Title ..XI.. .'R s^ Imprint. 18 — rods. CHEMICALS REQUIRED. Gasolene, sp. gr. 86° B., which leaves no residue. The dried and weighed residue from the determination of total solids is treated in a warm place with gasolene. After about fifteen minutes the gasolene is decanted off and the residue treated with a fresh portion. Six portions will suffice. Finally the dish is held with the pincers and the outside and rim carefully washed off by a stream of gasolene from the wash bottle, to insure the removal of all the fat. The residue is dried and weighed, and the difference between this weight and the original is the weight of the fat. The results are about 0.2 per cent, lower than those obtained by other methods, as the fat is not in a condition to be readily extracted. REFERENCE. Davenport, Alass. Stiite Bd. of Health Ref^l, xviii, 1886, p. 139. 7 2. Adams' Method. APPARATUS REQUIRED. Return flow condenser; Soxhlet's extractor ; wide-mouthed lOO cc. flasks ; large beaker for a water bath (if a special bath is not at hand) ; 5 cc. pipette ; extracted filter paper in strips 22 inches long by if inches wide. CHEMICALS REQUIRED. Gasolene, sp. gr. 86° B., which leaves no residue. Absorbent paper exercises a selective action on the constituents of mills, so that the fat is left on the surface of the paper mixed with only about one third of the non-fatty solids, and hence it is more easily extracted. The strip of paper may be pinned by one corner, so as to hang free ; 5 cc. of the milk to be tested are run on to the upper end from a pipette. When dry the paper is carefully rolled into a coil, placed in the extractor, and treated with gasolene for about two hours ; or the coil may be made first, and the milk run on to it. The flask, previously tared, is weighed after the evaporation of the gasolene, and the increase in weight is the fat. 3. The Babcock Method. The fat is freed from the other constituents and collected in the graduated neck of the bottle for measurement. APPARATUS REQUIRED. Centrifugal machine; whirling bottles ; 17.6 cc. pipette ; 17.5 cc. graduate ; wash bottle containing boiling water. CHEMICALS REQUIRED. Sulphuric acid, 1.835 ^P- S''- 17.6 cc. of the milk to be tested are delivered from a pipette into a long-necked graduated whirling bottle. 17.5 cc. sulphuric acid, 1.835 ^P- &•"•' ^'"^ gradually added, with vigorous shaking after each ad- dition of acid. The sulphuric acid should blacken the milk and cause a considerable amount of heat. After the acid is added and before the bottles are allowed to cool they should be whirled. They are to be placed in opposite pockets, in even numbers, and whirled for six to seven minutes, the large wheel making eighty to ninety revolutions per minute. The bottles are then removed, and the hot water, which should be ready by this time, is added to each bottle until the surface of the 8 mixture rises nearly to the top of tlie graduations on the neck. The bottles are again placed in the whirler, and turned at about the same rate for one minute. At the end of this time the bottles can be taken out and the length of the column of fat measured by dividers, one point of which is placed at the bottom and the other at the upper limit of the fat. If one point of the dividers is then placed at the zero mark of the scale on the bottle used, the other will be at the per cent, of fat in the milk examined. RELATION BETWEEN SPECIFIC GRAVITY, FAT, AND SOLIDS IN MILK. The specific gravity of milk is, in the main, a function of two fac- tors, namely, the percentage of solids not fat and that of the fat. The former raises it ; the latter lowers it. Taken by itself it afTords very little indication of the composition, but if any other item be known it should be possible to find, by calculation, the other quantities, pro- vided the assumption is true. The solids not fat are made up of sev- eral fluctuating constituents, but "normal milk " seems to contain them in such a constant ratio that a calculation serves at least to detect an abnormal sample. Specific gravity Of fat. •93 Of casein and albumen Of sugar. 1-34 1.65 Of ash. 3-0 Ash. Proteid. Sugar. Ratio in normal milif 2 9 13 EXAMPLE. Given the specific gravity and solids, to calculate the fat: Specific gravity = Gr. The amount which each per cent, of solids not fat raises the specific gravity = ^. The amount which each per cent, of fat lowers the specific gravity = f. Total solids = T. Solids not fat = 6". Fat = 7^ Gr =z Ss — Ff; or, substituting for 6' its value T — F, Gr ^:^ {T — F)s — Ff. The uncertainty of the calulation lies in the values of s and y^ which have not been quite satisfactorily deter- mined. 6 G The simple formula /^ = T — — answers within the limits of 5 4 experimental error for normal milk, but not for skimmed or watered milk. Data: Gr = 1.0323. G = Gr — i X 1000 = 32.3. T= 12.90. g 22 1 Calcul.ited. Found. ~jF = 12.90 — ~ = 4.02 3.99. A similar relation has been worked out for the proteids and sugar, so that from three determinations the whole composition may be cal- culated. Example as above : G Ash = .70 = A. Formula/^ = 2.8 T + 2.5.^ — 3.33/^" — .68 .. ^. ' or, F --= 36.12 + 1.75 — 13.32 — 21.28 = 3.27. Sugar = r — {A -\- P i- F). Sugar = 12.90 — (.70 + 3.27 + 4.02) = 4.91. CHIEF REFERENCES. Analyst, vol. vii, p. 129. Analyst, vol. xiii, pp. 26 and 49. With Chart. Analyst, vol. xv, p. 170. Analyst, vol. xvii, p. 169. Analyst, vol. x.x, pp. 7 and 57. SUGAR. /. By Titration with Fehlin^s Solution. APPARATUS REQUIRED. Pipettes 25 cc. and 5 cc, also i cc. divided into hundredths; 50 cc. graduate; 250 CO. bottle ; 500 cc. graduated flask ; burettes; thermometers; water bath ; beakers; 4-inch casserole or a 6- inch porcelain dish ; funnels; folded filters; medicine dropper ; extra small filters. CHEMICALS REQUIRED. 69.28 grams C. P. cojjper sulphate to i liter water; 346 grams sodium potassium tartrate and 80 grams sodium hydrate in i liter; acetic acid, 25 per cent, solution; potassium ferrocyanide, 1:50 made the day it is to be used; aluminum hydrate, as used for chlorine in water. See notes on Water Analysis. Any method of clarification may be used. The following answers very well. 25 CC. of milk are measured into a 250 cc. bottle ; 15 cc. of aluminum hydrate, 75 cc. of hot water, and 0.5 cc. of acetic acid (25 per cent, solu- tion) are added. The bottle is tightly stoppered, shaken vigorously, and lO placed on its side to allow the precipitate to settle, after which the almost clear liquid is decanted into a beaker. The precipitate in the bottle is washed three times with hot water, by decantation, the washings being poured into the beaker. The contents of beaker and bottle are then thrown upon a 6-inch filter, and the precipitate is washed until the volume of the filtrate reaches 500 cc. Titration. — z^ cc. of the copper solution and 5 cc. of the alkaline tartrate are accurately measured out into the 4-inch casserole or 6-inch porcelain dish, diluted with 40 cc. of water, and heated to boiling. The whey as above prepared is added from a burette as long as a blue color is seen in the liquid, which must be kept constantly boiling and made up to ^O cc. When the end-point is nearly reached a test for copper should be made in the solution. To this end a few drops of the liquid are run from a