REVISION OF THE AVIAN GENUS PASSERELLA WITH SPECIAL REFERENCE TO THE DISTRIBUTION AND MIGRATION OF THE RACES IN CALIFORNIA BY H. S. SWARTH University of California Publications in Zoology Vol. 21, pp. 75-224 UNIVERSITY OF CALIFORNIA PRESS BERKELEY 1920 Cornell Lab of Ornithology Library at Sapsucker Woods Illustration of Bank Swallow by Louis Agassiz Fuertes Cornell University Library QL 696.P2S97 Revision of the avian genus Passerellaw 3 1924 022 548 907 Laboratory of O ■ " •' <> 15« Sapsucker Viand* K086 Cornell Univertlty IttMca, New York 14851 All books are subject to recall after two weeks DATE DUE 1 GAYLORD PRINTED IN U.S A The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924022548907 n 597 UNIVERSITY OF CALIFORNIA PUBLICATIONS IN ZOOLOGY Vol. 21, No. 4, pp. 75-224, plates 4-7, 30 figures in text September 11, 1920 REVISION OF THE AVIAN GENUS PASSERELLA, WITH SPECIAL REFERENCE TO THE DISTRIBUTION AND MIGRATION OF THE. RACES IN CALIFORNIA BY H. S. SWARTH (Contribution from the Museum of Vertebrate Zoology of the University of California) CONTENTS PAGE Introduction 76 Material and methods of treatment 80 History. 84 Variation in Passerella iliaca 87 Distribution and migration 101 Passerella iliaca iliaca 114 Passerella iliaca altivagans 119 Passerella iliaca unalaschcensis 127 Passerella iliaca insularis 132 Passerella iliaca sinuosa ". 135 Passerella iliaca annectens 140 Passerella iliaca townsendi 144 Passerella iliaca fuliginosa 149 Passerella iliaca schistacea < 152 Passerella iliaca fulva 158 Passerella iliaca megarhynchus 161 Passerella iliaca brevicauda : 165 Passerella iliaca canescens 169 Passerella iliaca monoensis 170 Passerella iliaca mariposae 173 Passerella iliaca stephensi -. 176 Table of measurements 182 List of specimens examined 184 Literature cited 212 76 University of California Publications in Zoology [Vol. 21 INTRODUCTION The genus Passerella contains but one species, Passerella iliaca, the fox sparrow, confined to North America, and distributed over the greater part of the continent. In the eastern two-thirds of the general habitat, from northern Alaska to Newfoundland, the fox sparrow occurs as one relatively uniform subspecies, in the Pacific Coast and Rocky Mountain regions as a number of more or less widely differentiated local races. Sixteen subspecies are formally recognized in the present paper. All the forms of Passerella are migratory, most of them traveling long distances between their summer and winter homes. The eastern race, iUaca, occurs in Cali- fornia as a rare straggler in winter ; each of the other fifteen subspecies occurs more or lesss abundantly in California at one season of the year or another. The Museum of Vertebrate Zoology has gradually acquired a fairly extensive representation of fox sparrows, mostly collected by various field expeditions. This collection consists in part of series of most of the subspecies collected upon their breeding grounds, in part of miscellaneous non-breeders of various subspecies, taken as migrants and winter visitants. Of late years it has become increasingly difficult to coordinate the museum arrangement of many of these specimens with current classifications and descriptions. Inconsistencies were apparent but the remedy was not so readily seen. At the same time the Museum was being called upon constantly to identify specimens, sometimes single birds, sometimes extensive series, sent in for this purpose by collectors in various parts of the state; while every effort was made to satisfy such requests, the questions asked could not always be satisfactorily answered. It seemed desirable to bring together as large a series of Passerella as was to be had, in expectation that careful study of material known to be available would, among other things, permit of a clearer defini- tion of the subspecies and their characteristics than had yet been put forth. This study has proven far more complicated than was at first apparent, and while I feel that it has resulted in a much clearer perception of the nature of the problem, I cannot assume to have finally disposed of many of the -questions involved. In this connection my attention has been frequently called to the desirability of supplying bird students with a "key" or some similar 192 °] Swarth: Revision of Avicm Genus Passerella 77 means of identifying" specimens of Passerella, readily as captured. This, from the nature of the case, I fear can not be done, can never be done. I believe it to be possible, at the present state of our knowl- edge, to arrive at a solution of the fox sparrow problem that will enable anyone to identify accurately perhaps ninety per cent of the specimens taken, but it is to be doubted if diagnoses can be prepared enabling anyone positively to identify all specimens of Passerella secured. Even with the exceptional facilities I have enjoyed, I would not guarantee the correct allocation of all the birds examined in the present study. Intergrades between two forms may resemble a third and, taken in their winter home, may have their characters wrongly interpreted. Also immatures of one form may bear some resemblance to adults of another, as is sometimes seen in the thick-billed subspecies of the Schistacea group. Many specimens may be identified as of certain subspecies on the basis of large comparative series, whereas from one skin or a few it would be difficult to understand the allocation. Some of the difficulties encountered in identifying specimens may be understood from the following facts. Passerella is variable to an extreme degree, and intergradation of characters apparently occurs wherever two races come together. Thus, in the chain of subspecies extending along the coast of southern Alaska and British Columbia there are intergrades to be found between any two contiguous forms. There is also, judging from winter birds, intergradation between each of these coastal races and some inland form — iliaoa or altivagans, most likely. The nature and extent of this latter type of intergradation, as regards breeding birds, is still almost unknown, though winter specimens apparently indicative of its existence are not uncommon. It will be seen from this that in the Unalaschcensis group of sub- species as a whole there is an extremely wide range of variation possible. Of breeding birds of these six subspecies there are available fairly extensive series of each race from several localities, but the map (fig. N) will show what vast stretches of country there are as yet not rep- resented in collections. Series of summer birds from any one place are signally uniform in appearance, so it is evident that an explana- tion of the great range of variation seen in series of migrants and winter birds is to be sought in the study of specimens from many summer stations — the variation in a series of stations rather than among many specimens from one place. 78 University of California Publications in Zoology [ VoL - 21 As regards the difficulty of identifying the Unalaschcensis sub- species in the winter home in California, there are the following features of the problem to be considered. These races are arranged linearly along the coast in their summer habitats, and the centrally located subspecies intergrade, each one, with the races to the north and to the south. Thus, for example, there are individuals of sinuosa that approach unalaschcensis, and others that are close to mnectens. It might happen that a collector in California, in a limited series of birds, would have the two extremes represented. Comparison with sufficient material would show that his two diverse specimens must both be referred to the same form, in this case sinuosa, yet to his eye they would evidently be quite as unlike as other specimens labeled as of distinct subspecies. This is something that must be borne in mind in the consideration of small series of specimens. In the Schistacea group (the thick-billed and slate-colored spar- rows, so-called) there is another confusing factor. Among these subspecies the most variable feature is the bill (see fig. D). While the different subspecies are distinguished by smaller or larger bills, there is also (most noticeably among the larger-billed forms) some difference in this regard between young and old. For example, a young stephensi might have a bill of about the size of an adult nvariposae, and a young mariposae one of about the size of an adult monoensis (see fig. DD ) . It is not possible to tell, of course, whether the bill of an early fall migrant, taken far from its summer habitat, has or has not reached the maximum of growth, and in birds that are otherwise alike, it is easy to perceive the source of confusion that this variable feature may be. Aside from this complication there is in this group again undoubted intergradation between subspecies at the margins of their habitats. A thorough understanding of conditions in this section of the genus requires material from many more localities than are here represented. Osgood (1909, p. 17) in speaking of the mammalian genus Pero- myscus, makes the following comments, which, as is seen from the statements in the preceding paragraphs, apply as well to PassereUa. "Classification becomes, then, as has been said [by Ridgway, 1901, p. x], like dividing the spectrum and depends largely upon the standards set, for, theoretically at least, the possibilities of subdivision are unlimited. It is not strange, therefore, that hundreds and even thousands of specimens are intergrades almost equally resembling 1 920] Swarth: Revision of Avicm Genus Passer ella 79 two or more adjacent forms. Many of these intergrades for con- venience may be referred with some degree of assurance to the form they most closely resemble, but many specimens fall so near the imaginary line between two or more subspecies that it is practically impossible to classify them other than as intergrades." It must not be assumed from the above statements that I am dubious of the validity of the subspecies so far described. Far from it. Each of them is readily apparent in typical form, and certainly deserving of nomenclatural recognition. In the course of this study it has been considered necessary to supply names for races not here- tofore recognized but quite as distinct as some of older standing. Furthermore, it would have been possible to have bestowed some additional names, as will be seen in discussions of specific cases farther on, but it did not- seem to me that I should be justified in doing so at the present time. It is thus evident that this study of Passerella is not presented as the final word on the subject.. In fact, it has become clearly apparent that, abundant as the available material has been, compared with what previous students of the group have had, some future worker should certainly take up the subject with the advantage of vastly more extensive series. I am quite convinced, however, that whatever changes such future reviser may advance, he will not reduce the number of subspecies now recognized. I am inclined to believe that this future student will adopt an attitude similar to that assumed by Sumner (1918, p. 184) in one of his studies on variation and heredity in Peromyscus numiculatus: "I am not in the least con- cerned with characterizing and defining those taxonomic groups which have been called Peromyscus maniculatus gamb&li, rubidus and sonoriensis. I shall merely discuss the differences between (and within) four representative collections taken by me in widely sep- arated and climatically different regions of the state. The question as to what 'subspecies' a given mouse 'belongs to' is for my purposes a distinctly minor consideration. ' ' "With a thorough appreciation of this method of treating the finer divisions of a species, in the lesser degree of separation used in the present study of Passerella iUaoa it seemed advisable to attach a name to each skin. The necessity of cataloging museum specimens impels such action even were there no other reason to do so. It must be evident, however, that in the present case (the treatment of Passerella iliaoa) the usage of this system of nomenclature is purely an artificial contrivance, just such an aid as the drawing of checkerboard squares EBStnfary of OrnitM«gy US* Sapsucker W<*i i :■ : .; %^ v^ J £ ii' i fS IO0 IZB ISO 174 zoo tea 1 fo San Francisco through Mount Lyell C Monterey through Mount Whitney San Pedro through _San Jacinto Peak' Fig. G. Four cross-sectional profiles of California. The shaded areas indicate the altitudinal distribution in summer of the subspecies of fox sparrow occurring within the state at that season. a. Brevioauda on the Tolla Bolly Mountains; mariposae on Lassen Peak; fulva on the Warner Mountains. o. Mariposae on the Sierra Nevada west of Mount Lyell; monoensis in the Mono Lake region on the east sltfpe of the Sierra Nevada; caneseens on the White Mountains. o. Stephensi on Mount Pifios and on the Sierra Nevada west of Mount Whitney; mariposae on the Sierra Nevada east of Mount Whitney. d. Stephensi on San Jacinto Peak. through the interpolation of unsuitable territory, it never takes place through the interpolation of another race of fox sparrow. Regarding each isolated colony as a separate race (as in a sense they are, of course, though we can not always distinguish between them), and it is evident that any one line of variation extends continuously in a 1920] Swarth: Revision of Avian Genus Passerella 105 given direction, but that appreciable changes in appearance do not occur wherever there is discontinuity of range. Passerella is closely related to Melospiza. Fox sparrows are very like song sparrows, in structure, appearance, habits, and actions. The respective ranges also of Passerella iliaca and Melospiza melodia are suggestive of rather close relationship between the two, for their habitats are in a measure complementary since the two species, though frequently occurring in the same general regions, occupy different life zones. Passerella is mostly in the Boreal zone, descending occasion- ally into high Transition, Melospiza mostly in the Sonoran zones, though some races ascend to Transition and higher. On the coast of Alaska, at the northern extreme of the range of Melospiza, the general separation of the two does not hold, neither Sonoran nor Transition extending this far north, but at the latitude of Vancouver Island we find the birds thus separated, Passerella on the Boreal mountain tops, and Melospiza in the Transition zone valleys. All of the races of Passerella iUaca are migratory. Some are much more so than others, and in a general way it may be said that the subspecies breeding farthest north migrate farthest south. Passerella iliaca iliaca winters far south of its Boreal nesting grounds, concentrating into a relatively limited area in the South Atlantic and Gulf states. Thus the lines of migration in this sub- species converge, birds from Labrador and Newfoundland traveling southwest and those from Alaska and Mackenzie southeast, toward their common winter home. Birds of the Unalaschcensis group move directly southward along the Pacific coast, each into a more or less definitely circumscribed winter habitat. The six subspecies of this group are arranged in summer, from north to south (from the Alaska Peninsula to Puget Sound) in the following order: unalaschcensis, insularis, sinuosa, annectens, townsendi, fuliginosa. The. subspecies breeding at the northern extreme, unalaschcensis, insularis and sinuosa, move the farthest south in winter, passing completely over both summer and winter habitats of annectens, townsenM and fuUgimosa, and reaching the extreme southern limits of California. Annectens winters mainly on the coast of California in the Santa Cruz and San Francisco Bay regions, passing over the habitats of townsendi and fuliginosa to reach this destination. Townsendi in turn leap-frogs over fuliginosa and winters in numbers as far south as Humboldt Bay, California. Fuliginosa hardly migrates at all. Occasional stragglers reach San 106 University of California Publications im, Zoology [Vol.21 Francisco Bay and a few more winter in the Humboldt Bay region, but the main winter home of this subspecies certainly lies north of the latter point. Undoubtedly, however, it performs an altitudinal migration, moving down from the mountain tops which constitute its summer home, as on Vancouver Island. (See fig. H.) Fig. H. Map showing in somewhat conventionalized form the nature of the migrations performed by the subspecies of the Unalaseheensis group. 1, unalasch- eensis; 2, insularis; 3, sinuosa; 4, armectens; 5, townsendi; 6, fuliginosa. A solid line surrounds the summer habitat; a dotted line surrounds the main winter habitat. The broken lines connecting summer and winter habitats do- not necessarily indi- cate migration routes; they are inserted to emphasize the overleaping feature in the seasonal movements. The subspecies occupying the northernmost summer habitats (unalasohensis, insularis, and sirmosa) travel the farthest south. The subspecies occupying the southernmost summer habitat (Jvliginosa) is practically resident there. As bearing upon the manner in which these birds migrate, certain facts gathered during several seasons of field work carried on by the California Museum of Vertebrate Zoology on the coast of Alaska and !920] Swarth: Revision of Avian Genus Passerella 107 British Columbia are worthy of ; record, indicating a striking absence of migrants representing subspecies that might be supposed to occur commonly. An expedition to the Prince William Sound region, Alaska, coveringfthe migration period in part at least, took no speci- mens of Passerella other than the breeding form, sinuosa. Of two expeditions to the Sitkan district, Alaska, each one covering nearly the entire migration period, both spring and fall, one collected nothing but the breeding fox sparrow, 1 townsendi, the other secured, besides an abundance of tmwisendA, a single migrating individual of another race, insularis. An expedition to Vancouver Island, covering both spring and fall migration, secured, besides the breeding fuliginosa, one migrating individual of insularis and three of townsendi. I have at hand, however, two specimens ofimalaschcensis taken at Ketchikan, on the coast of southeastern Alaska, April 12 and 15, 1916, migrating birds; of course, and proof that some individuals, at least, do pass through this region. On the whole, however, it seems probable that these northern Alaskan fox sparrows do not as a rule travel south- ward by slow stages, passing continuously through the regions directly south of their breeding homes, >but rather that departure is made abruptly and by long flights to distant destinations. It may very well be ■ that unalaschcensis, insularis, and sinuosa, with their summer habitats lying so far to the westward, fly southeast over the open ocean for a long distance, the main body of the travelers not reaching the North American mainland until at about the northern boundary of the United States. In the above connection it may be noted that the one migrating example of insularis just referred to as having been collected in the Sitkan district, Alaska, was taken on Kuiu, one of the westernmost islands of the region, and a station where there was also found migrating the Aleutian Savannah sparrow (Passerculus sandwichensis sandwichensis), whose summer home lies in the same general region as this fox sparrow. In the Schistacea group the winter migration involves a moving down from the high mountains, but there is a decided southward passage also. The lines of travel and the respective ' winter habitats of the several subspecies are not so well understood as is the case in the other groups ; although the material at hand for the present study serves to explain some of the questions involved, there is need of further collecting of specimens and observation of migration at many points. 