. There were origin¬
ally three outlets available for the channel of the Ribble
through its estuary ; and though under ordinary conditions,
the direct central outlet would appear the most natural, a
study of the charts would have revealed that the southern
outlet afforded the most stable and deepest channel. The
indications, however, of nature were overlooked ; and the
channel was trained in a narrow channel near the northern
shore of the estuary. If Southport had been a seaport, its
interests could have hardly been so disregarded as in the
course chosen for the channel in 1853 ; and at that period
Southport had not developed into the important place it
has since become ; and its inhabitants were probably unaware
of the effect the training works proposed would have upon its
welfare. The extension of the training works since 1883 was
the natural sequence of the works of 1853 ; but this, and the
further prolongations proposed, if carried out, will by degrees
complete the deterioration of the South Channel. The
training of one channel in a sandy estuary subject to
accretion, necessarily involves the disappearance of the other
channels. Assuming that, in spite of the indications of the
early charts, the central channel was the most suitable for
Preston, it is evident that the trained channel should have
been widened out more seawards, even though this would
3°o Accretion in the Ribble Estuary. [chap.
have involved a larger amount of dredging ; for as the
accretions consequent on the training works progress, the
estuary will become very much restricted, and a wider
trained channel would have ensured a greater tidal scour
for maintaining the outlet The engineers, indeed, from the
outset deprecated reclamations, and merely aimed at im¬
proving the navigable channel : but unfortunately in an
estuary like the Ribble, with an abundance of sand outside,
it is impossible to carry out training works without promoting
accretion at the sides of the trained channel, which involves
the loss of tidal scour at the outlet and favours reclamation.
This inevitable result was undoubtedly hastened in the case
of the Ribble by constructing half-tide training walls, and
raising them by degrees ; but probably if the effect of training
works had been duly appreciated, a wider and more central
channel would have been designed, and dredging to provide
the necessary deepening would have been more largely
resorted to.
Lower and Outer Weser Training Works.
The river Weser drains an area of 18,600 square miles;
and its discharge varies between 196 cubic yards per second
at a low stage, and a maximum of 5,200 cubic yards during
a great flood, its average discharge being 387 cubic yards
per second. Formerly the tide did not flow higher up than
Bremen, 42 miles above Bremerhaven, which is situated at
the head of the actual estuary, where the rise of tide averages
joi feet1 (Plate 7, Figs. 3 and 4). About 10 miles below
Bremerhaven, the Weser estuary joins the estuaries of the
Jahde and the Elbe on either side, through which these rivers
flow in separate channels into the North Sea, which the Weser
reaches about 36 miles below Bremerhaven. At Heligoland
in the North Sea, beyond the combined estuaries of these
1 ' L'Amélioration des Fleuves dans leur partie maritime,'L. Franzius, Vme
Congrès International de Navigation intérieure, Paris, 1892.
xii.] Improvement of Weser by Training Works. 301
rivers, the rise of the tide is only 6 feet, so that the range of
tide is increased nearly 5 feet in flowing up to Bremerhaven.
Training Works in the Lower Weser. The river between
Bremen and Bremerhaven flows through flat alluvial land,
which may have originally formed an extension of the estuary.
The floods of the river and the changes of the channel have
been restricted by embankments on each side, at varying
distances back from the river. The strips of land, however,
thus enclosed with the river, though having about three times
the width of the river in some places, appear to serve only
as receptacles for floods, and are at too high a level to be of
value as reservoirs for tidal water to increase the tidal capacity
of the river. The training works, accordingly, along this
portion of the river, carried out since 1887, resemble the
regulation works on the Maas below Rotterdam and the
Nervion below Bilbao, rather than training works through an
estuary. They possess, however, a special intrest in systemati¬
cally training both a high-water and a low-water channel, by
double rows of fascine mattresses on each side, with enlarging
widths seawards carefully calculated to provide an adequate
channel for the full tidal flow as well as the fresh-water dis¬
charge (Plate 9, Figs. 8 and 10). The course of the river has
been straightened, bends have been eased, and secondary
channels suppressed, so as to facilitate the tidal flow and ebb,
and to concentrate the flow in a single channel ; whilst
dredging has been employed to increase the improved depth
obtained by scour (Plate 9, Fig. 9). The width of the low-
water channel has been adjusted, so that whilst large enough
to give free passage to the river from a little before to a little
after low water, it may not allow the low-water channel to
meander between the training walls. The material dredged
from the trained channel is deposited at the back of the
training walls, and is also used to fill up the abandoned
secondary channels. The works involve the removal of about
72 million cubic yards, of which it is estimated that qoi million
302 Training and Dredging in River Weser, [chap.
cubic yards will have to be dredged or excavated, the
remainder being removed by scour.
Training Works in Weser Estuary. The training works
were prolonged into the estuary below Bremerhaven in 1891,
where a submerged shoal exists in mid-channel in front of
the outlet, dividing the main channel into two branches.
A single training wall. 3J miles long, in continuation of the
left bank of the river, provides an enlarging outlet for the
river, which is bounded by the shore-line on the other side
(Plate 9, Fig. 8). This training wall skirts the edge of a
large sandbank, and is to be prolonged seawards to such an
extent as may be requisite to secure an adequate depth in
the outlet channel. This portion of the channel is also to be
dredged to the depth indicated on the section (Plate 9,
Fig. 9)-
Results of Weser Training Works. The training works
aided by dredging increased the available navigable depth
at high water up to Bremen, in the first four years, from 9 feet
to 14.2 feet ; and this depth will gradually be augmented to
about 22 feet by dredging (Plate 9, Fig. 9). The training
wall in the estuary, within six months after its completion,
had increased the average depth over the bar by if feet,
and 4J feet in the deepest channel.
The rectification and deepening of the channel has accele¬
rated the passage of the flood tide up the river, and has
extended the tidal limit from Bremen up to Flabenhausen,
4* miles higher up (Plate 7, Fig. 3). The low-water line had
already been lowered materially in 1891 between Bremerhaven
and Bremen, reaching a maximum of about 33 feet at Vegesack
(Plate 9, Fig. 9) ; and a further lowering may be confidently
anticipated from the proposed deepening by dredging. The
volume of tidal water passing Bremerhaven had been increased
by more than one-sixth in 189 t, and the efficiency of the ebb
correspondingly augmented.
Remarks on Weser Training Works. The training of
Effects of Weser Training Works. 303
a low-water, as well as a high-water channel, and the employ¬
ment of dredging to obtain the necessary depth, have enabled
any material contraction of the high-water channel to be
avoided. This maintenance of the width of the high-water
channel, combined with the lowering of the low-water line by
the improvement of the river, has produced an increase in the
tidal capacity above Bremerhaven, in spite of the shortening of
the channel by its rectification and the filling up of secondary
channels. The training works in the estuary below Bremer¬
haven have, accordingly, been commenced under the favourable
condition of an increased tidal flow. Moreover, the outlet is
being given a trumpet shape, so as not to impede the influx
of the flood tide up the river ; and as fairly deep water
approaches within 9 miles of Bremerhaven, it may be antici¬
pated that by a prolongation of the training wall in a
diverging line from the opposite shore, and the assistance of
dredging, a deep approach will be provided for the Weser,
and that sea-going vessels of large draught will have access
to Bremen which has already been converted by the training
works into a seaport.
CHAPTER XIII.
TRAINING WORKS IN ESTUARIES (continued).
Physical Conditions of the Loire and the Seine contrasted. Loire Estuary
Works: Description of Estuary; Training Works below Nantes, their extension
and results; Ship-Canal formed alongside part of Estuary to provide a better
navigable depth ; Dredging in Estuary, and in Trained Channel ; Improvement of
Navigable Depth up to Nantes ; Remarks on Loire Training Works, difficulties
involved. Seine Estuary Works : Original state of Tidal Seine ; Training Works
in Estuary, their extent and consolidation ; Effect of Training Works on Navigable
Channel ; Accretions in Seine Estuary, their origin, extent, and volume ; Alter¬
ations in Lines of Depths at Outlet; Tancarville Canal, description and object of
Canal ; Schemes for Completion of Seine Training Works, difficulties, variety of
proposals, their aims and merits ; Remarks on Seine Training Works, value for
Rouen, cause of anxiety for Havre ; Importance of Completion of Works, in
trumpet-shaped form aided by dredging. Remarks on Training Works : General
Conclusions deduced from Experience ; Enlargement of Trained Channel, indicated
by instances. Concluding Remarks.
Two rivers of France, the Loire and the Seine, presenting
considerable differences in their physical conditions, have
been partially trained through their extensive estuaries, for the
purpose of improving the navigable access to the ports of
Nantes and Rouen respectively. The Loire is a torrential
river, and brings down large quantities of detritus ; whilst
the Seine is comparatively gently-flowing, and its fresh-water
discharge is remarkably free from alluvium. The estuary
of the Loire, moreover, is somewhat narrow, and has a con¬
tracted outlet ; whilst the estuary of the Seine is much wider,
and has a large outlet ; but though free from the islands
which are found in the upper estuary of the Loire, the Seine
estuary is much more encumbered by sandbanks (Plate 9,
Figs, i and 4). The Loire drains an area of 45,000 square
miles about half as large again as the Seine basin with an area
Conditions of Loire and Seine contrasted. 305
of 30,370 square miles ; but the tidal rise at its mouth is
only i6| feet at springs, as compared with 33^ feet at the
mouth of the Seine ; whilst, owing to this smaller tidal range,
in conjunction with the greater slope of the country through
which it flows, the Loire is tidal for only about 44 miles
from the sea, whereas the Seine is tidal up to Martot, about
90 miles inland.
The discharge of the Loire is very irregular, falling to
a minimum of 160 cubic yards per second in a very dry
season, and attaining 7,600 cubic yards per second in excep¬
tional floods, or more than double the maximum discharge
of the Seine. This small summer flow renders the Loire
very shallow in dry weather, and therefore unsuitable for
regular navigation ; and the traffic on the river above Nantes
is very small. The Seine, on the contrary, has a much more
regular flow ; and its small fall between its tidal limit and
Paris, has rendered it capable of being greatly improved in
navigable depth by canalization. Rouen, accordingly, being on
the way to Paris, and connected with it by a very accessible
waterway, serves as the port of transit between the sea and
an important inland waterway leading direct to the capital ;
and its prosperity depends on its access to the sea. Nantes
being devoid of inland communication by water, and not
being on the route to any important town, is a port of
considerably minor importance ; and its trade is due to its
being near the head of the tidal Loire, and depends on its
connexion with the sea. Both these river ports, however,
are somewhat overshadowed by the sea-ports of Havre and
St. Nazaire. which are situated at the entrance of the Seine
and of the Loire respectively. The commercial value, indeed,
of an inland position at the head of a tidal navigation is
shown by these inland ports not having been eclipsed by
their formidable rivals on the sea-coast, and by their trade
flourishing in proportion to the improvement of their com¬
munication with the sea.
X
3o6 Loire Estuary described : Early Works, [chap.
Loire Estuary Works.