medicine dropper through a very small filter into a test-tube, or on to a porcelain plate, containing a dilute solution of potassium ferro- cyanide strongly acidulated with acetic acid, when, if copper be present, the characteristic rose coloration will appear. This will give the approx- imate number of cubic centimeters required to decolorize the copper solution. The exact number may be most conveniently found by adding the quantity of whey above used to a fresh portion (10 cc.) of Fehling's solution and 40 cc. of water, boiling exactly two minutes, filtering the whole through a 4-inch filter, and testing the filtrate as above described. If copper be still present, the operation is repeated with 0.2 cc. more whey until the end-point is reached; or, if the drop of solution be color- less, 0.2 cc. less of the whey is used each time until copper appears. If 10 cc. of Fehling's solution of the strength given are reduced by 0.067 500 X .067 gram of milk sugar, then -r; \ 3 = gram^ of milk sugar in ° ° ' No. cc. whey used ^ ° 25 cc. of the milk. The results are reported in per cent. From 27 to 2)2^ cc. of whey are usually required to reduce 10 cc. of Fehling's solu- tion. A standardization of the reagent with pure milk sugar should be made. The cuprous oxide formed may be filtered off, washed, dissolved in nitric acid, and the copper determined by the battery, in which case larger quantities may be employed. REFERENCES. Colby, New York State Bd. of Health Reft, 1882, p. 611. Beckmann, Fres. Zeit., xxv, 529. Beckmaiin, 7 lie Analyst, xi, 235. II 2. By the Saccharimeter. The necessary clarification is made either with basic ]ead acetate, acid mercuric nitrate, or mercuric iodide, and the resulting whey polar- ized. (See Bulletin No. 28, U. S. Department Agriculture, p. 208, 1890.) NITROGENOUS MATTERS CASEIN, ALBUMEN, ETC. 5 grams of milk are used and treated as in the notes upon the " Determination of Nitrogen by the Kjeldahl Process," which see. ASH. The residue in the platinum dish from the extraction of fat accord- ing to Method i is ignited at a low red heat. ADULTERANTS. Water. — The presence of nitrates indicates that the milk has been watered. Uffelman {77ie Afialyst, x, 146) recommends the follow- ing method for their detection. Diphenylamine the size of half a pea is dissolved in 25 drops sulphuric acid in a 3 inch porcelain dish. A few drops of the suspected milk are allowed to trickle down the sides of the dish, when, if nitrates be present, at the point of meeting a bluish stripe will form and ultimately tint the whole mixture. The picric acid method (see notes on Water Analysis) may also be used. Determination of Water Added to Milk. For a quantitative estimation of the amount of added water Radliscu {Mitth. a. d. p/iartn. Inst. u. Labor, d. Univ. Eriatigen, Hi/ger, Heft, iii, pp. 93-112) recommends the determination of the specific gravity of the whey, or " serum," as he terms it. APPARATUS REQUIRED. loocc. pijjette ; 2 cc. pipette; beakers; funnels ; vvate: Isatli ; thermometer; filters; Westphal balance. CHEMICALS REQUIRED. Acetic acid, 25 per cent, solution. 100 CC. of the milk are thoroughly mixed in a beaker with 2 cc. of the acetic acid, and heated in a water bath at 85° C. for five to ten minutes. The casein is by this treatment precipitated as a compact cake, and is easily filtered off. The contents of the beaker are now filtered, care being taken to bring as little of the precipitate upon the 12 filter as possible ; the filtrate, carefully mixed, is cooled to 15° C, and its specific gravity taken by the Westphal balance. Radliscu finds that the specific gravity of the whey or serum in a normal milk is never below 1.027, that it contains 6.3 to 7.5 per cent, total solids, of which .22 to .28 per cent, are fat. The addition of each 10 per cent, of water lowers the spe'cific gravity by .0005 to .0010, and the percentage of total solids from 0.3 to 0.5 per cent. OTHER SUBSTANCES. Salt. — Detected by the high percentage of ash, and determined volumetrically by silver nitrate after clarification with alumina. Caiie Sugar. — Detected by the polariscope. Starch. — 5 cc. of the milk warmed in a small beaker are treated with a few drops of iodine solution when the characteristic blue color indicates starch. Niter. — Detected in the ash. COLORING MATTERS. Annatto, caramel, and carrot extract are the substances usually employed to color watered milk. Thoms [Pharni. Zeit., xxxii, 59) found ultramarine in milk. A?inatto. — Dr. Davenport {Mass. State Bd. of Health Rep''t., xvi, 140) recommends the following method: A strip of filter paper is placed in the milk to be tested, previously made alkaline with carbonate of soda, and after being allowed to remain twelve hours is washed out. If annatto be present it will have acquired a salmon or light copper tint. This color is changed by sulphuric acid to dark blue ; by stannous chloride to a pink. Caramel. — See Fres. Zeit., xxiv, 30 ; The Analyst., x, 36. For other coloring matters see Leeds, '/Vie Analyst, xii, 150. PRESERVATIVES. Salicylic Acid. — A few drops of ferric chloride are added to 5 cc. of the milk. Should salicylic acid be present, a dirty purple to violet coloration is produced. Formalin or Fontiic Aldehyde. — This has recently come into use ; the odor usually betrays it. {Chem. News, vol. Ixxi, page 247.) Boric Acid. — Hilger {Nahr. und Genussniitteln) recommends the fol- lowing process: Five drops calcium hydrate are added to 10 (or 100) cc. 13 of the milk and evaporated to dryness on a water bath. The residue is charred, a few drops of water are added, the liquid is made slightly acid with hydrochloric acid, and filtered into a porcelain dish. The usual qualitative test with alcohol or turmeric paper is applied. (See also Kretschmar, The Analyst, xii, 159.) Benzoic Acid. — Fres. Zeit., xxi, 531 ; Jour. Anal. Chem., ii, 446. Carbonate of Soda. — Detected in the ash. Hydrogen Peroxide. — See Hilger, loc. cit., p. 58. See also Stokes, The Analyst, xvi, 123. BUTTER ANALYSIS. General Statements. — Butter consists of the fat of milk, together with a small percentage of water, salt, and curd; these exist in about the following proportions: Fat, 7S. 00-90.0 per cent.; average, 82 percent. Water, 5.00-20.0 " " " 12 " " Salt, 0.4 -15.0 " " " 5 " " Curd, 0.1 1- 5.3 " " " I " " The fat consists of a mixture of the glycerides of '' the fatty acids," which, when saponified, yield an increased weight of product, owing to the combination of water in the process. One sample gave 106.32 parts from one hundred parts of fat containing 12.5 per cent, of glycerine. (For further discussion see Analysis and Adulteration of Food, James Bell. On the constitution of butter fat see Analyst, vol. xvi, p. 161.) The fatty acids have a very complex composition, not yet perfectly understood. The Danish chemist Koefoed gives the following as the proportion of the acids thus far isolated : Oleic acid, C17H33COOH ) Per cent. Acid of the formula, C1SH28O4 f 34.0 Acid of the formula, C29H54O5 ) (The latter is probably a mixture.) Stearic acid, C17H35COOH 2.0 Palmitic acid, CisHsiCOOEi 2S.0 Myristic acid, C13H07COOH 22.0 Laurie acid, C11H23COOH 8.0 Capric acid, Cg H19COOH 2.0 Capryjic acid, C7 H15COOH 0.5 Caproic acid, C5 HuCOOH 2.0 Butyric acid, C3 H7 COOH 1.5 