108 University of California Publications in Zoology [Vol. 21 The winter habitats of certain subspecies are still practically unknown, the few winter specimens that have been collected being undoubtedly stragglers from the centers of abundance. Details of distribution are given elsewhere in this paper, under each of the subspecies, but the nature of the several problems may be briefly touched upon here. The winter homes of fulva and canescens are still practically unknown. Megarhynchus appears in numbers in winter in southern California from an- unknown breeding ground. Stephensi in departing from its summer habitat in the Sierra Nevada and the higher mountains of southern California does not visit the adjacent valleys, either as a transient or as a winter visitant. It leaves the region absolutely, in a similar manner to the Green-tailed Towhee {Oreospiza chlorura) and the Wright Flycatcher (Empidonax wrighti) of the same habitat. The latter two species, however, are known as migrants to the southeast, while the route and destination of stephensi, are facts that are still hidden. The manner in which megarhynchus, mariposae, and brevicauda are distributed locally in southern California in winter suggests that the winter homes of the other subspecies may be found by further explor- ing, at the proper season those regions whose winter avifaunas are still but slightly known. It may be pointed out, as regards the mountain ranges of southern California, that while we are well acquainted with the summer birds, there, are but one or two points where winter collecting or observations have been carried on at all extensively. Judging from the local distribution in winter of certain of the subspecies of Passerella, it may well be that those races whose winter homes are not known are just as definitely delimited, but in sections from which reports covering the proper season are lacking. The accompanying charts (tables 1, 2) are designed to show the winter distribution of the subspecies of Passerella in California. Under each of the several faunal areas are entered the number of specimens examined of each subspecies, taken within the specified areas between October 15 and April 15, and also the percentage of the total Passerella winter population that is formed by each sub- species. The restriction between limited dates is not quite fair, as it throws out a number of specimens of the more southerly wintering subspecies, taken unquestionably in their winter habitats, but some such arbitrary rule seemed necessary to allow for the presence of migrating individuals in the more northern parts of the state. It is also possible that October 15 is a little too early a date to insure the presence of a bird in its winter home in all cases. 1920] Swartk: Revision of Avkm Oewus Passerella 109 Table 1. — Manner of occurence in winter of the subspecies of Passerella iliaca in the different faunal areas of California. Specimens listed were taken between October 15 and March IS. The figures in the body of the table indicate the number of each subspecies in each faunal area. Column at extreme right indicates total number of specimens of each subspecies. Bottom row of figures indicates number of specimens of all subspecies from each faunal area. g 1 a s 'a el ■a OS &s Si So i 53 ! o fa g s bo "3 o *+> OS as 9 9 ■o'S 8.2 H 2 o OS t a to o o in '3 a es M a so 03 1 a ■*» a I g si to •0 a '3 0) 3 ►-9 a ES to o ■ BO 5 a es OS ■e el « es *> a el CO t S3 a o ■a es o *3 O i i i 3 3 2 3 12 1 2 8 4. 13 1 4 3 1 8 1 2 1 10 6 28 4 1 11 2 1 3 28 54 1 1 1 3. 2 10 15 27 4 9 1 4 i 3 4 1 1 3 1 7 32 36 25 124 16 1 14 4 82 24 8 11 8 i 2 1 3 2 3 1 1 12 46 56 60 213 127 40 6 49 16 15 4 92 38 9 20 9 4 6 21 28 17 4fi 63 90 71 5 9 5 396 12 1 770 Table 2. — Manner of occurrence in winter of the subspecies of Passerella iliaca in the different faunal areas of California ; based upon the same specimens as those listed in Table 1 . The figures indicate the percentage of the Passerella population of each faunal area that is formed by the indicated subspecies. 00 c *3 t § a 1 S el •O cS 55,, 55:9 £ o £ bo el ■a el Sa. 55 $ Si to I o fa t bo - '3' o +j o es a O +j OT3 11 9 9 WW ■S"9 3.2 U a) ^ * N 2 o es a es m 3? m o u .2 o a & a a to 5? 1 o s 2 o cS GO a? ■a u el V 5 a '3 ST a a o 93 3 a el CQ GO ■o 2 es 1 M es a es IS t* <9 S3 a o 1 o "3 O p 16.6 16.6 16.6 50 14.2 9.7 14.2 57 4.8 7.2 28.5 14.3 46.4 3.6 23.5 17.7 5.9 47 5.9 86.9 13.1 3.2 1.6 15.8 9.5 44.4 6.4 1.6 17.5 2.2 1.1 3.3 31.1 60 1.1 1.1 1.6 4.3 2.9 13.3 21.2 38.1 5.7 12.8 20 80 11.1 33.3 44.4 11.1 20 60 20 1.8 8.1 9.1 6.3 31.3 4.3 .3 3.5 1.1 20.9 6.1 2.1 2.8 2.1 1.1 16.6 8.3 25 16.6 25 8.3 p p. p . unalaschcensis *.. p p p p p p p p p 100 p p. p 110 University of California Publications in Zoology [Vol.21 From certain of the faunal areas .the numbers of birds examined were too small for the results to have much significance, but there are certain facts in winter distribution that the chart shows very clearly. Attention may be called to the demonstrated preponderance of annectens in the Marin, San Francisco Bay, and Santa Cruz sections, compared with its absence from the Humboldt Bay section directly to the northward, and the limited numbers found elsewhere in the state; the preponderance of townsendi in the Humboldt Bay section, Fig. I. Map of southern California indicating the position of the localities designated in the text as (1) the Santa Monica Mountains region, (2) the Pasa- dena region, (3) the Claremont region. and the few individuals taken at other points ; and similar restrictions in range shown by certain other subspecies. From the Sierra Nevada, Humboldt, Marin, Santa Cruz, San Francisco Bay, and San Diego regions, there are sufficient winter specimens to warrant some generalizations. From the San Diego region (with which is included the San Bernardino Mountain region) there are enough birds available to permit of analysis of conditions at several localities within this area, such as can not be done elsewhere in the state. Of the 396 winter collected specimens of Passeretta from the San Diego region, 326 were taken from three relatively restricted areas, in Los Angeles and San Bernardino counties. These points I have designated as follows (see fig. I) : (1) Santa Monica Mountains region (Coldwater Canon, Franklin Canon, Hollywood, etc.) ; (2) Pasadena region (Arroyo Seco, Millard Canon, Mount Wilson, etc.) ; (3) Claremont region (San Antonio Canon, San Dimas Caiion, Lytle 1920] Swarth: Revision of Avian Genus Passerella 111 Creek, etc.). Bach of these three areas includes places that were favorite collecting grounds of certain persons (Santa Monica Moun- tains: J. E. Law, L. B. 'Wyman, W. B. Judson, H. S. Swarth; Pasa- dena: J. Grinnell, P. S. Daggett, H. S. Swarth; Claremont: W. M. Pierce), and the series of Passerella are results of years of collecting at these points. I believe, therefore, that these series may be taken as fairly indicative of average conditions within the areas where they were obtained. As to the characteristics of the three areas, the Santa Monica Mountain region is widely different from the other two, while the Pasadena and Claremont regions are essentially alike. The part of the Santa Monica Mountains that is here included is a spur extending eastward from Santa Monica Canon about fifteen miles. These hills are low, nowhere reaching an elevation of 2000 feet, densely brush covered, with very little large timber and very little water. The mountains of the Pasadena arid Claremont regions are of a different character, much higher and more rugged, with deep canons containing large streams of water, and with an abundance of timber. Table 3. — Winter distribution of the subspecies of Passerella iliaca at three neighboring localities in the San Diego region (see map, fig. I). Specimens listed were collected between October 15 and March 15. Los Angeles and Santa Monica Mountains Pasadena and adjacent mountains Claremont and adjacent mountains 1 4 2 5 8 26 18 •2 2 2 6 7 17 9 10 2 19 1 4 3 2 17 26 1 83 2 7 26 P. i. canescens 6 7 1 Totals 68 80 178 There are certain features in the distribution of Passerella in the three sections that are at once apparent (see table 3) : the preponder- ance of schistacea in the Claremont region, the restriction of orevi- cawda to the Santa Monica Mountain region, the scarcity of any subspecies of the Unalaschensis group in the Santa Monica Mountain region, the preponderance of insukuris in the Pasadena region, and the preponderance of sinuosa in the Claremont region. 112 University of California Publications in Zoology [Vol. 21 The figures given for the southern half of the Sierra Nevada and for the Sierra foothill sections are based in part on material secured by expeditions from the California Museum of Vertebrate Zoology into the Yosemite region, covering every month of the year. The results apply only to this limited area, but judging from these figures the subspecies of the Schistacea group form but a small part of the winter population of the Sierra Nevada. I am convinced that there is significance in the results of these tabulations. I believe them to indicate that some forms really have a winter habitat nearly as sharply defined as the summer home, while others show preferences somewhat less marked. One lesson to be learned from this analysis of winter collections is that inasmuch as the extensive collecting of specimens at certain, limited spots has yielded definite knowledge of a sort that obviously could not have been obtained in any other way, similarly extensive collecting at places regarding which we are still very ignorant as to the winter avifauna, may be expected to explain many details of distribution in this group of birds of which at present we have no knowledge. In the writer's opinion, the solution of such complicated questions of relationship and distribution as are presented in the Passerelkt, group lies only in the wholesale but intelligently directed collecting of specimens. In the course of the writer's field experience he has from time to time noted facts in the distribution and migration of fox sparrows, which, disconnected though they be, are set down here as having some bearing on the subject in general. In a number of years of observa- tion about Los Angeles, fox sparrows were found to be extremely rare in the flat valley in which the city is situated, one or two of the Unalaschcensis group, and two altvwagans being all that were noted, and these at long intervals. In the low foothills of the Santa Monica Mountains, but a few miles away, certain subspecies could be found in fair numbers regularly every winter. This was also the case in the brush of the Arroyo Seco, a stream issuing from the mountains near Pasadena, fox sparrows being seen regularly in winter in the bordering willow thickets, practically as far as the junction of this stream with the Los Angeles River, at the edge of the city. In years past many southern California valleys were covered with miles of more or less dense chaparral. Much of this land has been cleared for agricultural purposes, but some fairly extensive areas of this nature still remain, the haunts of many typical Californian animal 1920] Swarth: Revision of Avian Genus- Passer ella 113 species. These tracts afford shelter for brush-loving birds, and it might be supposed that they would be occupied by the winter visiting fox sparrows; but such is not the case. In many years experience in such surroundings, while quail shooting and bird collecting, the writer has never seen a fox sparrow thus situated. In southern Cali- fornia fox sparrows in general apparently occur but rarely in the Lower Sonoran valleys. The brush covered hills, Upper Sonoran and as far up as low Transition, is the favored winter habitat, with local variations as elsewhere noted. In April, 1912, while the author was collecting at the eastern base of the Sierra Nevada, near Lone Pine, Inyo County, fox sparrows were observed migrating northward. All that were taken were schistacm, the forms that nest in the adjacent mountains, mariposae in the Sierra Nevada, and canescens in the White Mountains, not being detected in the lowlands at all. This fact, and also the non-occurrence of stephensi in the foothills of the mountains wherein the latter breeds, is, I believe,. in each case, indicative of rather abrupt arrival at, and departure from, the breeding grounds, with no tarrying in the low- lands nearby; suggestive of just such migration, in fact, as has been inferred on similar grounds in certain of the Alaskan subspecies. In September, 1916, in the Kings River Canon, in the Sierra Nevada, the first migrating fox sparrow was noted on September 10, and several others were observed during the next few days. Those collected proved to be all schistacea, evidently from some point con- siderably to the northward, and no specimens were taken of mariposae, the race breeding in the Sierra Nevada to the north and east of Kings River Canon. No Passerellat was found breeding on the floor of the canon. A few days later, at Horse Corral Meadow, on the adjacent table land to the southward, fox sparrows of several subspecies were found migrating abundantly. It seemed apparent that there was no migration from the east to the west side of the Sierras, or vice versa, in which the Kings River Canon could be used as a highway, and this was borne out by the manner of occurrence of many species of birds besides the fox sparrows. It was also apparent that migrants were travelling southward in numbers along the line of the Sierras, stopping at favorable places along the higher ridges, but not descending, except rarely, into the deep gorge that cut across their path. 114 University of California Publications m Zoology [Vol. 21 Passerella iliaca iliaca (Merrem) Eastern Fox Sparrow Original description. — "Fringilla iliaca Merrem, Beytr. besond. Gesch. Vogel, 11, 1786, pi. x. (North America.) " (A. O. U. Check-List, 1910, p. 277.) Type specimen. — Not known to exist. Range. — "North. America. Breeds in Boreal zones from tree limit in northeastern Alaska, northern Mackenzie, central Keewatin, north- ern Ontario (Moose Factory), and northern Ungava south to central Alberta, northern Manitoba, southern Keewatin, Magdalen Islands, and Newfoundland; winters from lower Ohio and Potomac valleys (occasionally further north) to central Texas and northern Florida; casual on the coast of southern Alaska and in California. ' ' (A. 0. U. Check-List, 1910, p. 277.) One record from Arizona (see p. 118). Specimens examined.— 48 (see list, p. 184). Distinguishing characters.— 3Prom all subspecies of the Schistacea group, iliaca is distinguished by different proportion of wing and tail lengths (wing longer than tail in timca, wing equal to, or shorter than, tail in the Schistacea group), and by the predominance of bright, ruddy coloration. From all subspecies of the Unalasehcensis group, iliaca is distinguished by the much brighter hue of its red markings, by the sharply contrasting red and gray streaking dorsally, and by the heavier and more stubby bill. Within the Iliaca group iliaca is distinguished from the one other subspecies, altivagans, by greater general size, and brighter, more ruddy coloration, with dorsal streaking much more sharply defined. Remarks. — There are at hand fourteen specimens of Passerella taken in California that are evidently referable to Passerella iliaca iliaca; a small enough number to allow of individual consideration being given to each. No. 4932, collection of A. B. Howell; male; Pasadena; December 13, 1907; collected by A. Van Eossem. (See Willett, 1912, p. 85.) This bird appears to be absolutely typical of iliaca, being closely matched in all respects by selected specimens from Illinois. It is, however, rather small for a male, though it is marked as such, approaching more nearly the dimensions of some females. No. 867, collection of J. B. . Pemberton ; female; Big Sur Biver, Monterey County; December 22, 1903; collected by J. B. Pemberton. (See Pemberton, 1908, p. 50.) Apparently a typical example of Uiaoa. Very closely similar to the specimen last described, and likewise duplicated by selected eastern birds. No. 384, collection of A. B. Howell; female; Burbank, Los Angeles County; November 11, 1911; collected by A. B. Howell. (See Howell, 1912, p. 41.) Ven- trally indistinguishable from eastern birds, but dorsal coloration generally more reddish. The streaks on the back are broader and consequently less sharply defined, and there is none of the clear gray on head and neck that is apparent on all the eastern specimens used in comparison. Wing bars but faintly indicated. No. 789, collection of J. Eugene Law; sex not determined; Lankershim, Los Angeles County; March 22, 1908; collected by J. Eugene Law. (See Willett, 1912, p. 85.) Closely similar to eastern examples, the main difference lying in 1920]' Simrth : Revision of Avian Genus Passerella 115 the duller dorsal coloration of the California bird. The dorsal streakings are less Tuddy than in eastern specimens, and less sharply defined. Top of head is gray, washed with brownish, as in many eastern birds, and the sides of neck are as clear gray as is seen in other winter collected specimens of iliaca. Ventrally it is indistinguishable from typical iliaca. Wing bars faintly indicated. ° Passerella iliaca iliaca DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY UHIVZD9ITT OF OU.IFOMIA Fig. J. Map showing stations of occurrence in California of Passerella iliaca iliaca, as established by specimens examined by the author. No. 30445,. Mus. Vert. Zool.; male adult; Seaside, Monterey County; Decem- ber 26, 1918; collected by H. G. White. Differs from average eastern birds mainly in somewhat darker coloration. Dorsal streaking and wing bars not so strongly indicated as in eastern collected specimens. No. 3233, collection of J. Grinnell; male; Mount Wilson, Los Angeles County; October 31, 1897; collected by J. Grinnell. (See Grinnell, 1915; p. 130.) This bird, together with the rest of the specimens here listed, is suggestive of iliaca 116 University of California Publications in Zoology [Vol. 21 rather than any other subspecies, but differs from the mode of that race in certain details of markings and color. Dorsally this bird has hardly a trace of streaking, the back being almost uniformly reddish, with a suggestion of under- lying gray about the head and neck. The ventral streakings, though of a brighter, more "foxy" red than in the TJnalaschcensis group, are not so red as in typical iliaca. Wing bars are but faintly indicated. No. 4630, collection of J. and J. W. Mailliard; male; San Geronimo, Marin County; January 26, 1901. (See Mailliard, 1901, p. 72.) Close to typical iliaca, differing from eastern birds mainly in less contrasted streaking above, and in more blackish markings below. Bill as in typical Uiaoa, and wing bars conspicuous. No. 4208, collection of J. and J. W. Mailliard; female; Plumas County, "alt. 4000"; September 27, 1898. Differing from eastern birds only in being less streaked dorsally. The general appearance of the back is reddish brown, mottled rathern than longitudinally streaked. Wing bars conspicuous. No. 750, collection of A. B. Howell; female; Santa Barbara; January 1, 1911; collected by J. H. Bowles. (See Bowles, 1911, p. 175.) Ventrally indistinguishable from selected eastern specimens. Back and top of head almost uniformly dull grayish brown, back obscurely mottled with somewhat darker brown. Reddish coloration on rump and upper surface of tail feathers rather paler than in typical Uiaoa. Wing bars fairly well defined. No. 2003, collection of W. Otto Emerson; male; Hayward, Alameda County; December 12, 1898; collected