Nantes is situated near the head of the Loire estuary,
about 35 miles above its outlet, and 9^ miles below Mauves,
the limit of the tidal flow at a low stage of the river. The
tidal rise of j6| feet at the mouth of the estuary opposite
St. Nazaire, is reduced to 6 feet at Nantes (Plate 9, Figs. 4
and 5) ; but during ordinary floods the tide is not felt within
12 miles of Nantes, being overpowered by the flood dis¬
charge; and in very high floods the flood-tide current is
imperceptible, except near the bed of the river at St. Nazaire.
The estuary between Nantes and La Martinière consisted
of a network of channels flowing between numerous islands,
probably a gradual growth from the ordinary state of an
estuary encumbered with sandbanks, resulting mainly from
the deposit of material brought down by the river. Islands
are also found in the more open estuary below La Martinière,
and nearly down to Paimbceuf ; but the alluvium in this part
of the estuary above low water was formerly of marine
origin.
Training Works in Upper Estuary of the Loire. Proposals
to improve the access to Nantes were made in the seventeenth
century; but the first works were carried out in 1756-68, at
which period vessels of 8 feet draught had difficulty in getting"
up to Nantesh These works consisted in barring the secondary
channels between the numerous islands and the left bank of
the estuary, by means of submerged rubble-stone dams, from
Bouguenais down to Couëron, so as to concentrate the
main current into the channel bordering the right bank, with
the view of increasing its depth by scour2. Groynes also
were carried out from the right bank into the channel, in
order to contract the waterway still more, and thereby
1 'Annales des Ponts et Chaussées, 1878 (2), p. 563.
2 4 Ports Maritimes de la France,' vol. v, p. 257.
xiii.] Training Works in Loire Estuary. 307
increase the scour. Considerable accretions, however, resulted
from these works ; and no material improvement in the depth
of the navigable channel was effected.
No further works were undertaken till 1834, when longi¬
tudinal training walls, raised to a level of 3* feet below high-
water spring tides, were constructed in different places between
Nantes and Le Pellerin. mainly on one side of the channel,
so as to narrow it to a width of 820 to 1,000 feet. Even these
works, however, completed in 1840, produced merely a slight
improvement in the channel, for in 1850 the depth over the
principal shoals had only been increased between 8 inches
and i foot.
In 1859-65 training works were again carried out along
the ten miles between Nantes and La Martinière. Rubble-
stone training walls, pitched on their exposed faces, were
formed along each side of the channel, with their tops at the
level of high water of low springs, placed 650 feet apart near
Nantes, and widening out gradually to 1,000 feet at Le
Pellerin (Plate 9, Figs. 4 and 7). A gradual increase in
depth was effected by these works, amounting in 1876 to
an average of 5 feet over the whole distance, and 3* feet
across the worst shoals, giving an average depth in the
navigable channel of 20 feet, and a minimum of 14I feet at
high-water spring tidesl. The shoals were found in places
where the old training walls had not been reconstructed, or
where side channels were left open for the convenience of
riparian owners ; and they might have been removed by the
completion of the training walls where the old walls were
left unaltered, and across the open gaps.
The improvement in depth of the trained channel was accom¬
panied by a large extension of the sandbanks in the estuary
below, between La Martinière and Paimbceuf, so that this
1 A plan and sections of the Loire were furnished me by Mr. Lefort, engineer-in-
chief of the tidal Loire, and also a summary of the works carried out, and their
results.
X 2
3o8 Shoaling in Loire Estuary : Ship-Canal. [chap.
section of the channel, which previously was deeper than the
portion above, became worse. This deterioration was due to
the deposit in this part of the material scoured out of the
trained channel, to the loss of tidal scour by the large
accretions which had taken place at the back of the training
walls, and to the direct discharge of the detritus of the river
into this section of the estuary. The mean depth of the
navigable channel through this part of the estuary was
reduced from 2of feet to 18 feet between 1859 and 1876;
and the length of the shoals with a less depth than 17^ feet
over them, was doubled in the same period ; whilst fresh
islands have been formed, or are in course of formation.
Ship-Canal alongside the Loire Estuary. In order to
provide the requisite depth in the deteriorated portion of the
estuary below La Martinière, the prolongation of the training
walls was proposed; but eventually, in 1879, the alternative
scheme of a lateral ship-canal was adopted, so as to avoid
the shoaling circuitous channel through the estuary below
the existing training works. This ship-canal has been con¬
structed through the land on the left bank, alongside the
estuary, starting at La Martinière from the trained channel,
and after a course of 9-J miles, rejoining the estuary at Carnet,
4! miles above Paimbceuf, where there is a deep channel close
to the shore, passing between Carnet Island and the left
bank of the estuary (Plate 9, Fig. 4). The canal was com¬
pleted in 1892 ; it has a bottom width of 78® feet, and a
depth of water of 19! feet; and it has a lock at each end,
558 feet long and 59 feet wide.
Dredging in the Loire Estuary. Since 1877, dredging
has been carried out in the estuary on a much larger scale
than previously; and up to the end of 1892 it was mainly
executed in the shallow part of the estuary, between La
Martinière and Paimbceuf, increasing the available depth in
the navigable channel through this section of the estuary
from 13 feet to 15! feet at high-water spring tides. The
xiii.] Dredging in Loire Estuary. 309
trained channel between Nantes and the entrance to the
ship-canal, is being deepened by dredging to a depth of
21 feet below high-water spring tides : and the channel
between the outlet of the canal and deep water in the estuary
is being deepened to 24 feet. Dredging will still be carried
on to a moderate extent in the portion of the estuary super¬
seded by the ship-canal for large vessels, in order that the
flood tide may not be impeded in its passage up the estuary
by shoaling in the channel, and that small vessels may continue
to go up to Nantes without being obliged to pass through
the ship-canal.
The ship-canal, and the increase in depth above and below
effected by dredging, will enable vessels drawing i6J feet to
reach Nantes from the sea in a single tide, even during the
lowest neaps, which formerly was only practicable on at
most fifty days in the year; and vessels of 19 feet draught
will be able to accomplish the passage during spring tides
and in flood-time.
Remarks on Loire Training Works. The estuary of the
Loire is not favourably circumstanced for improvement by
training works, on account of the large quantities of sand
brought down by the floods of the river, which, together
with alluvium brought in by the sea, readily deposit at
slack tide in sheltered parts of the estuary. The training
walls caused the river detritus to be discharged into the
portion of the estuary beyond their termination, and at the
same time reduced the tidal scour in this part by the diminu¬
tion of tidal capacity produced by accretion in the estuary
behind the walls. Accordingly, the training works terminating
at La Martinière merely shifted the position of the shallowest
channel lower down the estuary ; and the deteriorated portion
of the estuary, between La Martinière and Paimbceuf, has
ceased to be used as the navigable channel since the construc¬
tion of the lateral ship-canal. In spite, however, of its abandon¬
ment for navigation, this portion of the estuary will have to be
3io Considerations affecting Loire Works, [chap.
maintained by dredging, lest a reduction in tidal scour should
injure the lower estuary and the approach to St. Nazaire.
The enlargement in width between the existing training
walls as they descend towards the sea, has not been made
adequate, owing doubtless to the desire to effect the requisite
deepening of the channel by scour. Moreover, in order to
provide a navigable channel of sufficient depth through the
estuary between Nantes and the sea, it would have been
essential to continue the training walls down to Paimbceuf,
gradually enlarging the trained channel to the full width
of the narrowed estuary, and securing the required depth
by dredging. Nevertheless, a trained channel, however well
designed, would inevitably restrict the reservoir available for
the materials brought down by the river, and reduce the
tidal capacity of the estuary, and consequently diminish the
tidal scour past St. Nazaire and over the bar outside. Nantes
would have benefited by such works, for the depth in its
improved approach channel is less than the available depth
over the bar. With a large port, however, at the outlet of
the estuary like St. Nazaire, it would be inexpedient to carry
out works in the estuary which would reduce the tidal
capacity of the estuary, which serves as a sort of natural
scouring basin for maintaining the depth of the channel over
the bar, upon which a deep approach channel to St. Nazaire
depends.
Seine Estuary Works.
The river Seine flows from Rouen down to La Mailleraye,
a distance of 39 miles, in a stable bed, through rock, clay,
and gravel ; and throughout this portion of its course, which
is about half the distance between Rouen and the sea, it
possesses a good natural depth, so that by merely dredging
a few shoals near Rouen, a depth of 24 feet at high-water
neap tides has been provided. Below La Mailleraye, however,
where the estuary may be said to have commenced, the river
xi«.] State of Seine Estuary : Training Works. 311
formerly flowed to the sea in a shallow, winding, and variable
channel through shifting sandbanks. In some places within
twelve miles of La Mailleraye, the shoals were so high that
the available depth over them at high-water spring tides was
only 10 feet ; vessels exceeding 200 tons could not ascend the
river; and vessels of between 100 and 200 tons often had to
wait for a high enough tide. The voyage between the sea and
Rouen never occupied less than four days ; and the perils of
navigating the shallow, shifting channel were aggravated by the
bore, which travelled up the river with considerable velocity,
as a breaking wave some feet in height, at the commencement
of the flood-tide during springs ; and consequently wrecks
frequently occurred. Various schemes were proposed during
the eighteenth century for remedying this disastrous condition
of the estuary, as for instance a ship-canal along the left bank
down to Honfleur in place of the estuary channel, and the
improvement of the channel through the estuary by weirs,
groynes, or training works of various kinds1.
Training Works in Seine Estuary. The first works under¬
taken for the improvement of the navigable channel through
the estuary were commenced in 1848, by the construction of
training walls from Villequier, 7 miles below La Mailleraye,
down to Quillebeuf, a distance of about 11 miles. These train¬
ing works, completed in 1851, formed an experimental portion
of a general scheme, proposed in 1845 by Mr. Bouniceau, for
training the channel from La Mailleraye to Honfleur, and
across the estuary to Havre2. As the depth over the shoals
was rapidly increased by these works from iof feet to 211 feet,
it was determined to extend them up to La Mailleraye, and
down to Tancarville on the northern side, and to La Roque
on the southern side, which extensions were completed in
1859 (Plate 9, Fig. 1). The training walls were placed
1 ( The River Seine,' L. F. Vernon-Harcourt, Minutes of Proceedings Institution
C. E., vol. lxxxiv, pp. 241-252, and plate 4.
2 ' Etude sur la navigation des rivières à marées,' P. Bouniceau, [845, p. 152.
312 Description of Seine Training Works, [chap.
980 feet apart at Villequier. widening out to 1,310 feet at
Quillebeuf, and 1,640 feet at Tancarville; and they provided
a deep stable channel down to Tancarville. Below this point,
however, the channel diverged northwards, and was shallow,
circuitous, and changeable, so that a prolongation of the walls
down to Berville was carried out, which was completed in
1869. A further extension of the works, proposed in 1871,
was relinquished in consequence of the opposition of Havre,
which had become alarmed by the changes in the estuary
produced by the training works already carried out ; so that
the training walls have been stopped in the middle of the
estuary, at the point they had reached at that period.