Analyst, vol. xvii, p. 130. The examination, from a sanitary standpoint, consists in the deter- mination of the quantity and condition of the curd and in the test for gelatin, since it is probably only in the possible decomposition of the nitrogenous portion that danger to health lies. It is now well recog- nized that the addition of oleomargarine is not injurious to health. It hns become, however, a not infrequent practice to secure an admixture of a large per cent, (as high as ;^^ per cent.) of curd or other nitrogenous material in certain fresh butters. If this mixture is eaten at once there is no danger; but on keeping, a decomposition occurs which is liable to produce serious effects. The percentage composition of the butter has, therefore, a sanitary as well as an economic aspect. The "aroma" of butter seems to be connected with the decomposi- tions caused by the growth of bacteria on the sugar and casein, and not by a change in the fats ; but there is no evidence that any unwholesome effect is produced. The usual examination consists in the determination of fat, water, salt, and curd. EXAMINATION OF THE FAT. The butter fat is isolated, saponified, and the volatile fatty acids are determined by distillation and titration, as in the Reichert method ; or they are washed out and the fixed fatty acids weighed, as in the Hehner method; or the weight of iodine required to saturate the acids is found by the Hiibl method. APPARATUS REQUIRED. Tall 30 cc. l)eakers; pipette with bulb, graduated to deliver 5.75 cc. water; 250 cc. round-bottomed flasks; no cc. graduated flasks; 100 cc. graduated flasks; 100 cc. graduated cylinder ; beakers ; funnels and filters. CHEMICALS REQUIRED. Absorbent cotton; sulphuric acid i : 40 ; potassium or sodium hydrate r : i; 95 per cent, alcohol redistilled from potassium hydrate; IJ, sodium hydrate; ^^ barium hydrate; phenolphthalein 1:500 in 90 per cent, alcohol; bits of ignited pumice dropped while hot into water and bottled for use. Procedure. — A piece of butter — about a cubic inch — is melted in a small narrow beaker placed in the water bath. After about fifteen minutes, during which time the temperature should have risen not above 60° to 70°, the water, salt, and curd will have settled to the bottom, and the clear fat may be decanted into a similar beaker through absorbent cotton or asbestos, care being taken that none of the water or curd is 15 brought upon the filter. When the filtered fat has cooled to about 40° the pipette is placed in the beaker and the whole weighed. By means of the pipette about 5 grams of fat are transferred to a dry 250 cc. round-bottomed flask, the pipette replaced in the beaker, and the whole again weighed. The difference in weight gives the exact quantity of fat taken. It is a saving of time, if several portions are to be weighed out, to make the weights one after the other, so that one weight will sufifice for a determination. Weigh off thus : Two portions of 5 grams each into the round-bottomed flasks for Reichert's method, a portion of 2.5 to 3 grams into a 500 cc. beaker for Hehner's process, two portions of about a gram each into 300 cc. bottles for Hiibl's process. /. Reichert-Meissl Number for Volatile Fatty Acids. To the fat in the 250 cc. round-bottomed flnsks are added 2 cc. of the caustic potash and 10 cc. of 95 per cent, alcohol. (The addition of 3 cc. ether, and then a vigorous boiling, reduce the time of saponifi- cation to a few minutes, or even seconds.) The flasks are then con- nected with long, straight glass tubes, serving as return flow condensers, and placed upon the water bath until saponification is complete ; fre- quent shaking hastens the process. When this has taken place the flasks are disconnected and the alcohol evaporated. After the complete removal of the alcohol, 100 cc. of boiled distilled water, at a temperature of about 50^^, are slowly added from a graduate, and the soap dissolved by gently warming on the water bath. Rapid addition of water may de- compose the soap, setting free the fatty acids : C17H35COOK + HoO = C17H35COOH ^ KOH. When the solution of soap has cooled to 60° or 70°, it being per- fectly clear, 50 cc. of the dilute sulphuric acid are added to set free the fatty acids. Two bits of pumice are dropped into the flask, which is then closed by a cork lied in with twine and immersed in boiling water until the fatty acids have melted to an oily layer floating on the top of the liquid. It is then cooled to 60°, the cork removed, and the flask attached to the condenser. The distillation should be so conducted that 110 cc. come over in thirty minutes. The distillate, after being thoroughly mixed, is poured through a dry filter, and 100 cc. are titrated with y^ barium or sodium hydrate, using phenolphthalein as an indicator. The number of cubic centime- i6 ters of alkali used is increased one tenth, and the weight of fat cor- rected for any number greater or less than 5 grams. For example, if 5.3 grams butter fat are used, and 100 cc. of the distillate require 27.4 cc. Ba(0H)2, iiocc. would require 27.4 -|- 2.74 = 30.14 cc. Then 5.3 : 30.14 :: 5 : a\ x ^= 28.4. X is the Reichert-Meissl number. Butters which give a number over 27 and under 30 may be considered genuine. Those which give a number between 25 and 27 are somewhat doubtful, and those which give less than 25 should be looked upon with suspicion. French Buttek. English Buttek. Fresh. Salt. Salt, Fat Water Solids not fat Salt : Reicliert luimber 84-39 13.98 1. 51 .12 29.1 83-44 12.86 ..63 2.07 29.1 82.9S 13-99 0.89 2.14 28.1 REFERENCES. Meissl, Dingier folyt. Jojtr., ccx.x.\i, 47S. " Zeit. Anal. C/iem., xviii. 63. Sendtner, Archiv. fiir Hygiene, i, 137 ; viii, 424. WoUney, Analyst, xii, 203; xiii, 8. 2. He/iner's Method for Direct Deter tnination of the Fixed Fatty Adds. APP,4RATUS REQUIRED. 500 CC. beakers ; funnels ; " weighing beakers; " filters dried over sulphuric acid. CHEMICALS REQUIRED. Potassium hydrate as in t ; hydrochloric acid, 1.12 sp. gr. ; 95 percent, alcohol. The portion of 2,5 grams weighed out into the 500 cc. beaker is saponified by using i cc. potassium hydrate diluted with 20 cc. 95 per cent, alcohol. As it is not essential to prevent the escape of the volatile acids, the precautions in Method i are unnecessary, but loss by spurting must be avoided. The aqueous solution of the soap, which should be 300 to 400 cc, is decomposed by 10 cc. hydrochloric acid and heated in a water bath almost to boiling until the clear oil floats. The beaker and contents are allowed to become quite cold ; the clear liquid, and finally the solid fats, are brought upon the thick weighed filter. When the beaker and fat are 17 well washed with cold water, the adhering fat is washed out with boiling water, which is poured through the filter, taking care that the filter is never more than two thirds full. The funnel is cooled by plunging it into cold water, the filter removed, placed in a weighing beaker dried at ioo° and weighed as soon as practicable, since long heating causes oxidation. A weight constant within 2 mgs. is sufficiently accurate. Eighty-seven and one half per cent, is usually taken as the propor- tion of fixed fatty acids in butter; 88 and 89 per cent, have been fre- quently found. All other fats yield from 95 to 96 per cent, insoluble fatty acids. REFERENCES. Bitlter^ its Analysis and Adulteration, Hehner and Angell. Uietzell and Kresner, Fres. Zeit., xviii, 83. Heintz, Fres. Zeit., xvii, 160. Heliner, FVes. Zeit.f xvi, 149. J. Method of Baron Hubl. This depends upon the fact that certain of the fatty acids, notably the "unsaturated acids," as oleic acid, C17H33COOH, take up the halo- gens with the formation of addition-products. ATPARATUS REQUIRED. 