by W. Otto Emerson. Differs from average eastern examples of Uiaoa in the darker, less reddish, coloration of the spots below, and in the almost uniformly brown back. The dorsal region is dull brown, faintly mottled with darker brown, and on the head and neck there is a dull brown wash over an underlying grayish color. Wing bars faintly indicated. No. 1811, collection of W. M. Pierce; male; Palmer's Canon, Los Angeles County; February 2, 1919; collected by W. M. Pierce. Differs from eastern birds in its generally darker coloration and heavier ventral markings. Upper parts dark grayish brown, washed with reddish; back obscurely streaked with dark reddish. Breast streakings of a brighter reddish color than the back. Wing bars faintly indicated. This bird is fairly intermediate between typical Uiaoa and unalasohcensis. No. 1810, collection of W. M. Pierce; male; Palmer's Canon, Los Angeles County; February 1, 1919; collected by W. M. Pierce. To all appearances this specimen lies between typical iliaca and altivagans. It is larger than average examples of the latter, darker colored and more heavily marked below. Upper parts nearly uniform grayish brown, more grayish on the head, more brownish on dorsum. No dorsal streakings and no wing bars. In color and markings this bird is close to altivagans, in dimensions it is nearer Uiaoa. No. 2805, collection of F. S. Daggett; male; Nicasio, Marin County; Febru- ary 6, 1906; collected by C. A. Allen. Spotting beneath of a darker brown than in any eastern specimen; top of head and back almost uniform dull brown; dorsal region faintly spotted. Wing bars faintly indicated. This bird is about inter- mediate in color and markings between typical iliaca and insularis. The bill is as in iliaca, and on the whole the specimen seems best referred to that subspecies. No. 1180, collection of L. E. Wyman; female; Santa Monica Mountains, Los Angeles County; January 31, 1917. Collected by L. E. Wyman. This bird can be considered an example of Uiaoa only by reason of the manner in which the series of variants just detailed connects this extreme with the mode of the subspecies. It is in a slight degree further from the type than the specimen last described. Under parts are rather heavily streaked, and with a dark shade of 192 °] Swarth : Revision of Avian Germs Pwserella 117 brown; upper parts almost uniform, a faint suggestion of gray, underlying the brown of the head and neck, the back slightly mottled, and the wing bars barely apparent. The character of the bill, however, is closely similar to that in iliaea. From the appearance of the specimens described above several deductions may be drawn. These birds are all to be referred to the subspecies iliaea, but, with two exceptions (no. 4932, coll. A. B. Howell, and no. 867, coll. J. R. Pemberton), they vary from typical iliaea sufficiently to be noticeable if inserted among a series of eastern collected skins. The specimens all resemble iliaea most nearly in the ventral markings ; the greatest variation is evident on the upper parts. These birds all show a tendency toward elimination of dorsal streak- Fig. K. Wings of the two subspecies of fox sparrow of the Iliaea group, natural size. a. Passerella iliaea iliaea, adult male; no. 6088, Mus. Vert. Zool. ; Eastford, Connecticut; April 12, 1876. b. Passerella Uiaca altivagans, adult male; no. 26039, Mus. Vert. Zool.;' Aspen Valley, Yosemite National Park, California; October 14, 1915. ings, with consequent greater uniformity of color on the upper parts. This color in some is reddish, of a darker shade than is seen in typical iUaca, while in others it is decidedly gray. Such variation as appears in the breast streakings tends toward the darker shades of red. In the subspecies UAaca there is, according to Ridgway (1901, p. 387), great variation in color, the extremes being designated by him as the "gray phase" and "red phase," respectively, and these differences are sufficiently correlated with different regions to suggest the possibility of the eventual recognition of two subspecies. "The extreme rufous phase seems to be represented only in the Atlantic Coast district, and is altogether wanting in Alaska and other north- western portions of the continent, where only birds representing the gray extreme occur" (Ridgway, loo. tit.). However, it is extremely unlikely that the reddish-colored birds occurring in California in win- ter are migrants from the Atlantic Coast, where the red phase of iliaea has its habitat. The probabilities are that the Californian migrants 118 University of California Publications in Zoology [Vol. 21 in question arrive from the northwest together with the other fox sparrows with which they are found associated. My personal convic- tion is that they hail from an intermediate locality, or localities, between the ranges of typical iliaoa of the interior of Alaska and those of some of the Unalaschcensis subspecies of the coast. Some of the specimens described in this paper, together with certain variants of insularis and unalaschcensis at hand, form, both in color and propor- tions, a graded series connecting the extremes by such short steps that any dividing line between the two must be arbitrarily drawn. Then, too, the reddish coloration of these birds is of an appreciably darker shade than in the few Atlantic coast winter birds available for comparison, tending more toward the color of the Unalaschcensis subspecies. The more grayish-backed birds may represent a step toward altivagans. I can not agree with the opinion that the character of any of these birds points toward interbreeding of iliaca with schistacea (see Grinnell, 1915, p. 130). Beside the California birds described, there are two other specimens from unusual localities. One is a female (no. 2524, coll. J. Eugene Law) taken by A. Van Rossem in Pinery Canon, at 6000 feet, Chiri- cahua Mountains, Arizona, November 6, 1914. This specimen, here placed on record, comprises the first recorded occurrence of the sub- species in Arizona. It is extremely close to typical iliaca, differing mainly in its more uniform dorsal coloration, the back being quite grayish, and the dorsal streakings but poorly defined. This may be indicative of intermediateness between iliaca and altivagans. The second specimen is a male (no. 967, Provincial Museum, Victoria, B. C), collected by John Fannin at Sicamous, British Columbia, September 25, 1893. This bird is absolutely indistinguishable from selected eastern examples of iliaca. The California specimens of iliaca are too few in number to indicate any definite manner of occurrence in any specified area, but it seems unquestionable that some individuals of this subspecies winter each year within the borders of the state. Most of the speci- mens taken are from the San Diego region (where, however, there has been more winter collecting of fox sparrows than anywhere else in California), and as these were all secured between October 31 and March 22, they may be assumed to have been in their winter home. The specimens taken in December in Monterey and Alameda counties, and in January and February in Marin County, indicate the prob- ability of the winter habitat including the whole coastal area north 1920] Swarth: Revision of Avian Genus Passerella 119 to the San Francisco Bay region. The one specimen from the Sierra Nevada of Plumas County, taken September 27, was probably a migrant on its way farther south. It is noteworthy that nearly all the California examples of iliaca were secured by collectors who had gathered specimens for years at the several points where these birds were taken, hence the small number of specimens of this race, compared with those of most of the other wintering subspecies of Passerella, may be taken as fairly indicative of the relative numbers of each actually present. Fig. L. Bills of the two subspecies of fox sparrow of tlie Iliaca group ; natural size. a. Passerella iliaca iliaca, adult male; no. 6088, Mus. Vert. Zool. ; Eastford, Connecticut; April 12, 1876. b. Passerella iliaca altivagans, adult male; no. 26039, Mus. Vert. Zool.; Aspen Valley", Yosemite National Park, California; October 14, 1915. Besides the specimens here listed there are others recorded from California that I have not seen, as follows: "Accidental in California (spec, in Mus. S. I.) " (Coues, 1874, p. 161) ; specimen from Poway, San Diego County, January 3, 1888 (W. E. Bryant, 1889, p. 90) ; specimen from Oakland, December 2, 1892 ("W. E. Bryant, 1893, p. 363) ; Saticoy, December 14, 1872 (Baird, Brewer and Ridgway, 1874, p. 516). Passerella iliaca altivagans Riley Alberta Fox Sparrow Original description. — Passerella iliaca altivagans Riley, 1911, p. 234. Type specimen. — No. 222832, U. S. Nat. Mus. ; male, immature ; Moose Branch of Smoky River (about 7000 feet altitude), Alberta; July 31, 1911; collected by J. H. Riley (original number, 2175). Range. — In summer known only from a few points in the interior of British Columbia and the adjacent part of Alberta (see map, fig. N) . In winter southward at least to extreme southern California. 120 University of California Publications in Zoology IT 0L - 21 Winters mainly in the San Diego region and on the western slope of the Sierra Nevada (see map, fig. M). Specimens examined. — 90 (see list, pp. 185-186). Distinguishing characters. — Of the Iliaca group (see p. 89). Somewhat smaller than iliaca; size and proportions about as in the subspecies of the Unalaschcensis group (see table 4). Coloration more subdued than in iliaca; rather brighter, with more strongly contrast- ing reds and grays, than in any of the Unalaschcensis group. Remarks. — While the breeding range of the Alberta fox sparrow cannot yet be exactly defined, it may be said in general terms to be in the interior of British Columbia and in the adjoining portion of western Alberta. The breeding stations afforded by Riley's type series are the Moose Pass region on the boundary line between British Columbia and Alberta, and Thudade Lake, British Columbia. To these I am now able to add Jasper Park, Alberta, and McGillvary Creek, and Mt. McLean, Lillooet District, British Columbia. The migration route, as indicated by specimens at hand, would be through the western part of the Great Basin, in a path whose width extends from western Montana to central Oregon, and southwestwardly at least as far as extreme southern California. The winter home appar- ently lies almost entirely within the confines of California. There is one specimen at hand (not typical, however) from Government Island, near Portland, Oregon, taken December 15. In California the line of migration seems to lie entirely east 'of the coast ranges. There are no records from these mountains, nor even from the floor of the Sacramento-San Joaquin Valley, immediately to the eastward, while there are specimens, migrants, from many points the whole length of the Sierra Nevada. In all probability, judging from the available data, the winter metropolis of altivagans is in the western foothills of the southern Sierra Nevada, perhaps as far north as Eldorado County, and in southern California west from the eastern bases of the several moun- tain ranges separating the desert and the Pacific slope, with sporadic occurrences in the San Francisco Bay region. It may be that the subspecies winters regularly in the coast ranges from San Francisco Bay southward. There is but one specimen at hand from this region, but there has been practically, no winter collecting carried on there. As to dates of arrival and departure in California, a bird taken September 22, at Horse Corral Meadow, in the Sierra Nevada of Fresno County, is the earliest fall migrant, and a specimen from the Sierras of Plumas County, ' ' alt. 4000 feet, ' ' April 24, is the latest in. 1920] Swarth: Revision of Avian Genus Passerella 121 the spring. The Horse Corral Meadow bird was taken on an expe- dition headed by the present writer, and, as observed by him, was one of a "wave" of migrating northern fox sparrows which had just o Passerella iliaca altivagans DISTRIBUTION MAP MUSEUM OP VERTEBRATE ZOOLOGY intjvuunr or cm-iforhia Fig. M. Map showing stations of occurrence of Passerella iliaca altivagans in California, as established by specimens examined by the author. The region enclosed within the dotted outline is believed to be the main winter habitat of the subspecies. arrived in a region where we had been collecting birds for some time. It may therefore be regarded as quite accurately representing the arrival of the subspecies there in that year. Passerella iliaca altwagtms was described by Riley- (1911, p. 234) upon the basis of four young birds, three from the Moose Pass region, 122 University of California Publications in Zoology [Vol. 21 on the British Columbia-Alberta boundary, and one from Thudade Lake, British Columbia. The subspecies is described (loc. cit.) as "Similar to Passerella iliaca schistacea, but middle of the back mars brown instead of mouse gray; wings and tail with more red in the brown (near burnt umber)." Of some additional specimens, migrants and winter birds, Eiley remarks that "from typical Passerella iliaca schistacea, to which they have been referred, they are quite distinct, having the mouse gray of the upper parts of that race replaced by broccoli brown and burnt umber and the spotting of the under parts prout's brown, not sepia. Apparently there is no difference in size." The A. 0. TJ. Committee (1912, p. 386) denied recognition to this subspecies on the ground that it was "too near P. i. schistacea." The form was recognized by Grinnell (1915, p. 130), and some years later Oberholser (1918, p. 186) reasserted the validity of altivagans, affirm- ing its near relationship to iliaca, and its unlikeness to any other race. Most of Riley's type series has been available for the present study. The three birds collected by him (in July) and described (loc. cit.) as "nearly adult" or "slightly immature, " are, speaking more exactly, molting from juvenal into first winter plumage. The type specimen has reached the most advanced stage in the change, but all three still show large tracts of juvenal feathers. However, there is enough of the new plumage in place to afford a fair basis for comparison with full feathered migrants from points farther south, and thus to estab- lish with fair certainty the color characters of altivagans in its typical form. A specimen in fresh fall plumage (no. 26039, Mus. Vert. Zool.), selected for its close resemblance in comparable parts with the molting type of altivagans, is colored as follows: Upper parts generally mummy brown, most ruddy on dorsum, dullest on rump. Feathers of dorsum with dark centers, giving an obscurely streaked effect. Sides of neck washed with grayish. Upper tail coverts auburn, with narrow edgings of paler brown; exposed portion of rectrices auburn. Under parts white, chin, throat and abdomen nearly immaculate, else- where marked with triangular spots, varying in color from some the same shade as the back, to others suffused with blackish. Exposed portion of wing coverts and remiges auburn, like tail. Median and greater wing coverts slightly tipped with whitish, forming two ill- defined bars. The above described specimen is an average fall-plumaged example of this subspecies, though there are others at hand showing diverse . extremes of dorsal coloration, of grayish or of reddish. Specimens 1920] Sumrth: Revision of Avian Genus Passerella 123 taken in the spring, and consequently in less fresh plumage, are duller, less ruddy, and more grayish than fall birds. From what can be seen of the juvenal plumage of the Moose Pass birds, in this stage the markings are more ruddy than in juvenals of schistacea, which they otherwise most nearly resemble. There are no young of iliaca available for comparison. The large series that I have been able to bring together, though yielding few summer birds additional to those of Mr. Riley, abundantly demonstrates the distinctness of this subspecies, and there is no difficulty in recognizing in the winter birds from California the same race as that from British Columbia described by Riley. The small series from McGillvary Creek, British Columbia, pre- sents several points worthy of emphasis. It includes five specimens, one in juvenal plumage, two well advanced in the post- juvenal molt, and two in fully acquired winter plumage, whether immatures or adults is not apparent. Prom the dates at which they were taken (August 10 to 18), and from the fact that one is still in the juvenal plumage, these birds may be assumed to have been taken upon their breeding ground. McGillvary Creek is, in a direct line, less than ninety miles east of the general line of the coast, much less than that distance east of several deeply penetrating arms of the sea, and may be assumed to represent a point on the western boundary of the habitat of altivagans. It indicates the occurrence of the subspecies as far west as the eastern slope of the ranges most closely adjoining the coast. The birds from this point, though properly referable to altivagans, differ appreciably from the mode of that subspecies. The under parts are of a richer brown, and more heavily streaked, and there is an entire absence of streaking above, producing an evident appearance of intergradation toward the character of the Unalaschcensis sub- species. They can be very nearly matched with some winter birds from California. There is a series of July birds from Mt. McLean, so short a distance east of McGillvary Creek that the two lots must perforce be of the same race. Yet the worn and faded plumage of the Mt. McLean birds betrays hardly a trace of the brown coloration seen in the McGillvary Creek fall specimens, and the former are superficially very much like schistacea. However, the breast spottings still show a browner tinge than is seen in that race. The California series as a whole exhibits some little variation in color, though not so much but that the birds can, with all certainty, 124 University of California Publications in Zoology [Vol.21 be placed with altivagans, as now denned. From a style that may be taken as typical of the subspecies, the mode of coloration most frequently seen — pale reddish back, obsoletely streaked with darker red, and underparts moderately streaked — there are two extremes of variation. One is darker colored, dark reddish brown above, heavily marked below, the other predominantly grayish above and but sparsely spotted on the breast. These specimens I take to represent various degrees of intergradation, the reddish colored birds toward the coastal subspecies of the Unalaschcensis group (as shown in the McGillvary Creek series), the grayish colored ones toward iliaca. It will be noted in these non-typical birds as well as in typical altivagans, that tail is always shorter than wing, as is the case in all fox sparrows outside the Schistacea group. There are a few apparent intergrades of another type, of duller, less reddish coloration and with tail of about the same length as wing. These I take to represent intermediate stages between altivagans and schistacea. Of a series of nine specimens taken by W. M. Pierce in certain canons near Claremont, southern California, during the winter of 1918-19, practically all are variants from typical altivagans. appar- ently tending toward some of -the darker colored coastal races. One of these birds (no. 1776, col. W. M. Pierce) is closely 'similar to one of the summer specimens from McGillvary Creek, British Columbia. Compared with the subspecies of the Schistacea group, altivagans is apparently always distinguishable by the proportionate lengths of wing and tail, wing longer than tail in altivagans, rarely the same length, wing usually shorter than tail, in all members of the Schistacea group. In the occasional