The training walls, originally formed of mounds of rubble
chalk obtained from the neighbouring cliffs adjoining the
river in places, were raised to high-water level down to
Tancarville on the north, and to La Roque on the south ; but
below these points, they were only carried up two or three
feet above low-water spring tides, with the object of avoiding
as far as possible further accretions in the estuary. The
high training walls, being eroded by the bore, and under¬
mined by the deepening of the channel, were disturbed, and
settled ; and after being raised by additions of rubble chalk
for several years, they were eventually protected against the
wash of the bore and undermining, by a concrete facing and
apron, and by piling and planking along the toe (Plate 9,
Fig- 3)- The low training walls, moreover, below Tancarville
and La Roque were so much injured by the rush of the tide
across them at certain periods, that they also have had to
be reconstructed and raised to mean high-water level.
Effects of Training Works on Channel. A comparison of
the longitudinal section of the river in 1824 with the sections
along the trained channel in 1880 and 1891, shows the
remarkable increase in depth that has been effected by the
training walls between La Mailleraye and Berville (Plate 9,
Fig. 2). The deepening averages about 15 feet, the minimum
xiii.] Effects of Seine Training Works. 313
reaching 6 feet ; and the minimum navigable depth between
Havre and Rouen at high-water neap tides has been increased
from 10 feet up to 18 feet. The low-water line also has been
lowered throughout the trained channel, and above, by the
increased depth, the lowering reaching a maximum of about
8 feet at Tancarville. The improvement was almost wholly
due to the training walls, for the only dredging effected in
the trained channel was the lowering of the hard Meules
bank about a mile and a half below La Mailleraye. The
arrival of the flood tide at Rouen has been advanced one hour
by the removal of the obstacles to its progress above Berville ;
but the bore is still experienced along the comparatively
narrow trained channel, reaching a maximum at Caudebec,
and disappearing at Rouen (Plate 7, Fig. 1).
The influence of the training walls on the channel ex¬
tends very little beyond their extremities. Fortunately as
regards the depth, the worst shoals were above Berville ; but
on emerging from between the training walls, the channel
remains circuitous and variable along the 11 miles of sand¬
banks separating Berville from deep water outside the
estuary. The training works, however, have brought deep
water in the river much nearer the open sea, so that the
intervening length of shallow channel is readily traversed
at high water in fine weather. Moreover, owing to the
greater range of tide towards the mouth of the estuary, the
depth near high water in the main channel approximates to
the available depth in the trained channel ; so that little
inconvenience would be experienced by the absence of an
improved depth between Berville and the sea, if the channel
was stable and direct. Though, however, the winding and
shifting channel is marked by buoys, it is difficult for vessels
of large draught to navigate the intricate channel during
storms, when alterations are most liable to occur, and in fogs.
Accordingly, in order to render the passage through the
estuary always practicable and safe at high tide, for the
3T4 . Anticipated Accretion in Seine Estuary, [chap.
sake of Rouen which has been raised by the training works to
the position of the fifth seaport of France, it will be necessary
to prolong the training walls.
Accretions in the Seine Estuary. About half of the deeply
recessed portion of the estuary lying between Quillebeuf and
La Roque, comprising mainly the Vernier marsh, had been
reclaimed previously to the commencement of the training
works ; and a smaller portion of the foreshore of the recess
on the northern side above Tancarville had also been enclosed.
The rest of the estuary, however, was covered by the tide,
which came up close to the base of the chalk cliffs which
bounded the estuary on the north between Tancarville and
Harfleur, and on the south between La Roque and Honfleur
(Plate 9, F"ig. i). In proposing his scheme of training works,
Mr. Bouniceau expressed the opinion that the accretion that
might result from them would be exceedingly slow in ac¬
cumulating, and that the filling up of the estuary behind the
training walls would extend over what might be regarded as a
geological period, amounting to 210 centuries. He considered
also that the deposit which might settle behind the training
walls would be derived in great measure from the sandbanks
further down the estuary, which would thereby be lowered,
constituting in fact an advantageous transposition of material.
Owing to the freedom of the Seine waters from sediment,
the estuary is practically in no danger of accretion from this
source. The flood tide, on the contrary, enters the estuary
laden with material eroded by waves from the Calvados cliffs,
and collected by the tide in passing along this extensive
stretch of coast, especially when storms stir up the alluvium
composing the foreshore. Formerly, however, the alluvium
brought in by the flood was carried out by the ebb, especially
when the banks were periodically stirred up by the shiftings
of the channel. The sands encumbering the estuary were
doubtless originally derived from the Calvados coast ; whilst
the cliffs to the north have supplied the shingle beach in front
xiii.] Accretion due to Seine Training Works. 315
of Havre, which extends inside the estuary to Hoc Point.
No records are available to indicate whether accretion was
taking place in the untrained estuary ; but it is clear that the
process was slow, except in sheltered recesses. As soon,
however, as the existing state of equilibrium of the estuary
was disturbed by the construction of training walls, which
concentrated the main tidal flow and ebb and the whole of
the fresh-water discharge into the trained channel, the settle¬
ment of alluvium brought in by the flood tide was promoted
in the slack water, under shelter of the training walls and
out of the run of the currents ; and accretion rapidly took
place. The silting up of the large recesses above La Roque,
already partially reclaimed, was undoubtedly accelerated by
bringing the training walls close against the projecting points
at Quillebeuf and Tancarville, and constructing a cross wall
from La Roque point to the south wall, thereby enclosing
these areas which have since been reclaimed (Plate 9, Fig. 1).
The accretion, however, resulting from the training works has
not been confined to these enclosed portions of the estuary,
nor even to the sides of the estuary at the back of the training
walls, but has crept down along the sides of the estuary to
Honfleur on the south side, and to Harfleur on the north side,
ó} and 9 miles respectively below the ends of the training
walls. Accordingly, though the training works lose their
effect in deepening and fixing the channel close to their
extremities, their influence on the accretion of the estuary
extends several miles lower down.
A survey of the estuary made in 1875 showed that within
thirty years of the commencement of the training walls, after
making allowance for the increase in depth of the trained
channel, the capacity of the estuary at high tide had been
diminished by 27a million cubic yards, owing to the accretion
resulting from the works1. The rate of accretion would
1 ' Reconnaissance hydrographique de 1875 à l'Embouchure de la Seine.' Report
by X. Estignard, 1877.
3i6 Progress of Accretion in Seine Estuary. [chap.
naturally attain a maximum about the period of the comple¬
tion of the training works ; and subsequently the estuary
would gradually approach again a position of equilibrium,
as soon as the accretion induced by the works has taken
place. Another survey made in 1880 showed that in the five
years that had elapsed since the previous survey, a further
reduction in the capacity of the estuary had occurred, to the
extent of 40 million cubic yards, raising the total accretion
in the estuary since the commencement of the works to
312 million cubic yards, more than a third of the amount
to the accumulation of which Mr. Bouniceau had allotted an
almost indefinite period. Accretion was evidently still pro¬
gressing in the estuary in 1880, though at a diminished rate.
Without a fresh survey, it would be impossible to determine
the amount of the subsequent accretion ; but a comparison of
the low-water charts of 18801 and 1891 (Plate 9, Fig. 1) shows
that the area of marsh lands has somewhat increased on the
north side, that the area of sandbanks above low water has
become greater, and that more sandbanks emerge near the
outlet. It appears therefore that the accretions due to the
training works have not yet reached their limit. The time of
high water at Havre has advanced about forty minutes since
the construction of the training works, which is evidently due
to the quicker filling of the estuary on account of its greatly
reduced tidal capacity.
The idea entertained by Mr. Bouniceau that the material
deposited at the back of the training walls would come from
the sandbanks at the mouth of the estuary, was at first so far
realized that the 5- and 10-metre lines of soundings advanced
into the estuary, after the commencement of the works, up to
1875. Since that period, however, they have been receding;
and in the chart of 1891, they were 3 and 3^ miles seaward of
their positions in 1875, and further out than in any previous
chart back to 1834. This advance and retreat of the lines of
1 Minutes of Proceedings Institution C. E.? vol. lxxxiv, plate 4, fig. 1.
xiii.] Reduction of Depth : Tancarville Canal. 317
soundings is readily explained. The sand brought up from
the mouth, as well as from outside, by the flood tide, found
a resting-place in the sheltered parts of the estuary during
the extension of the training walls, and for some time after
their completion, and therefore did not return on the ebb to
be deposited again ; and consequently the depth at the outlet
improved. As soon, however, as the main part of the accre¬
tions resulting from the training walls had taken place, the
reservoir for the sand brought up by the flood tide having
been filled up, the sand returned with the ebb tide ; and the
scour of the ebb having been greatly reduced by the diminution
in the tidal capacity of the estuary, the sand was more readily
deposited, and reduced the depth at the outlet. This will con¬
tinue till the cessation of accretion, and the reduction in the
section of the outlet channel, corresponding to the diminished
tidal scour, produce a fresh state of equilibrium in the estuary.
Tancarville Canal alongside Seine Estuary. The idea of
avoiding the dangers of the passage of the estuary, by con¬
necting Havre with the river above by a canal, appears to have
originated in the eighteenth century; and various projects have
been brought forward from time to time. At length in 1876,
at the request of the authorities of Havre, the question was
investigated ; the construction of a canal connecting the docks
at Havre with the river Seine at Tancarville was authorized
in 1880 ; and the canal was opened for traffic in 1887 l. The
canal which traverses the alluvial deposits between Harfleur
and Tancarville, has a lock at each end, to regulate the water-
level which is kept below the highest tides, and an enlarge¬
ment in width near each end to serve as a siding (Plate 9,
Fig. 1). The present object of the canal is to enable river
craft to pass between Havre and the river without traversing
the estuary ; and therefore the canal has only been excavated
to a depth of 11^ feet between Tancarville and Harfleur.
The locks, however, with an entrance width of 52í feet, have
1 ' Canal du Havre à Tancarville,' Maurice Widmer, Havre, 1887.
318 Tancarville Canal : Estuary Works, [chap.
had their sills placed low enough to afford a depth of water
over them of 23 feet ; and their width, and the width of the
canal have been made of such a size that it will be only
necessary to deepen the canal to convert it into a ship-canal,
capable of accommodating sea-going vessels, in the event of
the estuary becoming so much deteriorated by accretion as to
be no longer suitable for navigation. As the communication
of Harfleur with the sea has been cut off by the accretions in
front of it, and by the canal, a branch canal has been carried
up to Harfleur ; and the depth of the waterway between
Harfleur and Havre has been made 19! feet (Plate 11,
Fig. 22). The canal is 15i miles long; and its cost, including
a basin at Havre, quays, nine swing-bridges across the locks
and for roads, and land, amounted to .£928,000. The traffic
through the canal reached 287,500 tons in 1893.
Schemes for Prolongation of Seine Training Works. The
training walls which were stopped at Berville in 1869, form
part of an incomplete scheme ; whilst the provision made for
enabling the Tancarville Canal to be readily converted into
a ship-canal, indicates the uncertainty which exists in some
quarters as to the maintenance of the navigable capabilities
of the estuary. It would be impossible to obtain a satisfactory
and stable navigable channel between Berville and the sea
without a prolongation of the training walls ; whilst any
extension of the training walls will introduce a fresh dis¬
turbance in the state of equilibrium to which the estuary is
at present approaching. Moreover, any scheme must secure
Honfleur its access to the sea, without in any way endanger¬
ing the maintenance of the depth of the approach channel to
Havre. The existence of ports on opposite sides of the
estuary occasions a difficulty in forming a trained channel
convenient for both ; and it would be quite inadmissible to
prolong the trained channel to Honfleur, and then to carry it
across the estuary to Havre, as indicated in one of the original
schemes, for a channel trained across the estuary could not be
xiii.] Schemes for Completion of Training of Seine. 319
maintained with low training walls, and with high training
walls would lead to the immediate reclamation of the estuary.