300 cc. glass-stopi^eied white glass bottles; glass-stoppered burettes; graduates. CHEMICALS REQUIREt). 25 grams iodine in 500 cc. 95 per cent, alcohol ; 30 grams mercuric chloride in 500 cc. 95 per cent, alcohol; the two are mixed, and after standing 24 hours are filtered; f;, sodium thiosulphate (24.6 grams to i liter); starch paste i '. 200; potas- sium iodide 150 grams to i liter; potassium bichromate C. P. 3.874 grams to i liter; strong hydrochloric acid; chloroform. The butter fat, weighed into the 300 cc. bottles, is dissolved in 10 cc. of dry chloroform. About 30 cc. — in the case of a doubtful butter 50 cc. — of the iodo-mercuric solution are accurately measured from a glass-stoppered burette, and the bottle is allowed to stand, with frequent shaking, for three hours in a dark closet. About 100 cc. of distilled water, with 20 cc. potassium iodide, are then added — the latter for the purpose of keeping in solution mercuric salts — and the excess of iodine uncombined with fat is titrated with ^^^ thiosulphate. When the solution has become faintly yellow, a few drops of freshly prepared starch solution are added and the titration continued to disappearance of the blue color. i8 The results are calculated in grams of iodine absorbed by loo grams of fat. This is frequently called the Hiibl or Iodine absorbtion number. Standardization of the lodo-Merairic Solution. — Since the strength of this solution is liable to change, the relation between it and the thiosulphate should be determined by carrying through a blank in the same manner and with the same quantities of reagents as in the case of the fat. Stafidardization of the Thiosulphate Solution. — As this is not per- n)anent, its strength should be determined by means of the standard solution of potassium bichromate, i cc. of which liberates o.oi gram of iodine. About 20 cc. of potassium bichromate are accurately measured from a burette into a 300 cc. glass-stoppered bottle, 10 cc. of potassium iodide and 5 cc. strong hydrochloric acid added, and the iodine which is set free is titrated with thiosulphate as above directed: KoCrsO; + HHCi + 6KI = 8KC1 + Cr-jCls + 7H..O + 61. CALCULATION OF RESULTS EXAMPLE. STANDARDIZATION OF THIOSULPHATE SOLUTION. 17.2 CC. thiosulphate = 21.5 cc. bichroniaie = 0.215 gram iodine. I cc. thiosulphate = 0.0125 gram iodine. STANDARDIZATION OF IODINE SOLUTION. 31 cc. ioditie solution = 46 5 cc. thiosulphate. I cc. iodnie solution = 1.5 cc. thiosulphate. If 31 cc. iodine solution have been added to 1.049 g'ams of fat, then 31.0 X 1.5 = 46.5 cc. is the equivalent amount of thiosulphate solution; and if 19.4 cc. tliio- sulphate were used to titrate excess of free iodine, 46.5 — 19.4 = 27.1 cc. is the amount of thiosulphate equivalent to iodine combined with the fat. Then since i cc. thiosulphate is equivalent to 0.0125 free iodine, ~-Ll '. — :iJ X 100 = 32.29 grams I 049 of iodine combined witii 100 grams fat. It is assumed that 100 grams pure butter absorb 30 to 40 grams iodine; artificial butter, 55 grams; oleomargarine, 63 to 75 grams; olive oil, 83 grams ; and cottonseed oil, 106 grams. REFERENCES. Hiibl, Dinnler polyt. Jour., ccliii, 281. Join-. Soc. Chem. Ind., iii, 641. Pattinson, y : 7'== :: ^= : i^. 7958 : x :: 760 : (766 — 18.5) (18.5 = tension of aqueous vapor at 21°). z'^ = 7827 = capacity of bottle at 0° and 760"""-. i cc. CO., at 0° and 760"""- weighs ^ 10.8 c.co ^ ^^^ 1.9643 mgs. = 5.50 cc. -^-^ — = 7.02 parts COo per 10,000. 1.9643 7827 Note. — T7tii> samples are to be taken following the notes closely, and the results cal- culated by both methods before collecting more samples. Then some one room may he taken, and the quality of the air determined for the different hours of the day, or the same hour different days of the week, or a comparison of different rooms may be made, or a building may be tested as a whole. In making out the report of results they should be arranged in tabular form, attention being paid to the following points: Room, date, weather, barometer, time, place in room, number of people, or gas jets burning in room, and the condition of the doors, windows, and transoms. REFERENCE. Jour. Analyt. and Applied Chem. vol. vi., ;;63. 29 BIBLIOGRAPHY OF THE CHEMISTRY OF FOODS. A bibliography complete to 1882 may be found in the Second Annual Report of the New York State Board of Health. Some of the important works published since that time are given below: Allen, " Commercial Org. Analysis," 2d ed, ; Bell, "Analysis and Adulteration of Food ; " Church, " Food " (South Kensington Science Handbook); Battershall, "Food and its Adulterations;" Atwater, " Chem. and Economy of Food" (U. S. Dept. Agric. Bull. 21, 1895); Richards, " Food Materials and Their Adulteration;" Atkinson, " Sci- ence of Nutrition ;" Blyth, " Foods: Their Composition and Analysis;" Koenig, " Chemie der menschlichen Nahrungs-und Genussmittel," 3d auflage ; Hilger, " Vereinbarungen bet. d. Unters. u. Beurteilung v. Nahr. u, Genussmittel ; " " Bibliothek fiir Nahrungsmittel-Chemiker ; " especially, Rottger, " Kurz. Lehrb. d. Nahrungsmittel-Chemie ; " Eph- raim, " Originalarbeiten iiber Anal. d. Nahrungsmittel ; " Bujard and Baier, " Hilfsbuch fiir Nahr. Chemiker;" U. S. Dept. Agric. Bulletins 13, 43; Boards of Health Reports; Sadtler, " Indust. Org. Chem.;" Wiley, "Agricultural Analyses;" Addyman, "Agricultural Analysis." Periodicals — "Arbeiten aus d. Kaiserl. Gesundheitsamte zu Berlin;" "Analyst;" "Jour. Soc. Chem. Indust.;" " Vierteljahrsschrift fiir Chem. d. Nahr. u. Genussmittel ;" "Milch Zeitung;" "Deutsche Viertelj. fiir off offentl. Gesundheitspflege ; " " Rapports du Laboratoire Municipal Paris." LABORATORY NOTES ON WATER ANALYSIS. (Prepared for the use of students in the Laboratory of Sanitary Chemistry of the Massachusetts Institute of Te'chnology. Not pubHshed.) CLASSIFICATION OF WATERS, The examination of a water in order to determine its fitness for domestic use (a so-called sanitary analysis) comprises the determina- tion of three points : first, the amount, if any, of organic matter in a living or dead condition suspended or dissolved in the water ; sec- ond, the amount and character of the products of decomposition of organic matter, and their relative proportions to each other ; and third, the amount of certain mineral substances dissolved in the water. From these results we draw conclusions as to the present condition and past history of the water. To facilitate this examination waters may be divided into three classes: first, brook, pond, and river water — so-called surface water; second, spring and deep well water ; third, shallow wells and sewage efifiuents. The waters of the first class found in New England are gener- ally more or less colored, and contain more or less suspended organic life and its debris, which often impart a decided odor to the water. These