variant of altivagans in which coloration somewhat closely approaches some examples of schistacea, the wing and tail measurements suffice for identification. In typical altivagans the spots on the lower surface are more ruddy than in the Unalasch- censis group, and rather less thickly distributed, leaving more white showing through where the spots are thickest, and more extensive pure white areas on throat and belly. It is quite possible that the name altivagans as now used covers a composite of several recognizable subspecies, as was the case in the former "megarhynchus" and "schistacea,." The breeding birds available are few in numbers and from but two or three scattered points, while as regards the greater part of the interior of the vast northwest nothing is known of the Passerella population. In view of 192 °] Swarth: Revision of Avian Genus Passerella 125 the manner in which fox sparrows are distributed over much of the general range of the species elsewhere, locally segregated at scattered points, and considering the variation seen elsewhere in series of breed- ing birds taken from isolated, but relatively close colonies, it would not be surprising if the diverse specimens of alt wag cms from the winter home of that race, should prove to be representative of addi- tional recognizable subspecies, forming chains of variants comparable to those just mentioned. Riley's type series from the Moose Pass region was taken at high altitudes (see Eiley, 1912, p. 69), and the McGillvary Creek birds were collected at 5000 feet and 7000 feet altitude. These facts point to the probability of altivagans inhabiting in summer the higher mountains, and not occurring in the valleys, just as is the case with the related subspecies to the southward. Under such conditions there is added probability of the existence of distinguishable forms within the general range here ascribed to altwagans. Two skins from Penticton and one from Midway, British Columbia, show a leaning toward schistacea. While from the dates of capture (April 21 and May 5) they may have been migrants at the points where they were taken, it is probably fair to assume that they would have bred somewhere near-by, and that these places therefore lie just north of the boundary line between altivagans and schistacea. In typical form, therefore, altwagwis probably does not breed south of the British Columbia- Washington boundary, nor schistacea north of it. An interesting sidelight on the status of this subspecies is afforded by the comments written upon the labels of many specimens examined from contributing collections. There is hardly a bird collection of any size in California but contains one or more examples of this race, many of them collected years before altivagans was described. These specimens have almost invariably been a source of trouble to their owners, for various names are written and rewritten upon the labels, frequently with question marks appended. Some of these birds have been referred to different authorities for naming, and the labels reflect opinions of their relationships, which, though diversely worded, are mostly in agreement as to the general status of the race. "Schistacea — not typical," " schistacea X Uiaca," "iliaca — intergrade, " are the terms most frequently seen, terms which quite accurately indicate the systematic position of altwagans, and in their unanimity also express the need of, and justification for, the naming of the subspecies. 126 University of California Publications in Zoology [ VoL - 21 3> IA 'if f*/ H HP's \(l m •3 03 « E 03 03 69 03 . •3 b5 S g *s * EH 3 .= ■§ ! "& So § | IB' n-v. 05 .S 01 £ is » ^ 'S ^ ^ .^ b g gi*^ ., * « ^ bo ~s as zz o -m "« ~hS t S^. g°«i * -£ P ' I'g.gg.a 1920] Swarth: Revision of Avian Genus Passerella 127 Passerella iliaca unalaschcensis (Gmelin) Shumagin Pox Sparrow Original description. — Emberiza unalaschcensis Gmelin, 1788, p. 875. Based on "Unalashka Bunting. Arct. zool. 2. p. 364. u. 232. Lath. syn. II. 1. p. 202. n. 48. Habitat in Unalaschca. ..." Type specimen. — Not known to be in existence. Range. — In summer, the Alaska Peninsula, Shumagin Islands, and Unalaska Island; in winter, south at least to extreme southern Cali- fornia. Specimens examined. — 83 (see list, pp. 186-187). Distinguishing characters. — Of the Unalaschcensis group (see p. 89). Within this group, of relatively large size, with large bill, and with the minimum of red or brown coloration seen in this section of Passerella; and thus easily distinguished from fuUginosa, townsen&i or annectens. Compared with sinuosa, unalaschcensis (in both types of color and bill structure [see postea] ) is even less reddish, and with larger bill. Of about the same size as insularis, but with differently shaped bill, and of duller coloration. Remarks. — Unalaschcensis is apparently of rather rare occurrence in most parts of California, to judge from the few specimens con- tained in collections. The series of breeding birds at hand also is unsatisfactory, in that but few localities are represented within the probable summer habitat. The variety of characters represented in birds in the winter series is such as to raise questions regarding their proper classification; questions that can not be answered prior to the accumulation of much more information about the summer home of the subspecies, and some additional material from various points therein. California winter birds may be divided into two categories, one comprising birds of pale brownish coloration, rather ashy in general tone, and with long pointed bill ; the other, birds decidedly darker in appearance, rather plumbeous in tone, and with short, heavy bill. Extremes of these two types are as widely different as any two recog- nized subspecies within the Unalaschcensis group. They are here lumped under unalaschcensis, first, because of intergradation which links them so closely as to indicate nearer relationship between the two than to any other race, as seems to be shown by the winter speci- mens, and second, because of the present impossibility of a satisfactory understanding of conditions iu the summer habitat. Emberiza unalaschcensis Gmelin (1788, p. 875) was based upon the Aoonalashka Bunting of Latham (1782, p. 202), and the Unalashka Bunting of Pennant (1785, p. 364). Eidgway (1901, p. 389, footnote) 128 University of California Publications in- Zoology [Vol. 21 comments upon the applicability of the name as follows: "Although no Passer ella has hitherto been found on Unalaska (unless the Aoona- lashka Bunting of Latham really came from there), it is not at all unlikely that the present form may occur there, at least accidentally. Passerella iliaca unalaschcensis Fig. O. Map showing stations of occurrence of Passerella iliaca unalaschcensis in California, as established by specimens examined by the author. Certainly if any form of the genus does occur on that island it would be the present one rather than the Kadiak form, most of the land birds of the Shumagins and Unalaska being the same, e. g., Leucosticte griseowwcha, Passerima nivalis tonmsenM, Melospiza melodia, cinerea, etc. ' This form inhabits, on the Shumagins and the peninsula, alder 1920] Swarth: Revision of Avian Genus Passerella 129 thickets. No alders now grow on that portion of Unalaska Island within sight from the town of Unalaska or its vicinity, but they may have done so formerly, or may now occur in other parts of the island. At any rate,, it seems better to retain the name unalaschensis for the present form rather than to give it a new one, since the Emberiza unalaschensis of G-melin seems almost certain to have been the present form." There is in the Museum of Vertebrate Zoology a Passerella from the island of Unalaska. This specimen (no. 4585, male adult), col- lected by C. L. Hall, is labelled as from "Ounalaska, Alaska, June 4, 1894." There is no reason to doubt the authenticity of the skin. It is one of a large collection of Alaskan animals that came into the possession • of this Museum, all taken by the same man, and all carefully prepared and fully labelled; furthermore, the date of capture of this skin accords with other dates upon specimens, from the same island, of species well known to occur there. These facts, together with other cor- roborative evidence adduced from the collection as a whole, fix the origin of this fox sparrow beyond any reasonable doubt, even were there occasion to view it with suspicion. This skin is unfortunately a rather poorly made specimen, but it is apparently of the plumbeous colored, heavy- billed type above described. On the other hand, the considerable series of breeding birds from points on the Alaska Peninsula are all of the more ashy colored, slender billed type, also represented in California in winter. These facts point to the further restriction of unalaschcensis, as a plumbeous colored, heavy-billed bird, breeding on Unalaska Island and possibly some adjacent territory, with the consequent naming of a new sub- species (of ashy coloration and with longer, more slenderly pointed bill), breeding on at least the eastern two-thirds of the Alaska Penin- sula. In the lack of an adequate series of birds from Unalaska it is Fig. P. Bills of specimens of Passe- rella iliaca unalaschcensis, natural size. These birds are from different parts of the range of the subspecies, as it is now denned. Note the short, stubby bill of a, as compared with the long, slender one of 6. These structural differences are correlated with differences of coloration. a. Adult male; no. 4585, Mus. Vert. Zool. ; "Ounalaska," Alaska; June 4, 1894. b. Adult male; no. 90835, Amer. Mus. Nat. Hist. ; Muller Bay, Alaska Penin- sula, Alaska; June 1, 1903. 130 University of California Publications in Zoology [Vol. 21 undesirable to take any such formal action, and I prefer merely to indicate what may be the explanation of the two different types of birds I have here included under the subspecific name unalaschcensis. Birds of the Alaska Peninsula type show a range, of variation undoubtedly indicative of intergradation with siwuom, the subspecies next adjoining on the southeast. There are also some individuals apparently inclining toward insuUwis, of the closely adjacent Kadiak Island. This range of variation, including apparent intergrades as above indicated, as well as the two diverse types of unalaschcensis, all occurring among winter birds in California, has added greatly to the difficulty of recognizing the relationships of many specimens there taken. At the outset of this study the proper separation of unalasch- censis and insularis from the confused series brought together, was one of the most difficult problems encountered, and while the present solution of the question seems logical and fairly satisfactory, it must be borne in mind that it is largely hypothetical, and advanced merely as an apparent explanation of the observed facts. Absolute proof of its correctness awaits the study of additional summer specimens. In the range of variation seen in winter birds, all the Alaska Peninsula type show some trace of brown or rufous. Even the summer birds show it to some degree, especially in the breast spotting. The plumbeous colored type of bird is much less rufous, many of this type presenting a dull slaty effect with scarcely a trace of reddish anywhere. In its southward migration unalaschcensis, like sinuosa, apparently avoids the coast to a great extent. This is especially noticeable in California, where nearly all the record stations are at inland points until southern California is reached (see map, fig. 0). Specimens taken at Helena, Trinity County, February 17, at Drytown, Amador County, December 8 and February 29, at Lakeport, Lake County, December 13, and at Berkeley, February 15, are apparently indicative of winter visitants at these points, but they are only scattered indi- viduals, the main winter home of the subspecies apparently lying south of Tehachapi in southern California. In the accompanying list of specimens (pp. 186-187) birds of the plumbeous colored type are marked with a superimposed figure ( x ) above the collection number, and it will be seen that this type of bird in winter is practically restricted to southern California. The paler colored Alaska Peninsula bird, found in southern California too, 1920] Swarth: Revision of Avian Genus Passer ella 131 occurs also in the Sierra Nevada. October 6 is an early date of arrival in the Yosemite region, October 7 in southern California. In the spring unalaschcensis lingers rather late, as shown by two birds taken in Ventura County on May 2 and 3. Unalaschcensis is one subspecies that may prove to winter well south into Lower California, and from the data afforded by the series at hand I am inclined to believe that it does so. The San Pedro Martir Fig. Q. Wings of four subspecies of fox sparrow of the Unalaschcensis group, natural size. u. Passerella iliaoa unalasohcerisis, immature male; no. 26031, Mus. Vert. Zool. ; near Yosemite Point, Yosemite National Park, California; October 30, 1915. 6. Passerella iliaoa sinuosa, adult male; no. 1535, Mus. Vert. Zool.; Hinehin- brook Island, Prince William Sound, Alaska; June 25, 1908. u. Passerella iliaoa toimsendA, adult male; no. 495, Mus. Vert. Zool.; Admiralty Island, Alaska; May 4, 1907. d. Passerella iliaoa fuliginosa, male; no. 27152, Mus. Vert. Zool.; Blue Lake, Humboldt County, California; December 4, 1916. region, at least, should offer suitable surroundings for fox sparrows in winter. There is no information at hand regarding the winter bird population of those mountains, and it will be of interest to note the nature of occurrence of. Passerella in the range. A specimen from Tacoma, Washington, on January 10, and one from Portland, Oregon, November 7, are possibly indicative of winter visitants at these points, surely the extreme northern limits of the subspecies in winter. The Portland bird is not a typical example of unalaschcensis, and is apparently an intergrade toward sinuosa. One from Salem, Oregon, April 24, was probably a migrant. 132 University of California Publications in Zoology [Vol. 21 Passerella iliaca insularis Bidgway Kadiak Fox Sparrow Original description. — Passerella iliaca insularis Ridgway, 1900, p. 30. Type specimen. — No. 52475, U. S. Nat. Mus. ; male adult ; Kadiak, Alaska ; May 17, 1868 ; collected by Ferdinand Bischoff . Range.- — In summer, Kadiak Island, Alaska ; in winter, south along the Pacific Coast to extreme southern California. Specim&ns examined. — 100 (see list, pp. 188-189). Distmgtrishmg characters. — Of the Unalaschcensis group (see p. 89). Compared with fuUgmosa, toumsendi, and annectens, insularis is of paler coloration, of slightly greater general size, and with a notably larger bill. Compared with sinuosa and unalaschcensis, in- sularis is of brighter, more ruddy coloration. The bill of insularis is heavier than in siwwsa, about the bulk of that of unalaschcensis but differently shaped (see fig. S). A specimen of insularis in juvenal plumage at hand differs from the corresponding stage in the neigh- boring races sinuosa and annectens in just such color characters as are shown in the adults. It is of a brighter hue than the darker red annectens, and decidedly ruddy compared with the relatively grayish young sinuosa. Remarks. — Insularis is apparently restricted in summer to Kadiak Island. It has also been tentatively ascribed to Middleton Island, near Prince William Sound (Bidgway, 1900, p. 30), but the two specimens from this island that I have examined, and upon which this claim may have been based, hardly justify this conclusion. These two birds are not exactly alike, but to my mind they both most closely resemble the subspecies sinuosa (see p. 136). The series of winter birds does not suffice for as explicit an out- lining of the winter habitat as is possible with some of the related subspecies, but there are, nevertheless, certain features of the distri- bution at this season that seem fairly well demonstrated by the data at hand. Judging from the few migrants secured at intervening points, the Kadiak fox sparrow, like the other northern coast f&rms of Passerella, apparently makes long, direct flights in its migrations to and from the summer home. Once central California is reached, insularis does not seem to shun the seacoast, as do its near relatives unalaschcensis and sinuosa; on the contrary, it favors the coastal region and avoids the Sierras. There are no specimens of insularis at hand from any point on the California coast north of Marin County, though many Passerellas have been examined from this section ; in Marin County insularis is a fairly common winter visitant, but outnumbered by annectens. In the Sierra 1920] Swarth: Revision of Avian Genus Passerella 133 Nevada, though there are several record stations of capture, the number of individuals of insularis secured forms but a small per- centage of the total number of Passerellas taken. The Pacific slope of southern California apparently forms the main winter home of this Passerella iliaca insularis DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY VivnJURr or cAiar orhia Fig. E. Map showing stations of occurrence of Passerella iliaca insularis in California as established by specimens examined by the author. The region enclosed within the dotted outline is believed to be the main winter habitat of this subspecies; in all probability most of the birds winter in the southern half of this region. The general avoidance of the mountains of the interior is in marked contrast to the distribution of the closely related unalcsohoensis and sinuosa. subspecies, and even here there is apparently a local restriction to regions fairly near the coast. Insularis is the most common form of the Unalaschcensis group in the western part of the San Gabriel 134 University of California Publications in Zoology [Vol. 21 Mountains, in the vicinity of Pasadena. To the eastward, in the canons above Claremont and Pomona, it is far outnumbered by sinuosa, and this appears to be the case in the San Bernardino Moun- tains also. There are only a few specimens at hand from that range, and there are no examples of insularis among them. September 23, at Echo, Eldorado County, and September 25, at Hume, Fresno County, form early dates of arrival in the Sierra Nevada. There is one specimen at hand from Lakeport, Lake County, taken as early as September 7. In southern California dates of capture lie between October 19 and February 25, though it can hardly be doubted that the birds arrive there earlier and depart later than is indicated by these dates. Farther north there are specimens secured at Stony Ford, Colusa County, March 5, at Oakland, Alameda County, March 14, and at Stockton, San Joaquin County, April 8. Following are dates and places of capture between California and Alaska: Fort Klamath, Oregon, April 13; "Westminster Junction, British Columbia, April 27; New Westminster, British Columbia, March 22; Parksville, Vancouver Island, April 26; Howkan, Alaska, April 25; Kuiu Island, Alaska, April 30. These data indicate about the time of the northward migration, but the specimens of insularis thus represented form so small a percentage of the fox sparrows taken in the region that to my mind they should be considered as little more than strays from the regular line of travel. They will serve, perhaps, to indicate dates of passage, but to all appearances the mode of travel is mostly a long, direct flight, rather than a more leisurely series of short flights with frequent stops. One specimen from Clackamas County, Oregon, February 16, possibly indicates the occasional wintering of insularis as far north as this. It is rather singular that among the non-typical winter birds, referred perforce to insularis, the form they most nearly resemble, there should be so many apparently leaning toward iliaoa. It is a question whether there can be actual intergradation between the two subspecies, with the ranges of unalaschcensis or sinuosa interposed between them, and with insularis restricted to an island habitat. The variants referred to usually resemble insularis in size of bill and bright, ruddy coloration, but show more or less obscure mottling or streaking upon the back, more or less distinct wing bars, and rather sharply defined breast markings. They may really be intergrades between unalaschcensis and iliaca, owing their bright coloring, and 1920] Swarth: Revision of Avian Genus Passerella 135 consequent superficial resemblance to insularis, to their iliaca strain, but this is a point that can hardly be determined from winter birds alone. Passerella iliaca sinuosa Grinnell Valdez Pox Sparrow Original description. — Passerella iliaca sinuosa Grinnell, 1910, p. 405. Type specimen. — No. 1593, Calif. Mus. Vert. Zool. ; male adult; Drier Bay, Knight Island, Prince William Sound, Alaska ; August 26, 1908 ; collected by J. Dixon ; original number 309. Fig. S. Bills of three subspecies of fox sparrow of the TJnalaseheensis group; natural size. a. Passerella iliaca insularis, adult male; no. 9655, Mus. Vert. ZooL; Kuiu Island, Alaska; April 30, 1909. 