Various schemes have been proposed for the completion
of the Seine training works, presenting a great diversity in
design1 ; but they may be grouped under three distinct
classes. One set of schemes proposes to close three-fourths
of the outlet of the estuary, by a breakwater extending
from Villerville to the Amfard bank ; and the prolongation of
the south training wall in a continuous curve past Honfleur
towards Havre, would guide the channel to the narrow outlet
(Plate 10, Figs, i and a). In a second series of designs, the
channel would be trained in a sinuous course between two
training walls, widening out seawards, the southern wall
terminating at Honfleur, and the northern wall being either
stopped in mid-estuary or prolonged to Havre (Plate 10, Figs.
4 and 5). The third class of schemes would train the channel
in an expanding trumpet-shaped form to the outlet (Plate 10,
Figs. 3 and 6). The first class of designs would convert the
estuary into a huge sluicing basin, and thereby secure a deep
channel through the narrow outlet : the second aims at
forming a deep and stable channel by a succession of curves ;
and the third would endeavour to guide the channel in a
central course, whilst freely admitting the flood tide, and
reducing the capacity of the estuary as little as practicable
consistently with the guidance of the channel. A great
reduction in the width of the outlet, whilst improving the
depth in the narrowed channel, would check the influx of
the tide, and consequently reduce the tidal rise in the enclosed
estuary, involving a diminution in tidal capacity, and a loss
of depth in the navigable channel at high water. A sinuous
course, though producing a fairly stable channel, especially at
the concave bends, in a river where the flow is always in one
direction, does not afford a similar advantage in a tidal river,
owing to the divergence in the courses of the flood and ebb
1 Minutes of Proceedings Institution C. E., vol. lxxxiv, plate 4, fig. 9.
32°
Results of Seine Training Works.
tides in a wide curving channel ; and any scheme where
a training wall terminates in mid-estuary must be regarded
as incomplete (Plate 10, Fig. 5). A trumpet-shaped outlet
for the trained channel is the only form which follows the
indications of nature, though some uncertainty exists as to
the best rate of enlargement, and some difficulty would be
experienced in combining a wide outlet with an adequate
deepening of the channel. The rate of enlargement, however,
is capable of calculation for any special case ; and deepening
by dredging might be advantageously resorted to within the
trained channel, aided possibly by inner low-water training
walls or dipping cross dykes, in order to avoid an undue
restriction of the estuary.
Remarks on Seine Training Works. The training works
in the Seine estuary are remarkable for the great improve¬
ment in depth effected by them as far as they have been
carried, the large and unexpected accretions which they have
occasioned, as well as the distance down the estuary to which
these accretions extend, and the great diversity of opinions
and schemes to which the proposed completion of the training
works has given rise. Rouen has gained enormously by the
great increase in depth effected in the worst part of its
navigable access to the sea, lying between La Mailleraye and
Berville ; though the portion of the channel between Berville
and the sea still needs amelioration to satisfy the requirements
of this flourishing inland port. Anxiety has, at the same
time, naturally been aroused about the maintenance of the
deep-water access to Havre, in consequence of the large
deposits which have accumulated in the estuary, creeping
down to the neighbourhood of Havre ; and the retrogression
of the lines of soundings seawards since 1875 is not calculated
to allay these fears. It has, accordingly, been proposed that
the completion of the training works should be accompanied
by the formation of a large outer harbour on the sea-coast in
front of Havre, so as to provide the port with a new deep
xiii.] Extension and Expansion of Seine Works. 321
sheltered outlet beyond the influence of the deposits in the
Seine estuary.
The completion of the training works should not be longer
delayed in the interests of Rouen and Honfleur, which at
present are exposed to the inconveniences of a constantly
shifting outlet channel. The training walls have hitherto
been placed too close together to ensure adequately the free
admission of the flood tide, and to provide for the due
enlargement of the outlet. The trained channel was made
narrow in order to increase the deepening by scour ; but it
would have been much better to give the channel a greater
rate of enlargement, and to have gained the requisite depth
by the aid of dredging as well as scour. The narrowness of
the existing trained channel has proved a hindrance in the
designs for the completion of the works ; and if possible
the widening out should be commenced at Quillebeuf, so
as to render a suitable expansion to a wide outlet more
practicable ; for the northern training wall must not impede
at all the access to Havre, and the southern training wall
should not project from the coast so as to cause an advance
of the foreshore in front of Trouville (Plate 10, Fig. 1).
A trumpet-shaped outlet would offer the least interference
with the existing condition of the estuary and the depths
outside, and the best prospect of obtaining a fairly stable
central channel, which could be deepened and fixed by
dredging and low inner training works. Moreover, by widen¬
ing the trained channel above Berville, and regulating and
deepening the outlet channel below, the influx of the flood
tide would be facilitated, and the bore consequently reduced.
Even under existing conditions, vessels entering the estuary
a little before high water can almost always get up to Rouen
in a single tide, owing to the long duration of high water ;
but the removal of some shoals in the trained channel would
render the passage up the river still easier. The descent of
the river from Rouen to the sea can only be effected in two,
Y
322 Influences of Tranting Works. [chap.
or occasionally in three, tides ; and the vessels have to
anchor in one of the deep pools on their route during low
tide. The lowering of the shoals, however, in the trained
channel, as proposed, will enable vessels drawing less than
16\ feet to get down without a stop ; and vessels of larger
draught will never have to anchor more than once on the
journey1.
Remarks on Training Works.
The training works through estuaries described in this and
the preceding chapter, prove that whilst their influence in
deepening a channel extends hardly at all beyond their
termination, their effects in producing accretion at the sides
of an estuary may be apparent a long distance in advance of
the trained channel. These results show that, in order to
secure a deep stable outlet channel, training works must be
carried out to deep water, and that their influence in causing
accretion in an estuary must not be lost sight of. The
relative advantages of an improved channel, and the main¬
tenance of the tidal capacity of an estuary, must be duly
weighed ; and the decision must depend upon the conditions
of the case. When the maintenance of the access to a large
seaport at the mouth of an estuary depends upon the pre¬
servation of the tidal capacity of the estuary above it, no
training works should be permitted in the upper estuary ;
and therefore training walls in the upper estuary of the Mersey
were inadmissible in the interests of Liverpool. For the same
reason it would have been inexpedient to carry training
works down the Loire estuary, for in benefiting Nantes,
St. Nazaire would have been injured. On the other hand,
improved outlet channels for the ports in the estuary of the
Wash are of more importance than the gradual silting up of
the estuary. In the case of the Seine estuary, the claims
1 ' La Seine Maritime,' P. Mengin-Lecreulx, Vrae Congrès International de Navi¬
gation intérieure, Paris, 1892, p. 14.
xiii.] Prejudicial Effects of Training Works. 323
of Havre are paramount ; but with this form of estuary, and
as Havre is on the open sea-coast, it appears possible, by very
carefully designed works, to provide an adequate navigable
outlet for Rouen without endangering the approaches to
Havre. A more difficult question to decide is how far the
improvement of a port up an estuary should override
the interests of a seaside town at the outlet, by allowing
works to be carried out which would prejudice such towns
by producing a progression of the foreshore in front of them,
and thus driving the sea further away. Some of the schemes
for the prolongation of the Seine training works would
produce this effect in front of Trouville (Plate 10, Figs. 1, 2,
and 3) ; and the training walls in the Ribble estuary threaten
Southport with a similar fate. Such considerations have
hitherto been disregarded, because the wide-spread influence
of training walls in a sandy estuary have not been under¬
stood ; but in future, questions of this nature can hardly be
neglected.
Enlargement of Trained Channel. Though the rate of
enlargement in width between training walls must depend
upon the special conditions of an estuary, such as the tidal rise,
fresh-water discharge, slope of river-bed, and size of outlet,
some indications of the limits of this enlargement may be
obtained from estuaries possessing good outlet channels. The
enlargement in untrained estuaries is necessarily irregular ;
but in those estuaries which possess the most uniform navig¬
able channels, the rate of enlargement increases rapidly on
approaching the mouth, and is greater for the high-water
channel than for the low-water channel. The increase in
width of the Scheldt, between Antwerp and Lillo, is about
i in 59 in the low-water channel, and 1 in 50 in the high-
water channel ; whilst between Lillo and Flushing, where
both depths and widths vary considerably, it amounts to
about 1 in 2i at low water, and 1 in 20 at high water. In
the Thames, the rate of enlargement between London Bridge
Y 2
324 Rate of Enlargement of Tidal Channels, [chap.
and Gravesend is about 1 in 93 at low water, and 1 in 80 at
high water ; whilst between Gravesend and Canvey Island it
attains about 1 in 30 at low water, and 1 in 8 j at high water;
and beyond this latter point both channels widen out rapidly
(Plate 7, Fig. 7). In the Humber, the rate of enlargement,
from the confluence of the Ouse and the Trent to Sunk
Island, is 1 in 48 at low water, and 1 in 37 at high water ;
whilst from Sunk Island to the mouth at Spurn Point the
enlargement is rapid 1 (Plate 7, Fig. 9).
The rate of enlargement adopted for trained channels varies
considerably. Thus the widening out of the Seine training
walls is only 1 in 200, and of the Vire 1 in J 00; whilst the
trained channel of the Ribble has been made very nearly
uniform between Preston dock and the Naze, though below
the Naze to the end of the north training wall the enlargement
is about i in 80. The enlargement of the trained channel
of the Maas, from Schiedam to the mouth, averages about
] in 77 ; of the Weser from Bremen to the end of the training
walls, i in 50 ; of the Clyde from Renfrew to Dumbarton,
1 in 71 ; and of the Tyne from Newcastle to its mouth, 1 in
74. The increase in width between the Seine training walls
is acknowledged to be much too small ; and probably in
most cases the suitable rate of enlargement for a trained
channel would be found to be comprised between 1 in 80
and i in 50, with a more rapid expansion near the outlet.
In some cases of training rivers through wide estuaries, too
much attention has been devoted to the deepening of the
trained channel by scour, and too little regard paid to the
maintenance of the tidal capacity of the estuary, leading
to an undue narrowing of the width between the training
walls ; so that in some instances a narrow trained channel
emerges abruptly into a wide estuary, as exemplified by the
Seine, the Loire, the Ribble, and the Dee (Plate 9, Figs. 1,
1 'The Physical Conditions affecting Tidal Rivers; and the Principles applicable
to their Improvement,' L. F. Vernon-Harcourt, Manchester Congress, 1890, p. 10.
xiii.] Considerations affecting Training Works. 325
4, 11, and 13). Where, however, dredging is resorted to in
combination with training works, a more suitable rate of
enlargement can be provided, together with an adequate
depth of channel, preventing an excessive reduction of tidal
capacity, and abrupt alterations in width and depth at the
outlet of the trained channel, as indicated by the training
works of the Clyde and the Weser (Plate 8, Fig. 7, and
Plate 9, Fig. 8).