waters, draining for the most part wooded and sparsely popu- lated regions, are low in ammonia, nitrites, and nitrates ; low, also, in mineral salts, and with only a slight excess of chlorine over the normal. They are usually high in organic matter and albuminoid ammonia. The waters of the second class are generally odorless, colorless, without suspended matter or organic matter in solution, low in nitrates, and with nearly normal chlorine, but with higher mineral substances than surface waters. Waters of the third class present the greatest variety. They may be as clear and colorless and as free from organic matter as the second class, and they may contain more organic matter than some waters of the first class. As a rule, nitrates and chlorides, as well as mineral salts, are high ; ammonia and nitrites may or may not be high. It is always desirable to knov/ something about the origin and character of a water before beginning the analysis, and the foregoing general classification will be found helpful in planning the analysis. In addition to classifying a water in this way it will also be found useful to make three qualitative tests : first, with Nessler reagent to see if much free ammonia is present ; second, with silver nitrate for the amount of chlorine ; and third, with phenol-disulphonic acid for nitrates. Since the condition of the organic matter and the relation of the several products of decomposition to each other are constantly cfiang- ing, the determination of these should be begun without delay. The determination of the so-called " free ammonia " is first made after distilling the apparatus free from ammonia. Then follows the distilla- tion with alkaline permanganate in order to liberate the nitrogen in the undecomposed organic matter ; this product is called " albuminoid ammonia," The determination of the total combined nitrogen by the Kjeldahl process is to be carried on in waters of Class I at the same time if practicable. The test for the amount of nitrogen in the second product of decay, nitrites, is also made as soon as possible, and lastly the re- quired quantity of the sample is s§t aside to be evaporated for the determination of nitrates. The odor of the water is at times a valuable aid in judging the condition of a surface or well water. This should be determined on the first day and while the bottle is at least half full. It often happens that the color of the water gives also valuable information as to its condition. The examination for turbidity and sediment is made on the morn- ing of the second day after the bottle has stood over night. The color is estimated at any convenient time during the two days. The deter- mination of the carbon in the organic matter, which is in such a con- dition as to be oxidized to CO2 by the Kubel method by treatment with hot, acid, potassium permanganate solution, is next in order, and the results are expressed as " oxygen consumed." The final test for the presence of organic matter in waters of Class I is made by igniting the total solid residue on evaporation. This determination has little or no value in the case of Classes II and III, and is omitted as a rule. The residue on evaporation is often used for the determination of iron. The determination of chlorine is most essential in deciding upon the history of a water, since chlorine is not taken up by plant life as nitrogen is, and, being soluble in all combinations, it remains in the water when once it is there. Hardness is determined by means of a soap solution, and ex- presses approximately the amount of calcium and magnesium salts present. The determinations of the fixed solids, hardness, chlorine, nitrates, and iron, taken together, give a good idea of the character of the mineral matters dissolved in the water. Having obtained these several results, there remains the decision as to the present condition and past history of the water as shown by these tests. Briefly, the questions to be answered are these : I. Is the water a normal unpolluted water of its class ? that is, has it any more of any substance than it has a right to contain by virtue of its source ? A brook draining a meadow or swamp has color, gives high albuminoid ammonia and high oxygen consumed, but it is not on that account a " polluted " water. A well water may contain high nitrates and chlorine and yet be free from present pollution. A deep well water may contain high free ammonia without giving any evidence of pollution in historic time. To answer the question, there- fore, one should know the locality and surroundings of the water, its source, the normal chlorine, the ratio of the nitrogen compounds to each other, and its character as regards the permanence of the or- ganic matter it contains. II. Is the water, if normal, a good water for general domestic uses ? that is, is it hard or soft, has it much or little iron, has it any disagreeable odor? III. Is the water in any case safe for drinking ? To answer this question there is needed a knowledge wider than a chemist's of the relation of decaying organic matter and of the germ-carrying power of water to outbreaks of disease. To the chemist's knowledge must be added, therefore, that of the biologist, the engineer, and the sanitarian. METHODS AND REACTIONS. FREE AND ALBUMINOID AMMONIA. Since the condition of the nitrogen, as well as the relative propor- tion of each of the four forms in which it occurs, is one of the most important points to be decided, the determination of the first product of decay, the so-called free ammonia, is begun at once. The test is one of much delicacy, and requires the greatest care and cleanli- ness of manipulation. The top of the bottle is first rinsed under running water to free it from any possible dust. Under no circum- stances must the inside of the neck of the bottle or the stem of the stopper be touched by the hand or wiped with a cloth. After the contents of the bottle are well shaken, in order that an average sample may be obtained, a measured portion is distilled in the apparatus shown in the accompanying cut, after it has been freed from all traces of ammonia by dis- tilling in it ammonia-free water. Of Classes I and II 500 cc. are usually taken; of Class III 500 or less — sometimes only 10 cc. are used, according to the result of the qualitative test. To this latter class of waters about 0.5 gram of sodium carbonate is added to be sure that the reaction is not acid. Three portions of the distillate, of 50 cc. each, are caught in graduated flasks and set aside. If there is very much free ammonia present 200 cc. are distilled over. The time of distilling 50 cc. should not be more than eight or less than five minutes. After the free ammonia has been distilled ofiE and the contents of the flask have slightly cooled, 40 cc. of alkaline permanganate are added through a funnel, taking care that no alkali touches the neck Scaie,ij^in.