6. Passerella iliaca sinuosa, adult male; no. 1591, Mus. Vert. ZooL; Hawkins Island, Prince William Sound, Alaska; June 22, 1908. c. Passerella iliaca anneetens, adult male; no. 7198, Mus. Vert. Zool.; Nicasio, Marin County, California; February 21, 1909. Range. — In summer the Prince William Sound region (both on the islands and the mainland), Middleton Island, and the Kenai Peninsula, Alaska. In winter south on the Pacific slope at least as far as southern California. In its migrations and winter habitat sinuosa for the most part avoids the immediate vicinity of the coast. The main winter habitat appears to be the eastern part of the San Diegan region, southern California, but a certain percentage of the birds winters in the foothills along the west slope of the Sierra Nevada, and small numbers even as far north as central Oregon. Specimens examined. — 365 (see list, pp. 189-194). Distinguishing characters. — Of the Unalaschcensis group (see p. 89). Bill rather long and slender, not as heavy as in unalaschcensis and insularis, but larger than in the more stubby-billed anneetens, townsendi, and fuliginosa (see fig. S). Coloration rather more reddish than in unalaschcensis, but decidedly less so than in the brighter colored insularis, and the darker anneetens and townsendi. A notice- ably ashy tinge dorsally and on sides of neck. Remarks. — The simmer habitat of sinuosa lies between the ranges of unalaschcensis and anneetens, and, in physical characters also, 136 University of California Publications in Zoology [ YoL - 21 sinuom occupies an intermediate position between these two subspecies. It, is of less slaty, more brownish, coloration than unalaschcensis, but it is decidedly less brownish than the ruddy-colored armectens to the southward. In bill structure also it is intermediate, with bill smaller than in unalaschcensis, but rather larger than in annecte-ns. The con- siderable series of sinuosa at hand from Prince "William Sound includes many breeding adults, some juveniles, and some adults and im matures in fresh fall plumage. It is thus illustrative of age and seasonal varia- tion, and there is. considerable individual variation exhibited also, but nothing that can be correlated with local distribution. That is, birds from the several islands and from the mainland, as well as from the opposite sides of the sound, are all intrinsically alike. (See Grinnell, 1910, p. 407.) Two adult specimens from Montague Island (nos. 1569, 1572, Mus. Vert. Zool.), exhibit, perhaps, the extremes, one of brown, the other of gray coloration as seen in the entire series. Six specimens at hand from points on the Kenai Peninsula are clearly referable to smuosa. These birds, from the collection of the American Museum of Natural History, were referred to by Chapman (1902, p. 243; 1904, p. 404) under the names annectens and unalasch- censis, the subspecies sinuosa not having been described at that time. Two birds from Middleton Island (nos. 115823, 115824, U. S. Nat. Mus.) appear to me to be clearly referable to sinuosa. These birds, in fresh fall plumage, were taken on August 26, a date sufficiently late to allow of their being perhaps wanderers from some other point, though they are more probably representative of the form breeding upon this island. They are very slightly aberrant from the mode of the Prince William Sound series in that both are a trifle paler colored and have distinctly more buffy under tail coverts, and that one has a slightly larger bill. It is interesting to note that two winter speci- mens from Echo, Eldorado County, California (nos. 88757, 88758, Amer. Mus. Nat. Hist.) duplicate these Middleton Islands birds in every particular. It is not uncommon, of course, to find winter speci- mens so different from the normal type of any of the described sub- species as to be difficult to place satisfactorily, but there are only one or two instances in the large series of fox sparrows here assembled where such birds can be linked up with a summer habitat, as appears to be the case with the two just mentioned. There are many winter specimens at hand from points in Califor- nia, which, though referred perforce to sinuosa, exhibit among them- l 92 °] Stmrth: Revision of Avian Genus Passerella 137 selves a much wider range of variation than is seen in the series from the summer habitat, showing practically every degree of divergence from typical sinuosa, on the one hand toward unalaschcetisis, on the other toward wmectens. These different types of birds, it is to be ■inferred, hail from appropriately intermediate localities between the summer home of typical sinuosa and the homes of the races they apparently approach in appearance. "Whether the whole Passerella population from any one such place is uniformly of intermediate appearance, or whether in such regions a percentage of intermediates occurs among a preponderantly typical population, are questions which can not now be answered. The field work carried on at such points has as yet been insufficient to solve these problems. Among winter birds there are other non-typical specimens also, apparently referable to sinuosa, but showing a trend of variation not to be explained as an approach either to unalaschcensis or to annect&ns. There are four birds from Vancouver Island (nos. 1623, 4082, 4091, 4095, Victoria Mem. Mus.) that I have provisionally placed with sinuosa though they are certainly not typical examples of that sub- species. One of these (1623) taken at Victoria on April 29, and probably a north-bound migrant, is, perhaps, to be explained as an intermediate toward annectens; but the other three, winter birds from Departure Bay, present peculiarities of dark coloration and thick bill not to be explained as an approach toward any other neighboring race. It is always possible in such a case that while the superficial appear- ance of the specimen is most nearly like some one race (in this case sinuosa), its real affinities lie elsewhere ; and that its peculiar characters may be due to an intermediateness of distribution in summer at present not understood. There are such large expanses of territory in the Pacific northwest where fox sparrows must occur and where no collecting has been done, and the manner of variation throughout the group is as yet so imperfectly known, that it is too much to be expected that the peculiarities of all winter specimens collected be perfectly understood. Aside from these aberrant specimens there is no example of sinuosa at hand from any point between Prince William Sound and northern Oregon, though extensive collecting has been carried on at many places in this region. The inference to be drawn is that travel between summer and winter homes is for the greater part of the distance by one continuous flight with few or no stops. 138 University of California Publications in Zoology [Vol. 21 Most California specimens of sinuosa come from inland points. There is evidently very little travel immediately along the coast, and only a small proportion of the birds winters in the coastal region. O Passerella iliaca sinuosa Fig. T. Map showing stations of occurrence of Passerella iliaca sinuosa in California, as established by specimens examined by the author. The regions enclosed within the dotted outlines are believed to comprise the main winter habitat of the subspeeies. The stations outside these areas are points where the specimens of sirmosa taken were migrants, or, if winter visitants, where they formed but a small percentage of the total Passerella population. Prom Humboldt Bay there are two or three migrating individuals and no midwinter birds; from Marin County, six winter specimens out of a total of 63 Passerellas, and one or two additional migrating 1920] Swarth: Revision of Avian Genus Passer ella 139 sinuosa; from the Santa Cruz region 28 wintering sinuosa out of 90 Passerellas. On the other hand, from points in the Sierra Nevada there are 23 wintering sinuosa out of a total of 51 Passerellas, and from the San Diegan region, 124 out of a total of 396. An interesting feature of the winter distribution of sinuosa lies in the manner of its occurrence in the San Diegan region (see table 3) . Sinuosa is the predominant winter fox sparrow in the foothills and canons at the eastern extremity of the San Gabriel Mountains. An extensive series of Passerella from the "Wright M. Pierce collection, mostly from San Dimas and San Antonio canons, is available in the present study, and there are some additional birds from the same general region available from other sources. Out of a total of 178 winter specimens of Passerella from this section, 83 are sinuosa. Only 29 of the remaining 95 fox sparrows are representative of other sub- species of the Unalaschcensis group (unalaschcensis, insularis, and annectens). On the other hand, from the region about Pasadena and in the adjacent mountains, toward the western extremity of the San Gabriel Range, there are but nine wintering sinuosa out of a total of 80 Passerellas. Here insularis is the predominant form. Specimens, to all appearances referable to sinuosa, from Corvallis, Oregon, in January, from Netarts, Oregon, in December and January, and from Mercer, Oregon, in February, are indicative of the wintering of at least occasional individuals as far north as this. There are a few winter birds at hand from scattered points in the interior of northern California : the Siskiyou Mountains, October 31 ; Clear Lake, December and February; Colusa, January; Eagle Lake, October 27. Undoubtedly, however, the main winter home of sinuosa lies in the foothills at the western base of the Sierra Nevada, and in parts of the San Diegan region. The series from Prince William Sound includes specimens taken from June 25 to September 2. "While this gives no clue as to the date of arrival on the breeding ground, it seems probable that the latter date, September 2, is close to the time of final departure for the south. Among the birds from California and Oregon there are scattered specimens that were secured upon dates affording some information regarding the time of arrival at the winter habitat and departure therefrom. Of autumnal arrivals in the Sierra Nevada there are birds from Sierra City, Sierra County, September 22; and from Horse Corral Meadow, Fresno County, September 21. This latter date may 140 University of California Publications to Zoology [ VoL - 21 be accepted with especial confidence as the date of arrival at that point on that particular year (1916), since fox sparrows had been assiduously sought for some time before. There is a specimen from Kunz, Trinity County, taken September 22; from the coast region there are specimens from Fortuna, Humboldt County, September 23, and from Oakland, Alameda County, September 28. The earliest date of arrival from southern California is Strawberry Valley, San Jacinto Mountains, September 28. There are but few specimens at hand from California of later date than February, though occasional captures indicate that the last sinuosa does not leave for the north until much later. One bird from Santa Cruz Island was secured on April 27, much the latest date of capture in California shown in the series here available. Other late spring dates are: Newhall, Los Angeles County, March 4, Bolinas, Marin County, March 31, and Eureka, Humboldt County, March 3. From Oregon there are birds from Portland, Multnomah County, March 23, from Salem, Marion County, April 23, and from "Warm Spring, Cook County, April 30. Passerella iliaca annectens Eidgway Yakutat Fox Sparrow Original description. — Passerella iliaca annectens Bidgway, Auk, 17, Jan., 1900, p. 30. Type specimen. — No. 170222, U. S. Nat. Mus. ; male adult ; Yaku- tat, Alaska, June 20, 1899 ; collected by R. Eidgway. Range. — In summer on the coast of Alaska in the Yakutat Bay region ; probably extends northwestwardly nearly to the eastern edge of Prince William Sound, southeastwardly about to Cross Sound. In winter restricted almost entirely to the immediate vicinity of the coast in central California, from Marin County south through Monterey County. A few stragglers reach southern California on the coastal ■ slope. Specimens examined. — 170 (see list, pp. 194-197). Distinguishing characters. — Of the Unalaschcensis group (see p. 89). In geographical position annectens lies between sinuosa and townsendi, and in its characters it forms a link between the two. Coloration is brighter, more ruddy than in sinuosa, not so dark as in townsendi; breast spotting less heavy than in townsendi. In fresh fall plumage, general coloration about cinnamon-brown; breeding birds have lost much of the bright, ruddy appearance of the fall plumage, and are duller in appearance. Annectens is of slightly smaller general size than sinuosa, insularis and unalaschcensis, to the northward, slightly larger than townsendi, to the southward (see table 4). Bill notably smaller than in unalaschcensis and insularis, slightly smaller than in sinuosa (see fig. S). !920] Swarth: Revision of Avian Genus Passerella 141 Remarks. — Vigors (1839, p. 19) described Fringilhi meruloides, a species that "was met with at Monterey" by the naturalists of H.M.S. "Blossom." His description is brief, but evidently pertains to a fox sparrow, and to one of the Unalaschcensis group, although it is not sufficiently explicit to be pinned down to any one of the six subspecies within the latter aggregation. Therefore the proper application of the name meruloides rests to some extent upon outside facts, inde- pendent of the written description. Vigors gives no dates of occur- rence, but the time at which the "Blossom" visited Monterey, as ascertained from Beechey's "narrative" (1831) was in January, and again in October and November, 1827. It was undoubtedly during the second visit (October 29 to November 17) that the natural history collections were made, but either stay was at the proper season to find fox sparrows there. No Passerella is known to breed in that region (see Grinnell, 1902, p. 44). Although Vigors' description cannot be used to distinguish between the several closely related subspecies concerned, it so happens that at Monterey there is one particular race that occurs in greater numbers than any of the others. The Passerella that winters there in such abundance is, as verified by com- parison of specimens, the form breeding in the Yakutat Bay region, Alaska, named by Kidgway (1900, p. 30) Puss&rella, iMaoa annmtens. Collecting carried on at Monterey in December, 1901, by Grinnell (1902, p. 44) resulted in the capture of but the one subspecies; exam- ination of additional specimens from the same general region showed these also to be of the same form. On this basis Grinnell (loc. cit.) pointed out that Vigors' name, meruloides, could be used with pro- priety for this subspecies, as apparently the only form occurring in the region. "With this view the present writer was in perfect accord until a very recent date. "When the writing of the present report was begun there were at hand seventeen fox sparrows from the vicinity of Monterey. Of these, fifteen were of the Yakutat Bay subspecies, one was sinuosa and one insularis, a representation of subspecies that bore out Grinnell 's assertions, above indicated. A little later, however, additional material was received that put a different aspect upon the matter. In December, 1918, Mr. H. G. White, of this Museum, spent some weeks collecting birds and mammals in the immediate vicinity of Monterey. As one result of his labors he sent in thirty-one fox spar- rows. Of these, thirteen are of the Yakutat Bay race, sixteen are sinuosa, and there is one specimen each of iUaoa and megarhynchus. 