Concluding Remarks. It must be always borne in mind,
in designing training works through an estuary subject to
the introduction of sediment, that no training works can be
considered complete unless they guide the channel right out
to deep water, and that training walls cause accretion at the
sides of an estuary, even in advance of their extremities.
Moreover, reliance must not be placed solely upon the scour
of the currents for the deepening of a trained channel, since
this necessitates an undue restriction of its width, and loss
of tidal capacity ; but dredging must be employed as an
auxiliary to the training works, in order to provide a large
enough channel for the free admission of the flood tide.
The rate of enlargement of a trained channel seawards, to
ensure the free admission of the largest volume of tidal water
consistent with the training and improvement of the channel,
must depend upon the special physical conditions, the form
of the estuary, and the width of its outlet ; and it should,
as far as practicable, be kept within the limits which nature
and experience indicate as the most favourable. The study
also of the various states of an estuary, as shown on suc¬
cessive charts, is very valuable in designing the direction for
the trained channel, by indicating the action of the natural
forces in the estuary, and the position of the most stable
and deepest outlet channel. In determining the position of
the training walls, provision has to be made for preserving
the access to any ports which may exist on the shores of the
estuary ; and in view of the inevitable accretion resulting
326 Reclamations injurions to Navigation.
from training works, care should be taken in designing the
works to avoid as far as possible injuring seaside towns by
the accumulation of deposit in front of them.
Reclamation of land is unfavourable to the interests of
navigation in an estuary, by reducing the tidal capacity; and
though training works have frequently led to reclamation,
by stopping the wanderings of the channel and promoting
accretion, as in the case of the Ribble, training works should
be designed so as to avoid reclamations as far as possible,
instead of favouring them, as in the Seine training works, or
with the object of reclamation, which proved such a fatal
course in the Dee estuary.
CHAPTER XIV.
EXPERIMENTAL INVESTIGATIONS ON TRAINING
WORKS IN ESTUARIES.
Origin of Investigations. Advantages of experiments with working Models of
Estuaries. Material employed for Bed of Estuary. Arrangement of Models ;
Method of producing Tidal Flow and Fresh-water Discharge, and Formation of
Training Walls. Suitability of Seine Estuary for instituting Investigations on
Training Walls. Experiments with Model of Tidal Seine : Description of Model ;
Resemblance of Model to actual Estuary; Changes produced on inserting
Training Works; Effects produced on inserting various Schemes for prolonging
Training Works, namely, contracted outlet, sinuous channel, and trumpet-shaped
channel ; Comparison of Results of Schemes. Large Rouen Model described.
Experiments with Model of Mersey Estuary : Contrast to Seine Estuary ;
Description of Model ; Resemblance of Channels in Model to those in Estuary ;
Transformation produced by inserting Training Walls in Inner Estuary; Improve¬
ment effected by Outer Training Works. Concluding Remarks, reliability of
method, its great value for preliminary investigations.
The great differences of opinion exhibited by engineers
in the various schemes proposed for the prolongation of the
training walls in the Seine estuary, and the conflicting views
expressed as to the influence of training works in the upper
estuary of the Mersey, when proposed for the outlet of the
Manchester Ship-Canal in 1883 and 1884, led me to en¬
deavour to obtain a definite solution of these disputed
problems, which no arguments from analogy appeared capable
of effecting, and at the same time to establish the principles
of training works in estuaries on a more scientific basis, by
experimental investigations with working models of the Seine
and Mersey estuaries on a small scale, carried out in 1886-9.
The main causes of the uncertainties which exist as to the
effects of training works in estuaries, and the divergence
of opinion to which schemes of such works give rise, are
328 Value and Limits of Model Experiments, [chap.
undoubtedly the very great differences in the configuration
and tidal capacity of estuaries, and the varieties of tidal
range and fresh-water discharge, so that engineers hesitate
to apply the experience gained from training works in one
estuary to another estuary subject to more or less different
physical conditions. These are the kind of peculiarities
which models are specially qualified to reproduce; and in
a working model of the inner estuary of the Mersey, con¬
structed in 1885 with the object of showing that the slight
regulation of the Cheshire shore of the estuary contemplated
in the final design of the Manchester Ship-Canal, now
completed (Plate 7, Fig. 11), would not affect the state of
the estuary, the existing condition of the inner estuary
was very fairly attained h Though the configuration of the
shores of an estuary can be exactly imitated in a small scale
model, the vertical scale has necessarily to be made much
larger than the horizontal scale, in order to produce any
appreciable rise of tide and flow of currents ; whilst, not¬
withstanding this exaggeration of the tidal rise, the size of
the particles of the finest material that can be procured to
represent the shifting sands of an estuary, are very large in
a small model in proportion to the actual particles of sand
constituting the sandbanks of an estuary. Accordingly, it
is not practicable to obtain the same proportionate depths
of channels in a small model as in nature ; but, on the other
hand, a small model can be inspected very closely, and
changes in the channels or banks readily noted ; and owing
to the extreme shortness of the tidal period in a model,
alterations which may take years in an estuary, are effected
in a short period in a model. The action of waves in an
estuary, produced by the prevailing winds, cannot, indeed,
be reproduced in a model ; but fortunately, in the particular
estuaries experimented upon, the direction of the prevalent
1 ' The Régime of Rivers and Estuaries,' Professor O. Reynolds, Frankfort
Congress on Inland Navigation, 1888, p. 11 ; compare figs. i and 2.
xiv.] Materials for Bed of Model Estuary. 329
winds and of the flood tide entering the estuary approximately
coincide, and therefore the tidal impulse given in the model
practically represented the combined influences of the flood
tide and waves.
Material used for Bed of Model of Estuary. In order that
experiments on the effect of training works in an estuary
might have any value, it was absolutely essential to produce
in the model the accretion which training walls have been
found to produce in nature ; and the obtaining of this
accretion in the model constituted the most novel part of
the investigations, and the most difficult effect to reproduce.
Various finely divided substances of low specific gravity
were successively tried, but the fine particles became gradually
consolidated into too compact a mass to be readily affected
by the minute currents ; whilst the sawdust of heavy woods
became swollen in the water, and was too easily shifted by
the currents1. The only material tried which at all satisfied
the requisite conditions, was very fine sand containing peat,
procured from the Bagshot beds on Chobham Common.
This sand, being perfectly clean, did not become a compact
mass ; and as the water was readily able to percolate through
it, the finest particles could be transported by the currents.
Arrangement of Models for Investigations. The first step
required in an investigation of this kind is to reproduce
in a model a previous, or existing condition of the estuary
to be investigated. For this purpose, in the two estuaries
experimented on, the estuary, with the whole of the tidal
portion of the river above it, was moulded in Portland cement
to its exact form, though to a distorted scale vertically ;
and the model was also extended to an adequate distance
beyond the mouth of the estuary. The shifting portions
of the bed of the estuary were formed with fine Bagshot
1 ' The Principles of Training Rivers through Tidal Estuaries, as illustrated by
Investigations into the methods of improving the Navigation Channels of the
Estuary of the Seine,' L. F. Vernon-Harcourt, Proceedings of the Royal- Society,
v°l. 45, P. 511.
33° Arrangement and Working of Tidal Model, [chap.
sand, laid to an adequate depth over the cement bottom ;
and the fresh-water discharge was let in through a small
tap from a cistern at the tidal limit. The tidal flow and
ebb were effected by the raising and lowering of a zinc tray
of suitable size, enclosed on three sides, and connected with
the model on the open side by an indiarubber hinge. The
tray was fastened to the model at an angle so adjusted
that the water flowing into the model estuary, on lifting
the tray, entered in the direction that the flood tide actually
approaches the estuary. The size of the tray also, and
the amount its outer end was raised, were arranged so that
the rise of the water in the model corresponded with the
rise of tide at the mouth of the estuary, according to the
small vertical scale adopted. The time occupied in each
raising and lowering of the tray, or the tidal period of the
model, was adjusted so as to reproduce in the small-scale
model the tidal phenomena of the river under examination,
such, for instance, as the time required by the tide to reach
the end of the tidal portion of the river in relation to the
tidal period.
As soon as it was found that, after making suitable adjust¬
ments, the working model fairly represented in miniature
the actual condition of the untrained estuary, with its shifting
channels and other peculiarities, strips of tin representing
the training walls to the distorted vertical scale, bent to the
desired lines, were introduced ; and the gradual changes
which resulted from their introduction, after working the
model for a great number of tides, were carefully recorded
on a chart of the estuary.
Suitability of Seine Estuary for Primary Investigations.
The correspondence obtained between the state of the estuary
in a model and the natural condition of the estuary, might
afford a presumption that the effects of training works in
a model would represent approximately the effects such
works would produce in the actual estuary. Owing, however,
xiv.] Special Value of Seine Investigations. 331
to the influence exercised by accretion on the changes in
an estuary produced by training walls, it would be impossible
to prove, by the preliminary step in the investigations, that
the training walls in the model would be able to exercise
the same influences with regard to accretion as corresponding
works in the real estuary. In fact, the crucial question in the
investigations was, whether accretion could possibly be made
to result from the introduction of training walls in the model,
with the comparatively coarse material composing the bed of
the miniature estuary, in at all a similar manner as had been
witnessed in various estuaries. Fortunately the Seine estuary
was specially suitable for the application of such a test to
a model, for its state previously to the commencement of the
training works is recorded in early charts, as well as its
condition since the completion of the works, by various
surveys. If, besides the first step of making a model fairly
representing the original condition of the Seine estuary
previously to the works, it was possible to reproduce in the
model the actual changes produced by the training works, it
might fairly be inferred that the model sufficiently accorded
with the condition of the estuary to indicate future changes
from any prolongations of these works. On the accomplish¬
ment of this second step, the model becomes a most valuable
guide as to the effects, and relative advantages, of the various
conflicting schemes proposed for the completion of the
training walls in the estuary of the Seine ; and the system
can be applied with confidence to ascertain the effects of
training works in any estuary with shifting sandbanks.
Experiments with Model of Tidal Seine1.
The remarkable variety in the schemes proposed for the
completion of the Seine training works was referred to in
1 ' The Principles of Training Rivers through Tidal Estuaries, as illustrated by
Investigations into the Methods of improving the Navigation Channels of the
Estuary of the Seine,' L. F. Vernon-Harcourt, Proceedings of the Royal Society,
vol. 45, pp. 504-524, and plates 2, 3, and 4.
332 Object and Description of Seine Model, [chap.
the last chapter ; and the chief types are given in Plate 10,
Figs, i to 6, though several others have been put forward in
various publications h The frequent appearance of fresh
schemes, and the conflict of views thereby exhibited, led to
my undertaking these investigations ; and the Seine estuary
not only provided ample material for experiments with these
divergent schemes, but also afforded the means of determining
whether it was possible to reproduce with sufficient exactness
the effects of training works in a model.