=ifoot. of thc flask, and the distillation of 5 the albuminoid ammonia, that is to say, the determination of the nitro- gen of the undecomposed organic matter, is proceeded with. For Class I five portions of 50 cc. each are obtained ; for Classes II and III only four portions are taken. This process gives about one half of the total combined nitrogen in the waters of Class I. The contents of the 50 cc. receiving flasks are transferred to Nessler tubes to be compared with standards which are prepared as follows : To the Nessler tubes nearly filled with water free from am- monia is added varying quantities, for instance, 0.3, 0.5, 0.7, i.o, 1.3, 1.5, 2.0, 2.5, 4.0, 6.0 cc. of a standard solution of NH4CI containing .00001 gram NH3 in i cc. The contents of the tubes are rotated (never shaken like test tubes or stirred with a rod), allowed to stand two or three minutes, and 2cc. of the Nessler reagent added to the whole set and to the samples to be tested as rapidly as possible. At the end of ten minutes the colors are matched and the amount of ammonia recorded. As an example of a water of Class I may be given the following results from distilling 500 cc. : FREE AMMONIA. 1st 50 CC. 0.7 CC, 2d 50 cc. 0.3 CC. 3d 50 cc. 0.0 CC, ALBUMINOID AMMONIA. 1st 50 CC. 4.5 cc. 2d 50 cc. 2.8 cc. 3d 50 cc. 1.5 cc. 4th 50 cc. 1.0 cc. 5th 50 cc. 0,5 cc. In this case the free ammonia would be .0020, and the albuminoid ammonia .0206 parts per 100,000. The compounds produced by action of ammonia on mercuric so- lutions are considered as substitutions of iHg for 2 H in NH4 and are called mercur-ammoniums. Tetra-mercur-ammonium iodide (NHgj I) is a brown precipitate soluble in excess of KI in presence of KOH with a brown yellow color, proportionate to the amount of NH3. NH3 -f (2Hgl2 -f 2KI + 3KOH) = NHg.I -f sKI + 3H2O. If practicable, a determination of the total nitrogen in the organic matter should be carried on at the same time by the Kjeldahl process. (See page 7.) NITRITES. As soon as the foregoing determinations are well started the test for the second product of decay, the nitrites, is made. Waters of Class I must be freed from color by milk of alumina, and filtered through filters washed with nitrite-free water. loo cc, of the colorless water are treated in special tubes with three reagents, added in the following order and quantities : i cc. HCl i : 4, 2 cc. sulphanilic acid, 2 cc, naphthylamine hydrochlorate. The filtration and determination should be carried on within the half hour, since the air of any room in which gas is burned contains nitrites. The pink color of the azo a amidonaphthylic parabenzolsulphonic acid, N N H c c c / ^ / \ / w H — C C H — C C C — H II 1 1 1 1 H — C C H— C C C— H \ // ^ / \ ^ c c c / 1 1 S03H NH2 H which is formed when nitrites are present, is compared with that of known amounts of standard nitrite solution, containing .0000001 gram N in I cc. Sewage effluents and waters from bad wells often need to be diluted, 10 cc, or even i cc, being made up to 100 cc. before adding the reagents. * NITRATES. Nitrogen in the third stage, that of nitrates, is next determined. In case of colorless well waters, i cc. and 2 cc. are measured out with a capillary pipette into 2|-inch porcelain dishes and set away to evaporate spontaneously in a place free from dust. For surface waters, always low in nitrates, 10 cc. are taken from the decolorized portion filtered for nitrites, and the dishes are placed on the top of the water bath until the bulk is reduced to about 2 cc, when they are set away to evaporate spontaneously. This determination depends on the color given by an ammoniacal solution of trinitrophenol (picric acid), which in this case is formed by the action of the nitrates contained in the cold, dry residue upon the phenol-disulphonic acid with which it is moistened. Six or eight drops of the acid are run directly upon the residue and carefully rubbed into it with a short glass rod to insure complete contact of the acid and the residue in the dish. After a moment or two 7 cc. of water are added, and 3 cc. of an alkali (preferably ammonia), to distinct alkaline reaction. The yellow color thus produced is compared with standards, i cc. •oooooi gram N, in the same way as in the two previous tests.' HO HO I I C C / ^ / \ H — C C— SO3H NO2 — C C— NO2 II I +3 HNO3 = II I + 2 H2SO4 H — C C — H H — C C — H \ -^ \ ^ C C + H2O I I S03H N02 phenol- disulphonic acid. picric acid. KJELDAHL PROCESS FOR TOTAL ORGANIC NITROGEN. Five hundred cc. of the water are poured into a round-bottomed flask, of about 900 cc. capacity, and boiled until 200 cc. have been distilled off. The free ammonia which is thus expelled may, if desired, be determined by connecting the flask with a condenser. To the water remaining in the flask is added, after cooling, 10 cc. of pure concentrated sulphuric acid. After mixing, the flask is placed in an inclined position on wire gauze, on a ring-stand or other convenient support, and boiled cautiously, until all the water is driven off and the concentrated sulphuric acid is white or a very pale yellow. After cool- ing, 200 cc. of water free from ammonia are added, the neck of the flask being washed free from acid, and then 100 cc. of sodium hydrate solution. The flask is immediately connected with the condenser and then shaken to mix the contents. The distillation at the start is conducted rather slowly. After the first 50 cc. are condensed, the contents of the flask may be boiled more rapidly until 150 cc. to 175 cc. have altogether been collected. The total distillate is made up to 250 cc. with water free from ammonia, well mixed, and 50 cc. taken for nesslerization. The sulphuric acid oxidizes the carbon of the organic matter, thus liberating the nitrogen in the form of NHg which remains as (NH4)2SOi until released by distillation with NaOH. The Kjeldahl process is very much simpler as applied to waters than it is for many organic substances — flour, for instance. The organic 'References: Sprengel, Pogg. Ann. 121, p. 188; Grandval and Lajoux, Compt. Rend. loi, p. 62; Fox, Tech. Quar. i, p. i ; Hazen & Clark, Jour. Anal. & App. Chem. 5, p. I, Gill, Jour. Am. Chem. Soc. 16, 122. 8 matter in surface waters can be easily oxidized by sulphuric acid with- out the aid of potassium permanganate or mercury. CARBONACEOUS MATTER, OR " OXYGEN CONSUMED." KubeVs Hot Acid Method. loo cc. (or 25 cc. of very highly colored waters) are measured into a 250 cc. flask. 8 cc. of H2SO4 (i : 3), and 10 to 20 cc. of standardized, approximately — permanganate added, the whole boiled for five minutes and cooled one minute. The color is then discharged by 10 cc. of ex- actly — oxalic acid and the solution titrated with the permanganate to a faint permanent pink. Since i cc. of the oxalic acid corresponds to .00008 gram oxygen, the number of cubic centimeters of perman- ganate which have been used to oxidize the organic matter must be multiplied by its value referred to the oxalic acid as a standard. Example. 11 cc. permanganate solution are decolorized by 10 cc. oxalic acid; i cc. of permanganate has, therefore, a value of .000072 gram oxygen. 100 cc. of the sample of water-]- 10 cc. oxalic acid required 14 cc. permanganate. 