142 University of California Publications in Zoology [Vol. 21 Altogether there are at hand forty-six fox sparrows from Monterey, including twenty-six of the Yakutat Bay race, seventeen sinuosa, and one each of iliaca, insularis, and megarhynchus. The name meruloides can be applied to one particular subspecies only on the assumption that the chances are overwhelmingly against any other race having been taken at the specified locality. As shown by the recent collecting indicated above, while the one form is by far the most abundant in the region, there is at least one other that occurs in numbers. Altogether five subspecies are here recorded from that place. Under the circumstances it seems to me that there can be no absolute certainty as to the bird Vigors had in hand. On the contrary, it seems apparent that there is no alternative but to discard Vigors' name meruloides as undeterminable, as has been done by Ridgway (1901, p. 390). That name is replaced by Passerella iliaca annectens Ridgway (1900, p. 30), with type locality at Yakutat Bay, Alaska. There is another angle to the question, in that the winter birds bearing the closest resemblance to Yakutat Bay specimens, both as regards color and bill structure, and making due allowance for seasonal changes in color, are those from Marin County. The Yakutat speci- mens at hand are all midsummer birds, taken a month or so prior to the molt, but I can not conceive that these birds, even in the freshest of plumage, ever presented the bright, ruddy appearance so frequently seen in fall and winter specimens from the vicinity of Palo Alto and of Monterey Bay. These latter seem to present an extreme of differ- entiation within this subspecies, both as regards bright color and increased size of bill. Discovery of a definable summer range of birds with these characters may necessitate the recognition of still another race, with annectens occurring at Yakutat Bay in summer, and mainly in the region immediately north of San Francisco Bay in winter. At present, however, no such division can safely be made, but the birds from the several points indicated must all be lumped as exemplif3'ing the range of variation within the one subspecies. In the series of Pmserella here accumulated, annectens is abun- dantly represented from many points near the coast, from Sonoma County south through Monterey County, and by strikingly few speci- mens from other parts of the state. Although this is such a common winter bird in the San Francisco Bay region and in Marin County, to the northward, it is almost entirely absent from the California coast region but a little way farther north. In an extensive series of fox sparrows from the Humboldt Bay region, but one example of 1920] Swarth: Revision of Avian Genus Passerella 143 annecteps was discovered. This bird, taken September 20, was prob- ably a migrant at this point. Two others specimens, from Logan, Oregon, March 18, and from Tacoma, "Washington, November 10, Passerella iliaca annectens DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY uhivehhiy or CAuromiA Fig. TJ. Map showing stations of occurrence of Passerella iliaca annectens in California, as established by specimens examined by the author. The region enclosed within the dotted outline has been satisfactorily demonstrated to be the main winter habitat of the subspecies. indicate at least the occasional appearance of the subspecies at these points, but in all . probability it is of uncommon occurrence there. In the winter metropolis, the coastal region of central California, dates of occurrence range from September 25 (Sebastopol, Sonoma County) to April 19 (Hayward, Alameda County). In southern 144 University of California Publications in Zoology [Vol. 21 California, specimens have been taken from October 22 (Arroyo Seeo, near Pasadena) to February 14 (San Dimas Canon). The only inland points where specimens have been taken are Hume, Fresno County (in the Sierra Nevada at 5300 feet), September 26, at Stockton, San Joaquin County, November 28, and at Tower House, Shasta County, March 8, one specimen at each place. Specimens of wmectens from Clayoquot, Vancouver Island, September 17, from Kalama, Cowlitz County, "Washington, October 18, and from Beaverton, "Washington County, Oregon, from February 25 to April 22, are presumably migrants. It is a point worth noting that not only are the limits of the winter habitat sharply defined and not only are by far the greater number of these birds found in this habitat, but also a large propor- tion of the stragglers found elsewhere are not absolutely typical of the subspecies. Intergrades occur, linking amwctems with svnuosa, on the one hand, and with townsendi, on the other, and many of the individuals taken outside the usual winter habitat are of such equivocal character, though referred to annectens as the form they most nearly approach. Passerella iliaca townsendi (Audubon) Townsend Fox Sparrow Original description. — " Plectrophanes townsendi Audubon, Birds Amer. (folio), volume IV, part 85, plate 424, figure 7, engraved, printed and colored by Eobert Havell, 1838. "Fringilla townsendi Audubon, Ornithological Biography, vol. V, p. 236, 1839. ' Shores of the Columbia Biver. ' ' '* Type specimen.' — No. 2874, U. S. Nat. Mus. ; female adult ; Colum- bia River, Oregon ; February 15, 1836 ; collected by J. K. Townsend. Range. — In summer the coast region of southern Alaska, on islands and mainland, from Glacier Bay and Lynn Canal south to Forrester Island; also on the Queen Charlotte Islands, British Columbia. In winter south on the coast of northern California ; to Humboldt Bay, abundantly, to the San Francisco Bay region in small numbers. One specimen from Arizona. Specimens examined. — -155 (see list, pp. 197-199). * These references are copied verbatim, as given me by Mr. W. H. Osgood, Assistant Curator, Mammalogy and Ornithology, Field Museum of Natural History, from copies in the library of that institution, the works cited not being available for my own personal inspection. I wish here to express my appreciation of his courtesy in supplying me with the needed references. Authors in citing the original description of this subspecies give sometimes one and sometimes the other of the above volumes, or a mixture of the two under one heading (cf. A. O. U. Cheek-List, 1910, p. 278; Ridgway, 1901, p. 393), a procedure that is apt to cause confusion where the actual circumstances are not understood. 1920] Swarth: Revision of Avian Genus Passerella 145 Distinguishing characters. — Of the Unalaschcensis group (see p. 89). Compared with the subspecies to the northward, unalaschcensis, insularis, svnuosa, and wmectens, coloration is much darker, more rufescent, spots on breast larger and more crowded. Compared with fuliginosa, to the southward, townsendi is brighter colored, less sooty. General size slightly less, and size of bill appreciably less than in unalaschcensis, msularis, and swmosa, slightly less even than in annectens (see table 4). An average fall specimen just through the molt, in general coloration of upper parts (head, neck and back) is close to bister. Spring birds are appreciably paler colored, about vandyke brown. Remarks. — The Townsend fox sparrow is a notable example of a bird with a winter habitat nearly as sharply denned as its summer home. In the Humboldt Bay region of California it is an abundant winter visitant, but even here it is evidently restricted to a narrow strip of territory immediately adjoining the coast. Of forty-six fox sparrows examined from the Humboldt Bay region taken during the winter months, forty are toumsendi; of sixty examples of townsendi taken in California (migrants and winter visitants), forty-nine are from the Humboldt Bay region. From these figures it is apparent that in California townsendi is closely confined to this particular region ; it is also evident that it is the predominant form of Passerella there. The only other fox sparrows I have seen from that section are a few winter visitant fuliginosa, as listed beyond, several migrating sinuosa, and one migrating annectens. The last mentioned subspecies is so abundant a winter visitant to the California coast to the south- ward, in the regions about San Francisco and Monterey bays, as to make its rarity at Humboldt Bay a matter of surprise and of evident significance. Conversely, the subspecies townsendi is almost as rare in the winter metropolis of annectens. The winter habitats of the two forms are nearly as sharply defined as the breeding ranges, and there is very little straying on either side. The summer home of the Townsend fox sparrow includes a long stretch of coast in southeastern Alaska, but the bird apparently is not continuously distributed throughout this area. The 1907 Alexan- der Alaska Expedition found this subspecies breeding at Glacier Bay (about the northern limit of the form), and at various points on Admiralty, Baranof, and Chichagof islands (see Grinnell, 1909, p. 232). . The 1909 Alexander Alaska Expedition did not find the bird on any of the more southern islands of the Alexander Archipelago, nor at the more southern mainland points touched upon (see Swarth, 1911, p. 93). It is, however, known to be a common summer visitant to 146 University of California Publications in Zoology [Vol.21 Forrester Island, near the southern extremity of the Alexander Archi- pelago, and there is a June specimen at hand from Wrangell Island, where I myself failed to find it. Specimens from Howkan, Long Island, taken March 29 and April 6, were probably newly arrived migrants, but one secured there on April 27 may well have been upon its nesting ground. The available facts, however, all things considered, point to a rather uneven distribution within the known summer habitat. The specimen above mentioned as secured at Howkan (near the southern extremity of the Alexander Archipelago) on March 29, is probably representative of as early a date of arrival upon the breeding grounds as usually occurs. The Townsend sparrow was plentiful on Admiralty Island by April 17, and eggs were found there on May 3 (Grinnell, loc. cit.). The present writer found the subspecies abun- dant at Port Snettisham (on the mainland opposite Admiralty Island) at the end of August; as birds in juvenal plumage were secured the inference is that they were upon their breeding ground. Townsend fox sparrows were seen on the Taku River early in September. The last was noted September 19; as the writer remained in that region until the end of September without seeing another one, that date may be assumed to indicate the final departure for the year (Swarth, loc. cit.) The first arrival in the winter home may be represented by a bird (no. C. 292, Dickey coll.) taken at Fortuna, Humboldt County, California, September 19. There is another specimen at hand (no. 2432, Emerson coll.) from Hay wards, Alameda County, California (near the southern limit of the subspecies), taken September 20. The latest date in the spring represented in the available series is of a specimen (no. 21, Dickey coll.) taken at Eureka, Humboldt County, California, on March 3. It is very probable, however, that this sub- species really has not finally departed for the north until some weeks later. In California, as noted, the Townsend fox sparrow is quite closely restricted to the Humboldt Bay region. Additional specimens from scattered points in Sonoma, Marin, Alameda, and San Mateo counties are in some cases typical examples of townsendi; in others they show an obvious tendency toward annectens. As 'no stray specimens of townsendi have appeared among the hundreds of fox sparrows exam- ined from southern California, the occurrence of an individual of this subspecies from a point in southeastern Arizona seems doubly strange. There is, however, one specimen at hand (no. 3178, Law coll.) collected 192 °] Swarth: Revision of Avian Genus Passerella 147 in Pinery Canon (altitude, 6000 ft.), Chiricahua Mountains, Arizona, November 28, 1914, that so closely resembles typical townsendi that it must perforce be placed in that category. It is a trifle duller colored, less richly brown, than Alaskan specimens, but the difference is slight. The series of townsendi from throughout the breeding range is on the whole notably uniform in appearance, despite the segregation of the birds on different islands. From the northern extreme, Glacier Bay, there are, unfortunately, but two juveniles available. These are not appreciably different from young from other points, and they are unlike the one juvenal annectems at hand, but the material is too scanty for a decision as to the existence of a tendency toward annectens at this point. Birds from Forrester Island, an outlying island near the southern boundary of Alaska, do not depart appreciably from the mode of townsendi, nor, as far as I can see, do those from the Queen Charlotte Islands, still farther south. A series of the last mentioned was reported upon by Osgood (1901, p. 48), who remarks that the birds ' ' are not identical with breeding birds from Sitka, and perhaps should be considered intermediate between townsendi and fuliginosa." I do not know the amount of comparative material that was available when that report was written, but in the light of the extensive series now assembled, the Queen Charlotte Islands specimens can not be regarded as outside the range of variation seen elsewhere in the habitat of townsendi. They certainly show no appreciable approach toward fuliginosa, though such variation would not have been surprising. Birds in fresh fall plumage from the Taku River and Port Snetti- sham are notably dark-colored as compared with breeding birds from neighboring islands (see Swarth, 1911, p. 94), and this darkness is also apparent in comparison with a September bird from Sitka. There is a fall specimen at hand taken at another point on Baranof Island (Peril Strait) of the same dark type of coloration as is seen on the mainland birds. The type specimen of Plectrophanes townsendi Audubon has been available for the present study. This is a well made skin, considering the period at which it was collected, and in good condition. Taken late in the winter (February 15) the condition of the plumage would not be noticeably different from birds collected six weeks later upon the breeding ground, and to my eye the coloration of this specimen is not appreciably different from selected specimens from the Alex- 148 University of California Publications in Zoology [ VoL - 21 ander Archipelago and the Queen Charlotte Islands. There can be no shadow of a doubt as to the applicability of the name townsendi. The only label upon the skin (besides the red "type label") was ° Passerella iliaca townsendi A Passerella iliaca fuliginosa DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY 1 OF CM IFORNIA Fig. V. Map showing stations of occurrence of Passerella iliaca townsendi and P. i. fuUginosa in California, as established by specimens examined by the author: The region enclosed within the dotted outline has been satisfactorily demonstrated to be the only part of California visited commonly by townsendi; it is evidently the southern extremity of the normal winter habitat of the sub- Fuliginosa is of little more than casual occurrence in California. presumably attached to it by Professor Baird. Upon the face of the label the words "Smithsonian Institution, Washington, D. C." have been crossed out with a pen, and "Coll. of S. F. Baird" written in. Upon the back of the label are written the words "Prom Audubons 192 °] Swarth: Revision of Avian Genus Passerella 149 Coll. Property of S. F. Baird." The locality of capture, Columbia Eiver, Oregon, must necessarily be regarded as established by the statement in the original description, rather than that upon the label, which must have been attached at a much later date. Passerella iliaca fuliginosa Ridgway Sooty Pox Sparrow Original description. — Passerella iliaoa fuliginosa Ridgway, 1899, p. 36. Type specimen. — No. 157611, Biol. Surv., U. S. Dept. Agric. ; male adult; Neah Bay, Clallam County, Washington; June 10, 1897; col- lected by E. A. Preble; original number 319. Range. — Imperfectly known. In summer known to occupy parts of Vancouver Island and northwestern Washington; presumably the mainland coast of southern British Columbia also. In winter non- typical birds occur in small numbers along the coast as far as San Francisco Bay, rarely to southern California. Specimens examined. — 39 (see list, pp. 199-200). Distinguishing characters. — Of the Unalasehcensis group (see p. 89). Much darker than any of the other subspecies, and also more heavily marked. One specimen at hand in freshly acquired winter plumage is about clove brown on the head, and very slightly more castaneous on the lower back. The lateral under surface of the body and the spots on the breast are dull and sooty, with hardly an indica- tion of reddish. In general size fuliginosa is slightly larger than townsendi. Remarks. — This is one of the least known of the subspecies of fox sparrows. Its summer home is in a region where relatively little bird collecting has been done; its migrations at any season are extremely limited, and do not carry it into regions where specimens are much more likely to be taken than in the summer habitat. As a result there are but few skins in collections. Owing to the diversity of appearance exhibited by the specimens, the series here assembled, small though it is, raises certain questions that can not be answered prior to the accumulation of much additional material. The type specimen of fuliginosa (a summer bird from northwestern Washington) possesses a relatively long and slender bill. This is seen also in the more fully grown young birds from Vancouver Island, and again in some winter specimens from Van- couver Island. Other winter birds from the latter place, however, though of dark coloration and with the heavy markings of fuliginosa, have noticeably short, stubby bills. Winter birds from California, though here listed under the name fuliginosa, are so placed by the 150 University of California Publications in Zoology [Vol. 21 author reluctantly and with a strong sense of dissatisfaction. They are certainly dark colored, but of a duller tone than in typical fuli- ginosa, they are not so heavily marked below (that is, the streakings are less crowded), the under tail coverts are more nearly whitish, and they all have stubby bills. There are no specimens available from the mainland coast of British Columbia, and I have consequently no knowledge of the characters of the birds of that region. The nature of variation that may be assumed to exist at the outskirts of the range of fuliginosa, perhaps showing inter gradation between fuliginosa and toumsendi, perhaps between fuliginosa and altimagaMS, and fuliginosa and schistacea, is entirely unknown, and until series of specimens are gathered from many points where such intergrades are to be looked for,, and the summer habitats of the several subspecies concerned con- sequently platted in much greater detail than is now possible, there is little to be gained by speculation as to the nature and origin of the puzzling winter birds above described. It may be said with all certainty, however, that the California winter visitants here included under fuliginosa were not migrants from northwestern Washington or Vancouver Island. With the exception of one or two specimens from Humboldt Bay they do not even approach typical fuliginosa very closely in appearance ; they are placed in that category because their characters are such as to indicate a closer affinity to fuliginosa than to any other form, and probably illustrate intergradation between fuliginosa and some one of the adjacent subspecies. In what little is known of the summer habitat of fuliginosa there are certain apparent inconsistencies of occurrence. Dawson (1909, p. 152) found this subspecies breeding upon the San Juan Islands, Puget Sound, but the present writer did not meet with it at the lower levels of the closely adjacent southeastern portion of Vancouver Island. Here it was discovered breeding in the Canadian Zone of the higher mountains, from 2200 feet altitude upward. Farther north on Vancouver Island, at Nootka Sound, however, it was seen at sea level (Canadian Zone) (Swarth, 1912, p. 62). While Dawson (1909, p. 155) defines the summer habitat in Washington as on "the north and west slopes of the Olympic Mountains, together with the islands of Lower Puget Sound," a list of the birds of the Lake Crescent region, Olympic Mountains, by Rathbun (1916) includes no mention of Passerella. The present writer collected four specimens of fuliginosa in the summer habitat on Vancouver Island. Three are in juvenal plumage. 