Description of Tidal Seine Model. The model of the tidal
Seine, made to the scales of 4Tj£Tn> horizontal and vertical,
extended from the tidal limit at Martot down to about Dives
on the Calvados coast, to the south-west of Trouville. The
tidal rise at springs, produced by the extreme tipping of the
tray, amounted to 0.71 inch in the model; and the tidal
period was about 25 seconds. The bed of the estuary was
formed of a layer of fine Bagshot sand, except where rock
crops up to the surface in front of Villerville, and the solid
portions of the Amfard and Ratier banks at the outlet
consisting of gravel or rock, which were moulded in Portland
cement. The hinge of the tray was arranged so that the
tidal influx took place from a direction about 5° to the west
of north-west. The fresh-water discharge, proportionate to
the scale of the model, was admitted at the upper end from
a tank ; and a corresponding quantity was discharged at the
outer end of the model during the higher half of the tide, to
prevent an increase in the volume of water in the model.
Results obtained on working the Seine Model. From the
commencement, the bore was well marked by a sudden rise,
particularly at the point of the model representing Caudebec ;
and the reverse current, which flows out of the estuary near
Havre just before high water, was apparent, resulting from the
filling of the estuary by the second tide following the southern
shore, which was produced by the angle at which the tray
1 Minutes of Proceedings Institution C. E., vol. lxxxiv, plate 4, fig. 9.
xiv.] Production of Natural Changes in Model. 333
was placed to the estuary and the prolongation of the
southern coast. Even with the silver sand first used, shifting
channels were formed ; and a condition resembling fairly
closely a previous state of the estuary was obtained after the
model had been worked for some time. The silver sand,
however, proved too coarse to be readily moved by the feeble
currents of the model, and failed to produce accretion when
the training walls were inserted. Various other fine substances
also did not adequately fulfil the requisite conditions, or
reproduce satisfactorily in miniature the state of the Seine
estuary after the introduction of the training works.
Insertion of Seine Training Works in Model. At last, with
a bed of Bagshot sand, the condition of the estuary, as
indicated in the chart of 1834, was obtained in the model;
and accretion was produced on the insertion of the training
walls. Strips of tin were fixed in the model to the heights,
and in the positions corresponding to the actual works, and
as far as practicable in a similar sequence, so as to conform
as nearly as possible to the same conditions in the model as
in the estuary. The training walls in the model produced
a deepening in the trained channel, and an accretion at the
sides of the estuary behind the walls, closely resembling the
actual changes in the estuary ; and, moreover, the accretion
extended down to Hoc Point on the northern side, obliterating
the Harfleur channel, and down to Honfleur on the southern
shore, corresponding precisely in extent to the accretion
which has taken place in the estuary.
Effects of various Schemes on Model of Seine Estuary.
The success at last obtained in the second stage of the
investigations, enabled the third, and final stage to be entered
upon with confidence. As it had been proved that the model,
with a bed of Bagshot sand, could reproduce the results of
training works actually carried out, the effects of prolongations
of these works in the model, according to the different
schemes, might reasonably be regarded as foreshadowing the
334 Results of Schemes in Seine Model, [chap.
effects which these works would actually produce if carried
out in the estuary. Accordingly, some of the typical schemes
were inserted successively in the estuary; and the model was
worked for from 3500 to 7300 tides for each scheme, with
the results shown on Plate 10, Figs. 1 to 6.
The contracted outlet formed by a breakwater extending
from the southern shore at Villerville to the Amfard bank, as
proposed in some of the schemes, converting the estuary into
a large sluicing basin, naturally produced a deep outlet
channel between the Amfard bank and Havre in the model
(Plate 10, Figs. 1 and 2). The inside channel, however, was
tortuous in the first case, and shallow above Honfleur in the
second arrangement, owing to the divergence in the directions
of the flood and ebb currents, and the gradual silting up of
the estuary in the model, from the deposit resulting from the
loss of velocity experienced by the rapid silt-bearing flood-
tide current through the narrow outlet on expanding in the
sheltered estuary. Moreover, the contraction at the mouth
appreciably reduced the tidal rise inside, notwithstanding the
minute scale of the model ; whilst a bar appeared between the
deepened outlet channel and deep water in the sea. The
breakwater also, by checking the silt-bearing flood tide along
the southern coast, and protecting this shore from the out¬
flowing ebb, caused a great advance of the foreshore outside
it in front of Trouville in the model; and by the shelter it
afforded on its inner side to the southern shore of the estuary,
it occasioned a similar progression of the foreshore between
Villerville and Honfleur in the model. The accretion in the
model estuary gradually reduced the great initial depth of
the narrow outlet channel between Amfard and Havre, by
diminishing the tidal capacity of the scouring basin which
formed it.
The results of working the model with the prolongation of
the trained channel in a sinuous form, as advocated in some
proposals (Plate 10, Figs. 4 and 5), indicate clearly the
xiv.] Superiority of Trumpet-shaped Outlet. 335
disadvantages of the want of coincidence in the direct course
of the flood tide and the winding course of the ebb, occasioned
by this form of channel. In neither of the cases investigated
was a deep, direct, or continuous channel obtained by this
form of training works in the model.
The schemes providing a trumpet-shaped outlet channel
furnished the most satisfactory results in the model (Plate 10,
Figs. 3 and 6), giving wider and more uniform channels than
the other types of schemes. The arrangement in which the
channel was given the least expansion, and a continuously
concave southern training wall to the outlet, afforded a some¬
what more uniform channel in the model than the other ; but,
at the same time, the narrower outlet occasioned a greater
accretion inside the estuary, and a considerable progression
of the foreshore in front of Trouville in the model. Whilst
a restriction of width between the training walls produces
a rather more direct channel, the maintenance of the width
at the outlet by a greater expansion of the training walls
preserves a larger tidal capacity, and obviates great changes
in the estuary and near the outlet. On the whole, a more
expanded channel, in which deficiencies in depth or direction
could be remedied by dredging, or possibly by low inner
training works, appears preferable to a narrower trained
channel deepened entirely by scour, on account of the greater
loss of tidal capacity, and more important changes in the
natural conditions produced by a considerable reduction in
the width of the outlet.
Comparison of the Schemes investigated. The conversion
of the Seine estuary into a vast scouring basin by a break¬
water greatly contracting the outlet, though securing deep
water at the mouth, evidently would produce too extensive
modifications in the estuary and adjacent foreshore to be
admissible. Moreover, the irregular depth of the channel
thereby formed, the variation in the velocity of the currents
at different parts, and the great eventual loss of tidal capacity
336 Comparison of Schemes : Ronen Model, [chap.
in the estuary, due to inevitable accretion, producing a great
reduction in the tidal influx and efflux at the outlet, would
preclude the adoption of any scheme of this nature.
The sinuous type of trained channel, by ensuring a divergence
between the flood and ebb currents, and a consequently
unsatisfactory channel, should evidently be avoided, as far as
practicable, in tidal rivers. Accordingly, schemes of this form
are inexpedient for the training of rivers through tidal
estuaries, being based upon the stability of the channel of
a non-tidal river at curves, which does not hold good when
the current is alternately in opposite directions.
The only schemes for prolonging the Seine training walls
which gave satisfactory results in the model were those which
expanded to a trumpet-shaped outlet. Before any definite
scheme is adopted, the most suitable form and rate of
expansion should be determined by further experiments ; but
by resorting to dredging, aided possibly by low-water inner
training works, it should be possible to adopt a sufficiently
expanded channel to avoid the occurrence of serious changes
in the condition of the estuary and adjacent foreshores.
Rouen Model of Tidal Seine. In consequence of the
results of these experimental investigations, and in accordance
with the views expressed by me after giving a brief account
of my experiments at the Frankfort Congress in 1888 1, the
government engineers at Rouen have made a large model
of the tidal Seine, with the object of further investigating the
proposed schemes. At first a reservoir was adopted, in place
of the winding upper channel of the tidal river ; but it proved
necessary to complete the model up to the tidal limit, as done
at the outset in the earlier model. The Rouen model now
looks exactly like an enlarged copy of the little model ; and
the same angle was adopted for the introduction of the flood
tide. The rise of tide, however, is effected by the immersion
1 Verhandlungen der Allgemeinen und Abtheilungs-Sitzungen, III In 'er¬
nationaler Binnenschifffahrts-Congress zu Frankfurt am Main, 1888, p. 193.
xiv.] Value of Rouen Model of Tidal Seine. 337
of blocks in little wells, worked by a gas engine, instead of the
tipping of a tray. The horizontal scale of the model is Ttrorr >
and allowance has been made for using a vertical scale of
or j1^, with a tidal period of 60 or 100 seconds. A small
maregraph enables the miniature tidal curves to be registered
at certain points on the model.
The comparatively large size of the model should enable
very accurate investigations of the relative effects of the
various schemes to be carried out ; and the results obtained
will be of the highest interest, both in determining the general
form that should be adopted for the completion of the training
works in the Seine estuary, and also in indicating the value
of these experimental investigations for the solution of the
difficult and intricate problems involved in the training of
rivers through tidal estuaries. In working the Rouen model,
it appears that the preliminary steps of the investigations, of
reproducing a former state of the estuary, and then effecting
the changes produced by the existing training works, were
dispensed with, and that the influences of some of the schemes
for the prolongation of the training walls, on the model estuary,
were investigated in the first instance. These preliminary
experiments seem always expedient ; and the substitution of
fine Bagshot sand for the coarser sand hitherto used in the
model, will be advantageous.
Experiments with Model of Mersey Estuary1.
The Mersey estuary, with a wide inner estuary, followed by
a narrow neck opening abruptly into the wide expanse of
Liverpool Bay, and with a comparatively small amount of
fresh water flowing into it from the Mersey and the Weaver,
offers a striking contrast in form and condition to the Seine
1 Proceedings of the Royal Society of London, vol. 47, p. 142 ; ' Effects of
Training Walls in an Estuary like the Mersey,' L. F. Vernon-Harcourt, London,
890 ; and ' Amélioration de la Partie Maritime des Fleuves y compris leurs
Embouchures,' L. F. Vernon-Harcourt, Vm6 Congrès International de Navigation
intérieure, Paris, 1892, p. 27.
z
338 Description of Model of Mersey Eshiary. [CHAP.
estuary (compare Plate 7, Fig. u, with Plate 9, Fig. 1). It,
accordingly, afforded an excellent opportunity for the extension
of the system of experimenting on training works, by means
of models, to a totally different estuary. As no training
works have been carried out in the Mersey estuary, the
second stage of the Seine experiments could not be effected ;
but the satisfactory results obtained in the Seine model
dispensed with the necessity of a similar stage in the case
of the Mersey model, in which similar arrangements, and
a similar material for the bed of the estuary were adopted.
Description of Mersey Model. The model of the Mersey
extended from Warrington to beyond the bar in Liverpool
Bay, and was made to a horizontal scale of and
a vertical scale of 5^. The exaggeration of the vertical
scale was made less than in the Seine model, on account of
the greater rise of tide at the mouth of the Mersey, and the
greater fall of the bed of the estuary, than in the case of
the Seine. The shore-lines of the model were extended to
Formby Point on the northern side, and to the sandbank
separating the Mersey estuary from the estuary of the Dee
on the southern side ; and the sides of the model were
prolonged in slightly diverging lines seawards up to the
hinge of the tray, which was placed a little distance beyond
the site of the bar. The tray was so adjusted that the flood
tide was introduced into the model from a direction 17° 15'
to the north of west, which appears to be approximately the
direction from which the tide enters Liverpool Bay. The
tray, on being raised to its full extent, produced a rise of
I inch opposite the site of Liverpool, which corresponds to
the equinoctial spring-tide rise of 31 feet on the scale of the
model ; and the tidal period was 3a seconds in the model.