14 cc. — II cc. = 3 cc. X .000072 =: .0002 16 oxygen ; or expressed in parts per 100,000 — .2160.. CHLORINE, Colored surface waters are treated with milk of alumina to remove the color as follows: About 750 cc. are poured, without measuring, into a flask holding at least one liter, 3 to 5 cc. of the alumina added, and the water brought to the boiling point. The flask is now set aside in an inclined position so that the alumina may settle out and allow of the decantation of two measured portions of 250 cc. each, which are con- centrated in six-inch evaporating dishes on the steam bath to 25 cc. A clean feather, moistened with distilled water, is used to rub the sides of the dish and loosen any adhering residue. Three drops of a neutral solution of potassium chromate are added, and the solution titrated (in the same evaporating dish in which it was evaporated) with silver nitrate, with or without the addition of sodium chloride. The latter solution has always a strength of 0.00 1 gram CI in i cc. The sil- ver nitrate is usually one half the value. Colorless waters of Classes II and III do not need to be treated with milk of alumina. If high in chlorine, 25 cc. may be titrated directly; if low in chlorine, 250 cc. are concentrated as above. RESIDUE ON EVAPORATION. A portion of the water to be examined (200 cc. of waters of Class I, 100 cc. of waters of Classes II and III) is evaporated in a weighed plat- inum dish on a water bath. After drying in an oven at ioo°C. for two or three hours, the dish is left in the desiccator over sulphuric acid for some hours. This gives the total weight of inorganic matter contained in the water, and in case of the waters of Class I of the organic matter as well. This latter may be burned off in the radiator, leaving "fixed solids " or mineral matter. HARDNESS. The amount of soap solution required to give a foam or lather with 50 cc. of water, which remains five minutes after shaking in a 250 cc. bottle, is read from the burette, and the corresponding quantity of cal- cium carbonate (or other salts) is taken from the table given in Sutton's Volumetric Analysis, page 370. DISSOLVED OXYGEN. Witikler' s'' Method. When water is taken from a faucet, a glass-stoppered bottle of known capacity, holding from 50 to 250 cubic centimeters, is filled by means of a tube which passes to the bottom of the bottle. A con- siderable amount of water is allowed to pass through the bottle and overflow at the top. In taking samples from streams or ponds a stop- per with two holes is used. A tube passing through one of these holes is sunk in the water to the desired depth, and the other is connected with a larger bottle of at least four times the capacity of the smaller one and fitted in the same way. From the larger bottle the air is exhausted by the lungs or by an air pump until it is nearly filled with water. Unless the determination is to be made at once the rubber stopper in the smaller bottle is quickly replaced by the glass stopper, so that no air is left in the bottle. In making the determinations a small amount of a saturated solu- tion of manganous sulphate is added with a pipette having a long capillary point reaching below the surface of the water, and in the same way a concentrated solution of potassium iodide and sodium hydrate. The glass stopper is now inserted, leaving no bubble of ' Berichte der deutsch. chem. Gesell., Vol. XXI, p. 2843. 10 air, and the contents well mixed. Strong hydrochloric acid is added, after most of the precipitate has settled to the bottom of the bottle. The contents of the bottle are now poured into a flask or other con- venient vessel, and the liberated iodine (in amount proportional to the amount of dissolved oxygen in the water) is titrated with thiosulphate. In calculating the amount of oxygen, a correction must be made for the volume of the reagents used, which should not be more than i per cent of the total volume. If the precipitate had settled before the acid was added no allowance should be made for the amount of acid, since the water it displaces contains no oxygen or iodine. If water is collected in the ordinary way and transferred to the apparatus by pouring, there will inevitably be an absorption of oxygen, unless the water is already saturated. Thus a process which gives excellent results when the water is nearly or quite saturated may fail entirely to give accurate results when the dissolved oxygen is low or absent. The results are reported in " percentage of saturation," the amount of oxygen which water will take up at the observed temperature being used as a basis. Winkler has calculated these amounts from o° to 30°.' It frequently happens that percentages greater than one hundred are obtained, due to the supersaturation of the water with oxygen.^ Determinations of dissolved oxygen in ponds and streams are best made on the spot. The very simple apparatus required for the Winkler process can be packed in small space, and the entire determination requires only a few minutes. The absorption of the oxygen by the manganous hydrate is complete almost at once, and it is unnecessary to allow it to settle for a long time before adding acid. The titration can be made with a small burette or pipette with accurate results. The temperature of the water at the depth from which the sample is taken is conveniently determined by means of a thermometer fitted by a doubly-perforated stopper to a bottle of about 500 cc. capacity, which has been filled with some of the Water and then lowered to the desired depth. IRON. Treat the residue on evaporation of 200 cc. of water with about 5 cc. of HCl (strong acid diluted with an equal bulk of water) on a water bath, being careful to carry the acid nearly to the edge of the dish to insure its contact with all the residue. When the residue is ' Berichte, Vol. XXII, p. 1772. * Technology Quarterly, Vol. V, p. 250. II completely dissolved (with the exception of silica), the solution is rinsed with a loo cc. tube and diluted to about 50 cc. A solution of perman- ganate is added drop by drop until the liquid remains pink for at least ten minutes. After the pink color has faded 15 cc. of a solution of potassium sulphocyanate (5 grams to the liter) is added, and after twenty minutes the color is read, using standards made up at the same time from a solution of which i cc. contains 0,000 1 gram of iron. The greatest care is necessary to prevent access of dust and to exclude any possible contamination ; also to avoid loss of ferric chloride by overheating, and in ignited residues to secure the complete solution of the iron. The results, with due precaution, have been very satisfactory.' ODOR. The odor of waters is obtained by shaking violently the sample in one of the large collecting bottles when it is about one half full, then removing the stopper and quickly putting the nose to the mouth of the bottle. An odor can often be detected in this way which would be en- tirely inappreciable if the water were poured into a tumbler. The odor which is given off when a water is heated is sometimes the same as the odor of the water when cold, sometimes it is different. The hot odor is obtained by heating on an iron plate about 200 cc. of the water in a beaker of 500 cc. capacity covered with a watch glass. The water is quickly heated until the air bubbles have all been driven off and the water is about to boil. The beaker is then taken off the plate, and, after cooling for about five minutes, it is shaken by a rotary movement, the watch glass removed, and the nose put inside the beaker. It is only for an instant, as a rule, that an odor can be perceived. COLOR. Most of the surface waters of the State have a yellowish-brown color more or less pronounced. The tint corresponds, particularly in the lower grades, very closely to that of nesslerized ammonia, so that the standards for reading ammonia can be used also for the deter- mination of the color. The comparison is made in the same kind of 50 cc. tubes that are used for the ammonia determinations, but the tubes used for this purpose are kept separate from those used for the ammonia, since the least amount of alkali remaining in a tube (if im- ' Thomson, Journ. Chem. Soc, Vol. XLVII, p. 493, 1885. 