1920]- Swarth: Revision of Avian Genus Passer ella 151 and one in almost complete first winter plumage. The last mentioned (no. 16253, Mus. Vert. Zool.), in perfectly fresh, unfaded plumage, is of intensely deep brown coloration, about clove brown on the head, and very slightly more castaneous on the lower back. It is much more heavily marked below than is the case in townsendi, the dark markings - being so nearly confluent across the breast that in this case it is the white ground color that shows through as interrupted markings. The lateral under parts and the dark spots on the breast are dull and sooty, with but a suggestion of reddish. The three juveniles exhibit some diversity of color. In the darkest of them (no. 16252, Mus. Vert. Zool.) the general coloration is bister, as compared with snuff brown in typical young townsendi; the mark- ings below are so extensive, and the ground color so very little paler than these markings, that it appears almost "solid" brown below. The other two young (nos. 16251, 16254) are less heavily marked and not so dark-colored, but they are still appreciably darker than town- sendi at the same stage. Fuliginosa (speaking now of the non-typical birds previously described) is a rare winter visitant to the northern coast region of California. The earliest arrival noted is one from Fortuna, Humboldt County, September 19; the latest in the spring, one from southeast Parallone Island, May 31. As the subspecies had been represented in local collections heretofore only by an occasional specimen (nearly all from the Humboldt Bay region, included among large series of townsendi), it was rather unexpected that in the fall of 1918 there should be a sudden influx of the birds into the San Francisco Bay region. On October 1, 1918, R. M. Hunt, of the Museum of Vertebrate Zoology, noted numbers of fox sparrows in the hills east of the Uni- versity Campus. The following day he returned to the place and collected six of the birds. Four of the six proved to be fuliginosa. On October 18 four more fox sparrows were collected at the same place, and three of them were fuliginosa. Later collecting failed to disclose more of these dark-colored birds. On September 26, 1918, J. Grinnell collected an immature female fuliginosa at Morro, San Luis Obispo County. In the light of these several captures, and con- sidering the absence of the subspecies in the extensive series of fox sparrows available and collected in the San Francisco Bay region during previous years, there seems to be evidence of a rather unusual "wave" of this bird. There are certain other birds, more con- spicuous in actions and appearance (the varied thrush is a good 152 University of California Publications in Zoology [Vol.21 example) that we know to vary in numbers from year to year, and it may be that among the subspecies of the fox sparrows also there is this question of fluctuating numbers (or even total absence some winters compared with relative abundance during others) to make still more complicated the already involved problem of the migration and distribution of the different forms. One specimen of fuligmosa at hand (no. 27153, Mus. Vert. Zool.), like the Berkeley birds in appearance, was taken at Bear Flat, San Antonio Canon, San Gabriel Mountains, California, on November 30, 1916. This is by far the southernmost record for the subspecies. Fig. W. Wings of three subspecies of fox sparrow of the Sehistacea group, natural size. a. Passerella iliaca sehistacea, adult male; no. 9054, Mus. Vert. Zool.; Pine Forest Mountains, Nevada; June 19, 1909. o. Passerella iliaca mariposae, adult male; no. 29089, Mus. Vert. Zool.; Sisson, Siskiyou County, California; May 15, 1918. o. Passerella iliaca stephensi, adult male; no. 3426, coll. G. F. Moreom; Bear Valley, San Bernardino Mountains, California; June 10, 1886. Passerella iliaca sehistacea Baird Slate-colored Fox Sparrow Original description. — Passerella sehistacea Baird, 1858, p. 490. Type specimen. — No. 5718, U. S. Nat. Mus. ; female ; ' ' Platte river, K. T." (== "South Fork of Platte Eiver, about 25 miles east of northeastern corner of Colorado, Nebraska" [A. O. U. Check-List, 1910, p. 277]) ; July 19, 1856; collected by Lt. F. T. Bryan (original number 131). Range. — The summer habitat is principally in the high mountains of the Great Basin region, north into Alberta, south into Nevada and east probably to eastern Wyoming. In winter south to southern California and (casually, at least) to southern Arizona. Occasional in migration in western Nebraska and Kansas. 1920] Sumrth: Revision of Avian Genus Passerella 153 Specimens examined. — 98 (see list, pp. 200-201). Distinguishing characters. — Within the Schistacea group (see p. 89), schistacea is distinguished from all the other subspecies save canescens by its diminutive bill (see figs. E, P, and table 4). From canescens, schistacea is distinguished by its more slender pointed, less stubby bill, and by color, canescens being more grayish in general coloration; schistacea more brownish. In an average example of schistacea in fresh fall plumage (no. 23203, Mus. Vert. Zool., Blue Canon, Placer County, California, August 31, 1912) the upper parts generally are overlaid with brownish, the general effect being not far from sepia, while the ventral spots, each one blackish centrally, are laterally of a richer brown than the back. By April much of the brownish tinge has disappeared through wear and fading, but direct comparison of specimens of schistacea and canescens shows an appre- ciable difference, still, in this regard. In juvenal plumage schistacea exhibits throughout a brownish suffusion, just such as is seen in the later stages. Remarks. — The type specimen of Passerella schistacea Baird was collected upon the expedition headed by Lieut. Francis T. Bryan in exploration of a road from Fort Eiley to Bridger's Pass in the sum- mer of 1856. The locality of capture of the specimen as entered upon the label is simply "Platte river, K. T.," the date, July 19, 1856. The itinerary of Lieut. Bryan's party appears in a report contained among "The executive documents printed by order of the senate of the United States" (1858, pp. 455-457) ; and there is a map showing the route the party traversed, in the Annual Report of the Wheeler Survey (1876, opp. p. 36). In the itinerary, unfortunately, a detailed account of day to day travel begins only on July 23, but by comparison of the less detailed statements just preceding this date, with the table of distances (pp. 485, 486) a fairly accurate idea may be reached of the whereabouts of the party on July 19. On that day, from the above data, they would seem to have been traveling along the South Platte River, probably between Laramie Crossing and Goodale's Crossing, approximately two hundred miles west of Fort Kearny. This would fix the type locality of schistacea at about the point given by Cooke (1897, p. 107), upon the authority of Dr. T. S. Palmer, and as repeated by the A. O. U. Check-List (1910, p. 277). The latter authority gives it as " South Fork of Platte River, about 25 miles east of northeastern corner of Colorado, Nebraska." I myself, after studying the documents above mentioned, did not feel able to fix the locality quite so exactly, but it is, apparently, in Nebraska and on the South Platte River, not many miles east of the Nebraska-Colorado boundary line. It should be stated that on the map above cited the 154 University of California Publications in Zoology [Vol.21 only part of Lieut. Bryan's route that is clearly labelled as such is the stretch between the South Platte River in northeastern Colorado and the Solomon River, Kansas, the homeward route taken by the party later in the summer.* The type specimen of schistacea is an adult female in badly molting condition, of little value for color comparisons. The subspecies is evidently rare in Nebraska and in the adjacent portions of Colorado and Wyoming, and it is improbable that this bird was breeding at the place where it was taken. From the account in Bryan's itinerary, above cited, the headwaters of Lodgepole Creek (which empties into the South Platte not far west of the point of capture) would seem a not unlikely breeding ground for Passerella, so it may be that this bird had strayed down that stream, and then down the South Platte. In Baird's original description of Passerella schistacea (1858, p. 490), his comments were applied to a series of specimens that included two or more distinct subspecies, as now understood. As far as the description itself is concerned, both the small-type paragraph headed "Sp. Ch." and the general comments following, all laying stress upon the great size of the bill, as compared with iliaca, apply to the bird he later called megarhynchus, and not to the small-billed bird for which the name schistacea has been generally used. The definite statement is made, however, that :"The essential characters of the preceding diagnosis are based on a specimen (5118) from the head of the Platte, and collected by Lieutenant Bryan, in 1856." It was a curious slip to have made, especially by as keen an observer as Baird, for it is literally impossible that a comparison of Bryan's specimen from the Platte with examples of iliaca, could have inspired a diagnosis of the former containing the statement : ' ' Bill very thick ; the upper mandible much swollen at the base. ' ' It seems as though the descriptive paragraphs had been based upon Fort Tejon specimens, and the comment upon the Platte specimen inserted at some later time. It may, perhaps, have been put in with an idea of defining the geographic range of the species. However that may be, and however definitely every sentence in the diagnosis applies to the California bird, there is no doubt that the statement above quoted fixes the type of schistacea as being the Platte River specimen. In addition to this there is Baird's action in an appendix to the volume containing the description of schistacea, where * I am indebted to Dr. T. S. Palmer, of the United States Biological Survey, for putting me upon the track of the accounts from which the above statements are taken. 1920] Swarth: Revision of Avian Genus Passerella 155 he applies the name megarhynchus to birds from Port Tejon (1858, p. 925), and specifically mentions the Platte River specimen as the type of schistacea. Judging from the variation exhibited by the birds assembled for this study, it seems possible that the name schistacea, even as here restricted, covers a composite of two or more recognizable subspecies. That such is the case, however, is not capable of proof by the exam- ination either of winter birds alone, or of the few summer birds avail- able, which represent but a limited number of localities. The general summer range of schistacea consists of many scattered high mountain localities, more or less widely separated, and it would not be surpris- ing should some of the isolated colonies thereon prove to be as dis- tinguishably different from normal schistacea as is canescens. The rather extensive series of summer specimens from the Pine Forest Mountains, Nevada, are apparently similar in character to the type specimen of schistacea and to summer birds from the vicinity of Port Bridger, Wyoming. Schistacea, proper, is thus primarily a bird of the Great Basin. Birds of this type are of distinctly brownish coloration, as compared with canescens (both adults and juvenals), and, in the extreme of differentiation from that race, have a notably longer, more slender pointed bill. In midsummer plumage the pres- ence of the brown coloration is usually obscured by wear and fading, but molting birds from the Pine Forest Mountains show it unmistak- ably in the new-coming feathers. Most California winter taken speci- mens are evidently of this same type, slender-billed, and more brown of color than canescens. There is a small proportion of skins, from various scattered localities, that have this brown coloration carried to an extreme, nearly as much so as in some of the Unalaschcensis group, from which, however, they are readily distinguished by other characters. The capture of one of these extremely rufescent birds at Anthony, Baker County, Oregon (October 11), suggests the possibility of the summer habitat of this type lying somewhat farther to the northward, or perhaps it is an indication of intergradation between schistacea and altivagans. In support of this supposition attention may be drawn to a series of summer birds from Banff, Alberta (coll. Victoria Mem. Mus.). As compared with the Pine Forest Mountains series, the Banff birds are much more heavily marked below, and of appreciably browner coloration. One specimen in particular (no. 1175) is notably rufescent, and with a faint suggestion of streaking above. On the other hand, a specimen from a neighboring locality 156 University of California Publications in Zoology t VoL - 21 (no. 100358, Amer. Mus. Nat. Hist., Ptarmigan Lake, near Laggan, Alberta), is not appreciably different in coloration from the Nevada birds. The Alberta specimens are short-tailed, as compared with those farther south, this again being undoubtedly an approach toward altwagans. Schistacm occurs in California as a migrant and winter visitant, the specimens at hand indicating the occurrence of this subspecies over a large part of the state, though it is not a common bird at any point. The line of migration extends along the Sierra Nevada, with the main winter home in California in the San Diego region. There are many records from the Sierras extending from August to November, and a lesser number during March and April. One individual from Mari- posa County, November 29, might have been supposed to be in its winter home, but Tyler (1913, p. 85), speaking of the occurrence of this subspecies in the Fresno district (south of Mariposa County), asserts that while some of the birds remain until December, they all depart during this month, presumably for points still farther south. There is one winter bird (December 5) from the Piute Mountains, at the southern extremity of the Sierra Nevada, while from the San Diego region there are many specimens taken during December, Janu- ary, and February. These winter birds are all from points west of the divide that separates the Pacific slope from the desert, though there is no reason to doubt that the subspecies winters also on the desert slopes of these same mountains. The northernmost winter specimen at hand from the coast region is from Nordhoff, Ventura County, and there are many points of capture from that station south practically to the Mexican boundary. The subspecies has not as yet been reported from any of the islands off the coast of southern Cali- fornia, nor from any locality north of Ventura County and west of the San Joaquin and Sacramento rivers. The earliest arrival in the fall is one taken at Blue Canon, Placer County, August 31, though others have been secured far to the south of this but a few days later — at Cottonwood Lakes, 11,000 feet altitude, Inyo County, September 3, and at Lytle Creek, San Bernardino County, September 2. In Kings River Canon, in 1916, slate-colored fox sparrows were migrating in small numbers the middle of Septem- ber, the first taken on September 12, though others, believed to be of the same subspecies, had been seen on the 10th. A few days later, at Horse Corral Meadow, a high ridge just south of the Canon, the 1920] Swarth: Revision of Avian Genus Passerella 157 migration was evidently in full swing, and specimens of schistacea were shot from time to time along with several other subspecies of Passerella. The latest taken specimen in the spring is one from Ojai • P. iliaca schistacea, breeding P. iliaca schistacea, non-breeding ♦ P. iliaca canescens, breeding P. iliaca canescens, non-breeding DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY UNIVERSITY OP CALIFORNIA Fig. X. Map showing stations of occurrence of Passerella iliaca schistacea and P. i. canescens in Nevada and California, as established by specimens exam- ined by the author. The portion of southern California enclosed within the outline is believed to be the main winter habitat of schistacea in this state. Note the general avoidance of the coastal region. Valley, Ventura County, April 25. In the vicinity of Lone Pine, Inyo County, in 1912, slate-colored sparrows were seen daily about the middle of April, migrating northward, specimens being secured from April 10 to 16. 158 University of California Publications in Zoology [Vol. 21 The capture of an individual of this subspecies in midsummer (July 3, 1908) in the San Jacinto Mountains, Riverside County, Cali- fornia, remains an anomalous occurrence, difficult to understand (Grinnell and Swarth, 1913, p. 281). There are four specimens of schistaicea available from Arizona. The winter range of schistac&a is generally given as including Arizona and New Mexico (ef. A. 0. U. Check-List, 1910, p. 277; Ridgway, 1901, p. 396), but the published records of occurrence in the former state hardly suffice as a basis for the statement. I know of but four such locality records : forty miles south of Camp Apache, September 1, 1873, one specimen (Henshaw, 1875, p. 293) ; San Francisco Moun- tain, September 29, 1889, one specimen (Merriam, 1890, p. 97) ; Huachuca Mountains, November 20, 1894, one specimen (Fisher, 1904, p. 81) ; Big Sandy Creek, near Signal, Mohave County, February 6 and 9, 1880, three specimens (Stephens, 1914, p. 259). Brewster's (1882, p. 197) record from Tucson was a mistake, as pointed out by Stephens (loc. cit.). The additional specimens now at hand are two from the Sierra Ancha, Gila County, November 7 and 12, 1916 (nos. 1803, 1804, coll. G. Willett), one collected by A. Van Rossem in Bonita Canon, Chiri- cahua Mountains, January 24, 1915 (no. 3179, coll. J. E. Law), and one taken by Dr. E. A. Mearns at Payson, Yavapai [= Gila?] County, February 18, 1888 (no. 52996, Amer. Mus. Nat. Hist.). It may be only due to the failure to collect in the proper localities that the sub- species has not been found to be of regular and abundant winter occurrence in Arizona, but on the basis of the above mentioned speci- mens (all that are known to me as occurring in that state), schistacea can hardly be regarded as of much more than casual occurrence. The four Arizona specimens at hand are in appearance essentially like the breeding bird of northern Nevada, and are probably migrants from that general region. Passerella iliaca fulva Swarth "Warner Mountains Fox Sparrow Original description. — Passerella iliaca fulva Swarth, 1918, p. 162. Type specimen. — No. 14795, Calif. Mus. Vert. Zool. ; adult male ; Sugar Hill (5000 feet), Warner Mountains, Modoc County, Califor- nia; May 19, 1910; collected by "W. P. Taylor and H. C. Bryant; original number 2887. 1920] Swarth: Revision of Avian Genus Passerella 159 Range. — In summer the extreme northeastern corner of California, in the mountains of Modoc and Lassen counties; north into central Oregon; east of the Cascades. The northernmost specimen at hand is from "Warmspring, Crook County, the easternmost from Burns, Harney County. Winter home unknown. Specimens examined. — 89 (see list, pp. 201-202). Distinguishing characters. — Of the Schistacea group (see p. 89). Bill intermediate in size between schistacea and mariposae; of about the same size as in monoensis, but more slender and attenuated than the short and rather heavy bill of that subspecies (see fig. F). Color- ation more brownish than in mariposae and monoensis; about as in schistacea. Wing and tail measurements somewhat less than in mari- posae, about the same as in schistacea (see table 4). Remarks. — Size of bill will suffice to distinguish fulva from any of the other fox sparrows of the Schistacea group save monoensis. From the latter form fulva may be told by the somewhat differently shaped bill, and by difference in coloration. In worn summer plumage color differences are more or less obscured, but in freshly molted fall specimens they are readily apparent. The large series of breeding birds at hand from the Warner Mountains, California, is quite uniform in appearance throughout. Breeding birds from various points in central Oregon exhibit consider- able diversity in size and shape of bill and in relative grayness and brownness of color, though all, I believe, are best referred to fulva. This variation is in accordance with the geographical position of the several specimens, showing different degrees of intermediateness between fulva, schistacea, and mariposae. Birds from extreme southern Lake County are typical fulva. One specimen from Burns, Harney County, the easternmost point represented, in size of bill approaches schistacea. Examples from Fort Klamath and Sisters, the westernmost points of record, have larger bills, apparently tending toward mariposae. These latter specimens also are decidedly grayish in color, as compared with Warner Mountains birds, in this respect again showing an approach to mariposae. There is an abundance of breeding specimens of fulva at hand, and from enough different localities to give a fairly accurate idea of the summer habitat of this form, but there is practically no material that would serve to shed light on migration route and winter home. There is a specimen in the Mailliard collection (no. 4253), a typical example of fulva, collected at Eagle Lake, Lassen County, October 1, 1899; and one in the collection of Allan Brooks, collected by A. Van Rossem 160 University of California Publications in Zoology [Vol. 21 y^ at Sierra City, Sierra County, October 6, 1911. These would indicate a rather late tarrying near the summer habitat. An immature male (no. E. 301, coll. D. R. Dickey) taken at Lawson (= Lassen) Creek, Modoc County, August 23, is so different from midsummer birds of the same age and stage of plumage from the near-by Warner Moun- tains, that, although I include it under the name fulva, I believe that it was a migrant at the point of capture, probably from somewhere in central Oregon. It has the gray coloration and the type of bill seen in birds from that region. There is one specimen in the Emerson collection taken in the Volcan Mountains, San Diego County, February 9, 1884. This bird is unquestionably fulva. It can be very closely matched by selected specimens from the Warner Moun- tains (allowing for the different degrees of wear on the feathers), and it is absolutely the only speci- ^Wv_ ^v men of which this can be said, "~~^ ■- > that has come to light among the hundreds of winter collected southern Californian fox spar- rows examined. There has been very little bird collecting done in the mountains of San Diego County in winter, not over a dozen Passerellas from this whole county having been examined by the present writer, and it may be that future work will disclose the winter home of fulva to lie in this unexplored part of the state. There is a skin in the collection of A. B. Howell (no. 6827), taken at Bluff Lake, San Bernardino Mountains, October 11, 1918, and one in the collection of W. L. Dawson, taken in Santa Barbara County, January 11, 1913, that also appear to be fulva, though not so un- equivocally so as the Volcan Mountains specimen. The Bluff Lake specimen has the brown coloration of fulva, but not the attenuated bill characteristic of adults of this subspecies. The relatively stubby bill of this bird may be a result of immaturity, or it may, on the other hand, be an indication of intergradation, perhaps towards megarhynchus. Size and shape of bill are about intermediate between Fig. Y. Bills of two subspecies of fox sparrow, natural size. a. Passerella iliaca mariposae, adult female; no. 25691, Mus. Vert. Zool.; Yosemite Point, Yosemite Park, Cali- fornia; June 4, 1915. 