The proportionate small flows of water representing the fresh¬
water discharges of the Mersey and Weaver, were let in from
two small cisterns at Warrington and Frodsham respectively
in the model.
xiv.] Resemblance of Model Estuary to Mersey. 339
Results of working Mersey Model. The model was first
worked in its natural condition, to ascertain how nearly the
existing condition of the Mersey estuary could be reproduced
in the model. In working the model for a number of tides,
irregular channels appeared in the bed of Bagshot sand in
the inner estuary, which constantly shifted; and two more
permanent channels skirted each shore above the ' narrows,'
up to Eastham and Garston respectively in the model
(Plate 10, Fig. 7). A deep channel was formed in the
'narrows' between Liverpool and Birkenhead in the model,
and extended into the bay with a gradual reduction in depth
out to the bar, which appeared in the model approximately
in its actual site, showing a tendency to shift its position
slowly, like the Mersey bar. A channel was also formed
along the Cheshire shore in the bay, at right angles to the
main channel, corresponding to the well-known Rock Channel ;
and a shallow subsidiary channel diverged from the main
channel towards Formby Point, which exists in the actual
estuary (compare Plate 7, Fig. 11, with Plate 10, Fig. 7). On
the whole, considering that it is impossible to produce the
action of breaking waves in a model, which undoubtedly
causes the rugged appearance of the sandbanks in Liverpool
Bay, the model reproduced remarkably closely the main
features of the Mersey estuary, confirming, in the case of
a very different estuary, the practicability of exhibiting in
a small model the general conditions of an estuary out to the
open sea, previously established with regard to the Seine.
Experiment with Training Walls in Inner Mersey Estuary.
Two problems of considerable interest present themselves for
investigation with regard to an estuary of the peculiar form of
the Mersey, namely, the effect of training works in the inner
estuary on the general condition of the estuary, and the
influence that training works in the outer estuary might have
on the outlet channel.
The proposals made in 1883 and 1884, to provide an outlet
Z a
34o Silting up of Inner Estuary in Model. [chap.
for the Manchester Ship-Canal from Runcorn to deep water,
by training works through the inner estuary of the Mersey,
gave rise to most divergent opinions amongst engineers. It
was admitted that the maintenance of the depth in the outlet
channel below Liverpool, and over the bar, depended on the
preservation of the tidal capacity of the inner estuary. The
difference of opinion arose on the question whether the
formation of a trained channel for the river through the inner
estuary would, or would not, give rise to accretion, and
consequently affect the tidal capacity of this part of the
estuary. Though this scheme was abandoned on its rejection
by a Select Committee of the House of Commons in 1884,
after passing through the House of Lords, and the present
line of canal, skirting the Cheshire shore down to Eastham,
was consequently substituted, the effect of training works
under such conditions is of considerable scientific interest ;
and it is not advisable that this subject should remain in
doubt, as a similar problem might at any time arise in some
other estuary. Accordingly, strips of tin were inserted in the
Mersey model, representing the training walls proposed for
the outlet channel of the canal through the upper estuary.
After working the model for some time with this modification,
a remarkable change had become apparent in the estuary.
The channel between the training walls had been scoured out,
but the rest of the upper estuary had been filled up ; and the
consequent reduction in tidal capacity had produced a great
deterioration in the outlet channel beyond the ' narrows,' thus
fully confirming a view expressed in 1881, before the canal
scheme was contemplated 1, and the fears entertained by the
1 1 The condition of the river above Liverpool has formed the subject of consider¬
able discussion. It would be impossible to improve its depth without confining
its channel by training banks, which would lead eventually to the reclamation of
a large portion of the existing tide-covered area between Runcorn and Liverpool.
The gain to the river above Liverpool from such works would be considerable ;
but the reduction of the tidal receptacle would reduce the scouring power of the
current in the estuary below, and would in all probability lead to a deterioration
of the outlet.' ' Rivers and Canals/ ist edition, p. 257.
xiv.] Training Walls in Outer Mersey Estuary. 341
Liverpool authorities on the introduction of the project
(Plate 10, Fig. 8). The result of this experiment proves
conclusively that training works should never be put into
a portion of an estuary subject to the introduction of sediment,
whenever the maintenance of its tidal capacity is essential for
the preservation of the depth of the outlet channel of a large
port situated lower down on the estuary.
Experiments with Training Walls in Outer Mersey
Estuary. The improvement in the depth over the Mersey
bar, which is being effected by dredging, has been previously
referred to (p. 379). The only other way in which the depth
over the bar might be increased, is by regulating the outlet
channel beyond New Brighton by training works, which
would render the channel more stable, and produce a greater
scour across the bar. Training works would be costly in
such an exposed situation, especially if they had to be carried
some distance out ; but they would obviate the constant
maintenance necessitated in deepening the channel by dredging
alone. The question, however, possesses a general interest
irrespectively of its special application to the Mersey ; and it
was mainly as a novel problem that the experiments with
training walls in the outer estuary of the Mersey model were
undertaken. The training walls inserted in the model in this
instance, were designed to form a trumpet-shaped outlet in
continuation of the shore-lines of the ' narrows,' so as to
promote, rather than impede, the influx of the tide into the
' narrows ' and inner estuary, whilst restricting the ebb and
flow to a single channel (Plate 10, Fig. 9). On working the
model with this arrangement, the outlet channel became fixed
and deeper ; whilst the maintenance of the inner estuary was
not affected. The general direction and depth of the outlet
channel was improved by removing the northern portion of
the southern bank, which protruded somewhat into the trained
channel, showing that dredging is a very useful auxiliary to
training works. It appears probable from the experiments
342 Value of Experiments on Training Works.
that in the actual estuary, the outlet channel might be regu¬
lated and deepened by a single training wall extending sea¬
wards for a moderate distance from New Brighton, diverging
somewhat from the Lancashire shore, and shutting off the
Rock Channel ; and this arrangement, besides rendering the
channel over the bar more stable, would reduce the amount
of dredging required for maintenance.
The experiments with the Mersey model show that in
a very irregular and defective estuary, training works may be
injurious in the upper portion, and advantageous below where
accretion at the sides does not affect the maintenance of the
outlet channel.
Concluding Remarks. The experiments carried out with
the models of the Seine and the Mersey show that definite,
and apparently reliable, indications can be obtained in this
manner of the effects of training works in any tidal estuary,
since definite and concordant results were obtained from
models of two such very dissimilar estuaries as the Seine and
the Mersey. Considering the unexpected results that have
followed training works in estuaries, the divergence of opinion
on the subject exhibited by engineers, and the facility with
which the influence of various schemes may be tested in
a working model, it may be hoped that in future no training
works will be carried out in estuaries without seeking the
guidance so readily afforded by the experimental method.
By this means, injurious works leading to irremediable conse¬
quences may be avoided, and the best out of several alternative
schemes may be selected. The system of experimenting
with models of estuaries promises to prove of inestimable
value to engineers in the design of training works through
sandy estuaries ; and each set of such experiments will aid in
the establishment of definite principles with reference to this
complex branch of the science of river engineering.
HYDROLOGY OF RIVERS.
PLATE I
Fig.l.
RIVER SEINE BASIN.
SHOWING DISTRIBUTION OF RAINFALL
Ft#. 2.
RIVER SEINE BASIN.
SHOWING TRIBUTARY BASINS AND NATURE OF STRATA
Perrrveables Strata;
Imp errrveecble; Strata (Flat.)
Imperrrvmbley Strata f Sloping .J
AVERAGE ANNUAL RAINFALL
jBelow
24 Inches
2A bo
32 „
32 to
40
40 to
48
Above/
48
Miles id s o
'""'""i
Fig. 3.
DIAGRAM SHOWING DAILY HEIGHT OF SEINE AT POISSY.
"Warm Season '878 i88i. Cold/%
JULY. AUGUST. SEPTEMBER. OCTOBER. NOVEMBER. DECEMBER. JANUARY7^ FEBRUARY^ß MARCH.
yy
d.atixm'i
li7f
m
FLOOD DIAGRAMS.
Fig. 7. 1882-1883.
NOGENT-SUR - SEINE .
NOVEMBER. DECEMBER. JANUARY.
/-
Iniu ¿Lxhons
t'orna i
er ce
Fig. 13.
SEINE AT PARIS.
NOVEMBER. DECEMBER. JANUARY.
Fig. 8.
BRENNE AT MONTBARD.
Feet jo
A
_lriun¿jzlzi>n
1
/V>
/
ft
J
\
A
Fig.9.
YONNE AT SENS.
Fig. 14.
AISNE AT PONTAVERT.
FeetJS
À
I'M.
ûH't
ns
C ifn r
rtzrix
V
n
r-
A
J"
y
\J
Feetw
J
V/
-r~\
X
V"S
~Zy
/-I
niau
lit
pre
%A
ftimr
nene
\7
Fig. 10.
SEINE AT MONTEREAU.
Feetis
iM
V
Fig. 4.
DIAGRAM SHOWING DAILY HEIGHT OF SEINE AT BEZONS.
I 881 - 1884 .
Feetzo
woe
18
Irxinljalarte
Fig . 5.
DIAGRAMS SHOWING DAILY HEIGHT OF SEINE AT BEZONS.
Warm Season/. 1884-1887. Coli^
JUNE. JULY. AUGUST. SEPTEMBER. OCTOBER. NOVEMBER. DECEMBER. JANUARY. FEBRUARY
X
■VJ
Fig. 15.
OISE AT VENETTE.
Feetzo
Vv
■m&.zce;
MARCH. APRIL.
X
■X
It
&<v-
X
'A
rs
Fig. 11,
MARNE AT ST DIZIER
FeetJS
Fig. 6.
1887-1890 .
A
=N
\
Fig. 16.
SEINE AT MANTES.
Feet 25
mx
Ti'jj'jdcdic
X8
Commence
-A/M
A
Ax/
fc
A
isSL
it
Aj 13
18 ÍS
Fig. 12.
MARNE AT CHAL1FERT.
Feeti5
-
It
¿w 0
T_S_
C 7m
à*
nze__
—
\
(
K
-
i
X
jt'tns
\J
Joli Ii ü az-Aoiomew & Co-EdinT
DREDGING AND EXCAVATING MACHINES
Fig. I.
BUCKET-LADDER HOPPER DREDGER
LONGITUDINAL SECTION.
Fig. 9.
STEAM NAVVY
*AC!TY/y
TONS
=» Fig. 5.
THREE-BLADED GRAB BUCKET
Fig. 10.
ROCK BREAKER.
SEMI-CYLINDRICAL GRAB BUCKET.
Fig. 7.
DREDGER WITH LONG SHOOT
GRAB BUCKET HOPPER DREDGER
Fig. 2.
LONGITUDINAL SECTION
Fig. 3.
CROSS SECTION
Fig. 8.
SAND-PUMP DREDGER
opacity ¡àf hopper
\ 8ch \tons. i \
H O P¡f E R
Scale/ to Fig? 1,2,3,6,8,&10 = z4o
/ \ /"\ /
/ \ CAPJfCIfY OF J
GRAB BUCKET HYDRAULIC DREDGER
\ 500
FCO'C'-K'
Fig. 5. RIVER RHONE. Fi9 6
26-9 MILES TO 28-6 MILES BELOW LYONS. 54 MILES TO 55-3 "MILES BELOW LYONS.