12 perfectly washed) alters the color of the water. The scale used corre- sponds with the amount in the standards. Thus a color of i.o is that corresponding to the nesslerization of i cc. of the standard ammonium chloride solution ; o.i is the color produced with o.i cc. of this solution. In the higher grades of color, over i. the tint varies considerably from the nesslerized ammonia, and the degree of color is then better deter- mined in wider tubes and in less depth. Standards made from very dark water from cedar swamps by various degrees of dilution, and veri- fied by direct comparison with nesslerized ammonia, are also used. TURBIDITY AND SEDIMENT. The suspended matter remaining in the water after it has rested quietly in the collecting bottle for twelve hours or more is called its turbidity, and that which has settled to the bottom of the bottle its sediment. Good ground waters are often entirely free from turbidity and sedi- ment, but surface waters are seldom free from suspended matter. The turbidity is very various in character and amount, sometimes milky from clay, but more generally it consists of fine pollen- like particles. These are generally living algae or infusoria, and a practiced eye can, not in- frequently, recognize their forms. Some of the lower animal forms can also be seen by the naked eye, and the larger Entomostraca are quite noticeable in many waters. The sediment may be earthy or flocculent, in the latter case it is generally dkbris of organic matter of various kinds. The degree of tur- bidity is expressed by the terms " very slight," " slight," " distinct," and " decided," and the degree of sediment by " very slight," " slight," "considerable," and "heavy." REAGENTS. Nessler Reagent. — Dissolve 61.750 grams KI in 250 cc. distilled water. Add 415 cc. of a cold solution of HgCl2 which has been satu- rated by boiling an excess of the salt and allowing it to crystallize out. Dissolve the slight precipitate of Hglj by adding 0.750 gram powdered KI. Then add 300 grams of KOH dissolved in 250 cc. of water. Make up to the liter and allow it to stand over night to settle. This solution should give the required color with ammonia within five min- utes, and should not precipitate within two hours. Alkaline Permanganate. — Dissolve 400 grams of refined potassium carbonate in 2 liters of distilled water and heat to boiling in a silver 13 dish. After removing the lamp add 250 grams of quicklime of the best quality piece by piece. When all the lime is slaked, heat to boiling again, and transfer to a large bottle or crock which can be tightly closed to exclude the air. Allow it to settle over night, siphon off the clear liquid, and add water uiftil the sp. gr. = 1.125. One liter of this caustic potash and 8 grams of permanganate crystals are boiled for thirty minutes, the water lost on evaporation being replaced. When the chemicals used are all good there should be no correction needed for ammonia in this solution. Standard Nitrite Solution. — The pure silver nitrite used in making this solution is prepared by the double decomposition of silver nitrate and potassium nitrite and repeated crystallizations from water of the rather insoluble silver nitrite, i.i grams of this silver nitrite are dissolved in water nitrite-free, the silver completely precipitated by the addition of the standard salt solution used in the determination of chlorine, and the solution made up to one liter. 100 cc. of this strong solution are diluted to i liter, and 10 cc. of this last solution again diluted to i liter. The final solution is the one used in preparing standards, i cc. = .0000001 gram of nitrogen. Sulphanilic Acid. — Dissolve 8 grams (Kahlbaum's) in i liter of water. This is a saturated solution. NaphthylatJiine Hydrochlorate. — Dissolve 8 grams in 992 cc. of water and add 8 cc. strong HCl. (Keep in the dark.) Standard Nitrate Solution. — 0.720 gram of pure KNO3 is dis- solved in I liter of water. Of this strong solution 10 cc. are diluted to I liter. I cc. of this dilute solution corresponds to .000001 gram nitrogen. Phenol- Disulphonic Acid. — Heat together 3 grams synthetic phenol with 37 grams pure concentrated H2SO4 on a boiling water bath for six hours. Sodium Hydrate. — For Kjeldahl process. Dissolve 200 grams good quality caustic soda in i liter of water. Boil with 3 grams of permanga- nate crystals to free the solution from nitrogen. Sulphuric Acid. — For Kjeldahl process. This should be free from nitrogen. May be obtained from Baker & Adamson, Easton, Penn. Calcium Chloride and Soap Solutio?is. — For hardness. Dissolve 0.200 gram of pure Iceland spar in dilute HCl, and evaporate several times to remove excess of acid. Dissolve the calcium chloride thus formed in i liter of water. Dissolve 100 grams best dry white Castile soap in 80 per cent alcohol. From this strong solution make a weaker solution (about 100 cc. to a liter of 70 per cent alcohol) of such a 14 strength that 14.25 cc. will give the required lather with 50 cc. of the above CaCL solution. Reagefits for Wmkkr's Process. — 360 grams of NaOH and 100 grams of KI in i liter of water. 48 grams of MnSOi -}- 4H2O in i liter of water. HCl sp. gr, 1-125. Thiosulphate y^ -f- 4 grams ammonia carbonate. Dilute to ^$0 for use. Standardize by potassium bichromate. The following books will be found on the laboratory shelves for consultation as to methods of analysis and interpretation of the results obtained : Wanklyn — Water Analysis. Frankland — Water Analysis. Nichols — Water Supply. State Board of Health Reports for Massachusetts. " " " " " Michigan. " ' " " " " " Illinois. National Board of Health Report for 1882. Tiemann-Gaertner — Untersuchung des Wassers. Fischer — Die Chemische Technologie des Wassers. The following papers on special topics relating to water supply and water analysis will be found in the reports of the Massachusetts State Board of Health, the Pro- ceedings of the Society of Arts, and the Technology Quarterly. They are also avail- able in pamphlet form : The Chemical Examination of Water. The Interpretation of Water Analyses. Chemical Examination of Drinking Water. Discussion of Special Topics relating to the Quality of Public Water Supplies. The Analysis of Water — Chemical, Microscopical, and Bacteriological. On the Determination of the Organic Nitrogen in Natural Waters by the Kjeldahl Method. On the Amount of Dissolved Oxygen contained in Waters of Ponds and Reservoirs at Different Depths. On the Amount of Dissolved Oxygen contained in the Waters of Ponds and Reser- voirs at Different Depths in Winter, under the Ice. The Odor and Color of Surface Waters. The Effect of the Aeration of Natural Waters. The Filtration of Natural Waters. On the Mineral Contents of Some Natural Waters in Massachusetts. The Purification of Water by Freezing.