PLATE III
TEDD1NGTON WEIR.
RIVER THAMES.
Fig. 15. Fig.16.
SECTION. ELEVATION.
Fig. 19.
LA MONNAIE WEIR.
RIVER SEINE, PARIS.
.
REGULATION WORKS: FIXED AND DRAWDOOR WEIRS.
Fig. 8. Fig. 10.
SECTION OF WEIR ON RIVER SEVERN. SECTION OF LA TEMPLE
Fig. 14.
GODAVERY ANICUT, INDIA.
SECTION .
SoaZey to Fig. 19
Fig. 7.
PLAN OF WEIR ON RIVER SEVERN.
Fig. 11.
PLAN OF FOSSAT WEIR, RIVER LOT.
FRANCE.
Fig. 13.
Sighest FloocL Lev eh.
DEHREE ANICUT, SONE CANAL, INDIA.
SECTION .
Fig.12.
OF FOSSAT
WEIR .
>
River Lot.
Scale to Fig? 8,10,12,13,&U.
Ft. ,, ,-f,,, w 2<? so 40 Feet.
SECTION
Fig.l.
DEPTHS OF WINDING RIVER
VARIATION IN
NIEDER WALLUF.
MOMBACH
GRUNEWALDE.
FROHSE
SCHONEBECK
ST ALBAN.
Ft. 1000 500 O
I 1111111111
Scales to Fig ? 2,3,4-, 5, & 6 - 4o,ooo
5,000
1o,ooo Feet.
j i I
PLAN OF
Fig. 9.
LA TEMPLE WEIR,
FRANCE.
RIVER LOT.
River Lot.
SECTION
SLUICE-GATES WITH FREE ROLLERS
Fig.<5.
LONGITUDINAL. SECTION OF RIVER MAIN
MONTEREAU
FRANKFORT.
NIEOERRAD.
COURBETON.
HÖCHST.
VARENNES.
MADELEINE.
FLÖRSHEIM.
PLAN OF PORT-À-L'ANGLAIS LOCK AND WEIR
LONGITUDINAL SECTION FROM PARIS TO MONTEREAU
CHAMPAGNE.
SAMOIS
CAVE.
MAINZ,
PARIS
MELUN
VIVES EAUX.
CITANGUETTE.
NAVIGABLE PASS.
COUDRAY.
EVRY.
-LOCK
Miles 20
RIVER SEINE
PAßlS.—
"TA MONNAIE.
NAVIGABLE PASS.
Scale to Fig. 5.
lOOO
BOUGIVAL.
POISSY.
CARRJERES...
LOCK
Datum- Line- 80 Feet- above- Zero of .Marine- Charts.
M ERICOURT.
PORT VILLEZ.
-SwjrvrrteK.
Exgh- Water Ori
\160 MILES
Fig. 11.
POSES WEIR.
NAVIGABLE PASS
69.2
Datum-Line- 30 Feet? helxr.v Zero of Marine Charts.
Fig.lZ.
HYDRAULIC SHUTTER WEIR
RIVER YONNE, FRANCE.
: 6,000
NAVIGABLE PASS.
Frame- with- Curtains.
THROSTLE-NEST WEIR
RIVER IRWELL.
Cúrtain,
boTtede
V ixo - \
PANELS ENLARCED
SURESNES WEIR
NAVICABLE PASS-
Frame -with- Panels.
PLAN OF POSES WEIR AND LOCKS
Flood Level S^fareh-^1816.
Highest NasvigcCbT^Level.
r Fig • m •
PORT-À-L'A N G LAIS WEIR
Shutter Weir.
Fig. 16.
CHAR LOTTEN B U RC DRUM WEIR
RIVER SPREE, GERMANY.
Fig. 15.
BEAR TRAP WEIR, LA NEUVILLE, RIVER MARNE
FRANCE.
RIVER SEINE
POINT ISLAND
NAVIGABLE
Drum
LOCKS, AND MOVABLE WEIRS. PLATE IV
OUTLETS
OF RIVERS FORMING DELTAS.
PLATE V
Fig. 2.
Fig.l.
MOUTH OF THE RHONE.
1891.
MISSISSIPPI SOUTH PASS JETTIES.
18 92.
;#
' ■- / _ ' lAQr> ?i .Kl •
If -Si ; F . yd--' ■ jJtfJLV
¿o?>
-- /;
JaiinBarlnoloniew& Co..Ei3in£
i
WORKS AT MOUTHS OF TIDAL RIVERS.
■
PLATE VI
PLANS, 50,000. LONG-r SECTIONS: HOR«r söööö, VERTV göo
Fig. 14.
SECTION OF JETTY.
OUTER PORTION.
River SicLes.
O'CONNELL
BRIDGE
Jc✠Bartkolomew& Co.,Eäiii?
SIMULTANEOUS TIDAL CURVES, RIVER SEINE
Spring Tides 13th- September 1876.
RIVER THAMES
Fig. 7.
plan 1894-
canvey island
iholc havem. __
Faihoi
The No re
Light Ship
crosby.
Fig. 6.
LONGITUDINAL SECTION
uncorn
<^0 ; / sheerni
LIVERPOOL
hale head.
:rodsh am
garston
Datum, line 30. O below Zero of Marine, Charts.
Tig . 2.
tidal diagram, river seine.
Fig. 5.
SIMULTANEOUS TIDAL CURVES, RIVER MERSEY.
equinoctial spring tide . 0-
Datum, Lint' 80-Ö below T.S.W.M. or 67.6 below Ordnances Datum/.
eastham.
ellesmere port.
Scale to Fig? 7,9, &lL,and Horizontals Scale to Fig? 8, 10, &-12 = zso. ooo
'helbre point.
RIVER HUMBER
Fig. $.
plan 1893.
SUNK ISLAND
line. 21.0 below Old, Doch Sill.
SPURN POINT.
¡5 2O 25 so 35 Miles.
Fig.6.
TIDAL DIAGRAM, RIVER MERSEY.
equinoctial spring tide.
250.000.
Fig. 12.
LONGITUDINAL SECTION
CK art s.
5 7 Hours.
Fig.3.
SIMULTANEOUS TIDAL CURVES, RIVER WESER
AVERAGE TIDES.
VECESACK. FARCE. BRAKE. BRI
Line,
LIVERPOOL.
garston
stan low pt
GRIMSBY
Fig. 10.
LONGITUDINAL SECTION
GRIMSBY
SPURN POINT
tidal diagram, river weser
AVERAGE TIDES.
I Uli
Datum, Fmc 9F-67 below Ordnance' Datum,, or 90. Ó below Old/ Dock, Sill/.
Datum, Line, 88.10 beUnv Ordnance Datum, or 73 10 below the SM of fírimxlrv Royal Doch.
o 5 10 15
fig1-
TIDAL DIAGRAMS AND ESTUARIES. plate vn
\^KScotswood Railway
TIDAL RIVER IMPROVEMENTS.
Plcuis.2Miles to 1 Inch. i
Long I Sections. Eorl 2 Miles to 1 Inch. Vert* 60d
PLATE VIII
RIVER TEES.
Fig. 4i.
PLAN 1891.-
j¿Fairway
} Burn.
6 Miles
Jaim.Bai<lu).UaTa«w3s CO..E3TB^
TRAINING WORKS IN ESTUARIES.
PRESTON DOCK.
RIVER SEINE
Fig. 1.
PLAN 1891.
RIVER RIBBLE
Fig .11.
PLAN 1893 94.
SALWICK.
NANTES
Fig. 3.
WALL AS RE-CONSTRUCTED
RIVER SEINE.
ST ANNES.
HAVRE;
LYTHAM.
COUERON.
LOCK
PAIMBŒUF.
ACCRETiorr^^
S.W.W.T.
Fig. 6.
LONGITUDINAL SECTION
CORDEMAIS .
CXtK MARTINIERE
LOCK. LE PELLER1N.
HESKETH,
BANK /
Fig. 7.
TRAINING WALL, RIVER LOIRE
LA MARTI N I ERE
PAIMBŒUF. LAVAU
S.W. S. T. 1892
Scale, to Fig ? 3,7,&10 = w
BERVILLE
Fig. 2.
LONGITUDINAL SECTION
BERVILLE.
Fig . 5.
DINAL SECTION
SOUTHPORT
TANCARVILLE
LONGITU
PAIMBŒUF .
ST NAZAIRE
LA MARTI NI ERE
NANTES
Fig. 12.
LONGITUDINAL SECTION
PRESTON
DOCK.
LYTHAM
S.W. O.S.T. 1891.
Plans 250.000
long]' Sections. Sor I' 250L.ooo.
^XBVELOF^PROPÛSED^REDdlNG.
39 Miles
fHELBRE
,,POtNT.
RIVER DEE
Fig. 13.
PLAN 1892.
jó) line.
LANGWARDEN
BRAKE.
Fig. 10.
TRAINING WALL, RIVER WESER
B U RTON
ivPOINT^
¡HORDENHAMM
MOSTYN QUAY.
GREENFIELD
CONN AH S
QUAY.
HASENBUREN
LONGITUDINAL SECTION.
SmELD" S.W.S.T. 1843.
.CHESTER.
CONNAH'S QUAY. QUEEN'S FERRY.
SALTNEY.
MOSTYN DEEP
z)LiO%.
Fig. 9.
LONGITUDINAL SECTION
BREMERHAVEN.
WORDENHAMM
ELSFLETH
BRAKE
BREMERHAVEN
F. W. Ay erjage Tides. 1891.
10 Miles.
2 Miles frorru
Br&merv
50 Miles
30 MJJLES
PLATE X
EXPERIMENTS ON TRAINING WORKS.
Fig.l.
SEINE
Fig . 5.
ADMINISTRATION SCHEME, 1887.
Fig.2.
M.M. DE COENE AND LEBRUN's SCHEME, 1888.
Fig. 3 .
LAVOINNE'S SCHEME,
1884- .
Scale/ for Fig? 1. to 9.
q £ ig 15Miles.
«John .B 82'tiuiLaimyw & Co-Etlm?
Miles. 5
E.
IAVRE.
LA ROQUE.
iOUVJLLE
Fig. 6.
L. F. VERNON-HARCOURT'S SCHEME, 1888.
Fix,. 7.
MODEL CHART OF
INCORN.
LIVERPOOL.
BIRKENHEAD.
EASTHAM.
:ORMBY PT
INCORN
LIVERPOOL.
HALE
BIRKENHEAD"
EASTHAM
ELLESMERE PORT.
MERSEY ESTUARY.
Fig. 8.
SHIP-CANAL SCHEME, 1884.
BY THE SAME AUTHOR
In Two Vols., demy 8vo, price 25s.
HARBOURS AND DOCKS
THEIR
Physical Features, History, Construction
Equipment and Maintenance
WITH
STATISTICS AS TO THEIR COMMERCIAL
DEVELOPMENT
O*foïî>
AT THE CLARENDON PRESS
LONDON: HENRY FROWDE
OXFORD UNIVERSITY PRESS WAREHOUSE, AMEN CORNER, E.C.
p_/W
3 5556 041 997230