6. Passerella iliaca megarhyncJms, female; no. 12402, XL S. Nat. Mus., type speeimen; Port Tejon, Kern County, California. 1S20] Swarth:' Revision of Avian Genus Passerella 161 the two subspecies. The Santa Barbara specimen is close to typical fulva. Two specimens in the "W. M. Pierce collection (nos. 1772, 1796), from Cucamonga and San Dimas canons, southern California, I have also considered as fulva, but they are far from typical of the form. If these several points in San Bernardino and Los Angeles counties represent northern and western limits of the winter habitat of fulva, as seems possible, it would not be surprising to find at these outposts non-typical birds, visitors from the borderland of the summer home of the race. Passerella iliaca megarhynchus Baird Thick-billed Fox Sparrow Original description. — P[asserella~\. megarhyndhus Baird, 1858, p. 925. [Note that I use this name, megarhynchus, as it is written by the original describer, Baird. He, too, used it in this form in conjunction with the generic term Passe- rella.} Type specimen. — No. 12402, U. S. Nat. Mus. ; female ; Fort Tejon, Kern County, California; collected by J. Xantus de Vesey; original number 1397 ; date of capture not on label. Range. — Common winter visitant to the Pacific slope of southern California. Summer home unknown. Specimens examined. — 104 (see list, pp. 202-204). Distinguishing characters. — Of the Schistacea group (see p. 89). From all other subspecies of this group except brevicauda, megar- hynchus may be distinguished by its heavy, stubby bill and relatively brownish coloration. In both these features it approaches brevicauda, but it has a less heavy bill, and it is not as brown colored as in the extreme of that form. In bill structure megarhynchus is about inter- mediate between the type of brevicauda on the one hand and fulva, of southern Oregon, on the other (see fig. Y). Remarks. -^-la. assorting the confused series of "thick-billed spar- rows" contained in the several collections available from southern California, there proved to be in each lot a rather large percentage of birds, winter visitants only, with the prominent characteristics of broad, stubby bill, and relatively brownish coloration. Through the courtesy of the authorities of the United States National Museum I have been permitted to examine the type specimen of Passerella megarhynchus Baird, and it proves to be a bird of this character. The name megarhynchus is thus shown to be properly applied (as it has been in the past) to the most common form of ' ' thick-billed spar- row" found in winter in southern California, but it can no longer 162 University of California Publications in Zoology [Vol.21 be used for the breeding bird of the Sierra Nevada. Series of summer specimens from various parts of those mountains are available for comparison, and prove to be of a different type of bird (see page 1-73). It is not possible at present to indicate the breeding range of megar- hynchus, but the relatively thorough working-out of the summer homes of stephensi, mariposae, monoensis, fulva, and brevicauda, made pos- sible by the series of specimens here assembled, narrows down the territory in which this breeding area may be expected to be found. Megarhynchus as here restricted, is, it may be remarked, perfectly distinct from any of the above mentioned subspecies, and our ignor- ance of its summer habitat is not to be explained by a mistaking of seasonal or other variation of individual birds for the subspecific variation of geographic races. Although there are only winter speci- mens of megarhynchus at hand, there is sufficient material representa- tive of the other closely related subspecies to demonstrate the reality of the indicated differences, regardless of season. Megarhynchus is most nearly like brevicauda, .in fact the material at hand indicates perfect intergradation between the two. Both are brownish colored birds, compared with mariposae, monoensis, and stephensi, and both have the same shaped bill, in different sizes. Series of breeding birds at hand show that the subspecies of the Sierras are decidedly grayish. Stephensi, mariposae, and monoensis, extending from the San Jacinto Mountains to Mount Shasta, are all of this pale type, and all have similarly shaped, rather slenderly-pointed, bills. Obviously there is no room here for the insertion of a brownish- colored, stubby-billed subspecies; also there are enough breeding stations represented along these mountain chains to show that there are no fox sparrows there of the megarhynchus type. On the other hand, the available material shows the existence of a chain of brownish-colored subspecies of the Schistacea group, extend- ing from schistacea of northern Nevada, through fulva of southern Oregon, and the "Warner Mountains, California,, and ending with brevicauda of the Yolla Bolly Mountains. In its salient characters megarhynchus apparently fits in between fulva and brevicauda; it is intermediate between the two, though appreciably nearer the latter. According to this reasoning the breeding grounds of megarhynchus would lie in the coastal mountains of extreme northwestern California, and, probably, southwestern Oregon. This section is not represented by a single specimen in the material at hand, and apparently there has been little or no bird collecting done there. As far as I know 1920] Swarth: Revision of Avian Genus Passerella 163 there are no published references to the occurrence of fox sparrows in that region, at any rate nothing to shed light on this problem. Among the winter birds from southern California here ascribed to megarhynchus, there are many that are difficult to distinguish by color from mariposae. It is worth noting that rather grayish-colored individuals are frequent in the Claremont series, while brown-colored ones predominate about Pasadena and Los Angeles. The gray-colored Claremont specimens, however, have the stubby bill characteristic of megarhynchus, and I consequently place them in that category. I do not claim to possess a perfect understanding of the exact relation- ships existing between megwrhynckus, mariposae, and fulva, which, indeed, can not be attained without much more information than I possess regarding the summer habitats of all three; I can not do more than point out the features of the winter birds examined, as correlated with places of capture. The suggestion may be made, however, that here again is evidence of fairly close restriction of winter habitat; furthermore, winter birds with the various characters above described should in the future be checked up with summer birds from points as yet unrepresented, in the more minute working out of relationships and routes of migration. Megarhynchus is a common winter visitant to the Pacific slope of southern California, but there are very few specimens from other points indicating the route traversed between this region and its breed- ing ground. There are three specimens at hand from Tower House, Shasta County, taken on March 2 and 7, and one from Sisson, Siskiyou County, September 5; whether these birds were winter visitants at the points of capture, or merely migrants, there is no way of telling. At any rate they may be taken as evidence that the breeding ground of megarhynchus lies at least as far north as this. The subspecies certainly does not breed in the immediate vicinity of either Tower House or Sisson. One bird from Lakeport, Lake County, taken on February 27 (no. 11436, Calif. Acad. Sci.)', is in large part albino, but is un- equivocally megarhynchus nevertheless. There is one specimen of megarhynchus from Nicasio, Marin County, taken January 30, the only one of. this subspecies that has come to light among the many fox sparrows examined from this county. The dates of capture of the two birds last mentioned are indicative of their being winter visitants, though the subspecies is evidently extremely rare at the points where they were taken. There is a record of megarhynchus 164 University of California Publications in Zoology [Vol. 21 as seen at Alton Junction (near Eureka), Humboldt County, in late October (Perry, 1908, p. 43), but in the absence of specimens there is no certainty as to the subspecies actually encountered. O Passerella iliaca mariposae + Passerella iliaca megarhyncJius j. Passerella iliaca fulva Fig. Z. Map showing winter occurrence of Passerella iliaca mariposae, P. i. megarhynchus, and P. i. fulva in California, as established by specimens examined by the author. The region enclosed within the dotted outline is believed to be the main winter habitat of megarhynchus. The locality stations of mariposae and fulva denote the capture of but a single specimen at each point, apparently strag- glers from the still unknown center of winter abundance. Along the line of the Sierra Nevada, among the scores of fox sparrows collected, there are but two examples of megarhynchus. One is from Blue Canon, Placer County, taken on October 19, the other from El Portal, Mariposa County, on November 28. The sub- species is evidently rare in these mountains, but may be assumed 1920] Swarth: Revision of Avian Genus Passerella 165 to winter in small numbers as far north as the latter station. The Pacific slope of southern California is evidently the winter metropolis of megarhynchus. In the foothills of this section it is in winter per- haps the commonest of the fox sparrows, certainly the most abundant of the Schistacea group. Dates of capture, in the region about Pasa- dena, range from September 15 to April 17; in the Santa Monica Mountains, from October 26 to February 22 ; from the Ojai Valley, Ventura County, there is one as late as April 27. There is but one specimen (no. 30446, Mus. Vert. Zool., Seaside, Monterey County, December 30) from a coastal point between Santa Barbara and San Francisco, but it would not be surprising should the subspecies be found regularly, migrating at least, along the coast ranges of this section. On the whole there is a notable contrast in the abundance of this subspecies in its winter home in southern California, and its scarcity as a migrant farther north in the state. An explanation of this con- dition may be found in this bird again being one that leaves its breed- ing ground abruptly and departs by a long flight to its winter home, passing over much of the intervening country. Passerella iliaca brevicauda Mailliard Yolla Bolly Fox Sparrow Original description. — Passerella iliaca brevicauda Mailliard, 1918, p. 138. Type specimen. — No. 23924, Calif. Mus. Vert. Zool. ; female adult ; one-half mile south of South Yolla Bolly Mountain, in Trinity County, California; August 7, 1913; collected by A. C. Shelton and George Stone (original number 385). Range. — There are breeding birds at hand from South Yolla Bolly Mountain, which lies at the junction of Trinity, Tehama and Mendo- cino counties, from Snow Mountain, in the northwestern corner of Colusa County, and also one juvenal from Mount Sanhedrin, in north- eastern Mendocino County. "Winter birds have been examined from the near vicinity of the coast, from Marin County, central California, and from Santa Barbara, Ventura and Los Angeles counties, includ- ing Santa Catalina Island, southern California. Specimens examined. — 73 (see list, pp. 204-205). Distinguishing characters. — Of the Schistacea group (see p. 89). Of the subspecies within this group brevicauda is superficially most nearly like stephensi, the two being noticeably alike in the enormous development of the bill. Between stephensi and brevicauda there are differences of color and proportions. Brevicauda is brownish colored, stephensi grayish ; brevicauda has the tail shorter than stephensi (see table 4), has weaker claws (see fig. AA), and a somewhat differently shaped bill (see fig. F). 166 University of California Publications in Zoology [Vol.21 Remarks. — Although brevicauda is one of the more recently differ- entiated local races of the species, it so happens that there is more information at hand regarding the seasonal distribution of this sub- species than there is regarding that of any of its near relatives. The available data include facts of decided interest, the locally restricted winter habitat being especially worthy of emphasis. In its summer home brevicauda occupies the westernmost point of any of the schistacea group, and in its winter home also it keeps just as far to the westward as possible. Other closely related subspecies occur quite as far west, it is true, but brevicauda in winter does not extend east- ward (inland) for any distance, as do the others. It hugs the near vicinity of the coast. The local restriction of the race in Marin Fig. AA. Feet of two subspecies of fox sparrow, natural size. a. Passerella iliaca stephensi, adult female; no. 2096, Mus. Vert. Zool. ; San Jacinto Mountains, California; July 10, 1908. b. Passerella iliaca 'brevicauda, adult male; no. 23922, Mus. Vert. Zool.; South Yolla Bolly Mountain, Tehama County, California; August 1, 1913. County, just north of San Francisco Bay, has been commented upon by Mailliard (1912, p. 63). Here it occurs in winter apparently with fair regularity upon certain rocky, brush-covered ridges near the coast, and nowhere else. East of San Francisco Bay, on the University of California campus, this subspecies has not yet been detected among the several kinds of fox sparrows that frequent these grounds in winter. Neither is it included in the extensive series of Passerella collected over a period of years at Hayward, a few miles to the southward, by W. Otto Emerson. South of San Francisco Bay there are no record stations for this subspecies until Santa Barbara is reached, though it can hardly be doubted that it occurs in winter at suitable spots. In this region, as elsewhere, it is probably restricted to certain limited areas. In Los Angeles County the peculiarly local distribution of the race receives special emphasis from the large number of specimens of Passerella available from many points within the county, rendering it out of all question that the predominance of certain subspecies at certain spots 1920] Suiarth: Revision of Avian Genus P'asserella 167 should be purely a coincidence (see pp. 110-111). Some of the writer's earliest experiences in bird collecting (in the vicinity of Los Angeles) were with fox sparrows, and he soon discovered that the place to hunt for "thick-billed sparrows" (quite satisfactorily thick-billed they were, top !) was in certain brushy canons between Los Angeles and the ocean, while for "Townsend's sparrows" (any of the Unalaschcensis group) it was usually necessary to make trips to the mountain tops and canons to the eastward above Pasadena. In its history thus far, brevicauda has been regularly confused with stephensi. Winter specimens of the former at one time were cited by Mailliard (1912, p. 63) as demonstrating the winter occurrence of the latter in Marin County. Oberholser (1900, p. 233) lists speci- mens of stephensi from Santa Catalina Island in April, and Bowles (1911, p. 175) from Santa Barbara in August. The specimens upon which the above records were based have all been examined by the present writer, and all prove to be examples of brevicauda. It will be noted that Oberholser {loc. tit.) comments upon the brown dorsal coloration of the Santa Catalina Island specimens, one of the dis- tinguishing features of brevicauda as compared with stephensi. While brevicauda has thus been confused with stephensi wherever the former has been collected, I nevertheless do not believe the rela- tionships of the two to be especially close. The large bill, common to both, is, of course, a most conspicuous feature, the presence of which would be apt to lead the observer astray, for without strongly con- firmatory evidence one would hardly assume the existence of two subspecies of the same species, each characterized by exceptional development of the bill. With this feature conspicuously in view, the differentiating characters of color and proportions are not such as to impress the observer so strongly, especially with but a limited number of specimens for comparison. With an adequate series, how- ever,- it can be seen that the peculiar characters of brevicauda are further accentuations of the features, not of stephensi, but of megar- hynchus. The latter is a rather brownish colored bird, compared with mariposae and stephensi, and with a broad, stubby bill. Brevi- cauda is still browner in color, and with the same shaped bill greatly enlarged (see fig. P). In typical stephensi, at least in adult birds, the bill is rather longer and more slenderly pointed, very much the shape seen in mariposae, though, of course, of greater size (see fig. E ) . Individual variation in subspecific characters is manifested in the series of brevicauda to about the same extent as in the other races 168 University of California Publications iw Zoology [Vol. 21 of Passerella. Such variation in most cases takes the form of inter- gradation toward megarhynchus, sometimes in slight degree, occasion- ally to an extent that makes the placing of the individual specimen a matter of arbitrary decision. It is noteworthy that the birds from Snow Mountain, the southernmost known breeding station of brev.i- cauda, have the bill perceptibly larger than in the Yolla Bolly birds, to the northward. The latter thus appear to show in some degree intermediateness toward megnrhynchus. In but a few cases, among winter birds, is there any occasion for confusion with stephensi. In faded and worn plumage there may sometimes be difficulty in deciding between these two subspecies, but there are no fresh-plumaged birds at hand that are at all equivocal. In addition to the normal variation just described, there is one specimen at hand (coll. W. L. Dawson, male, "Santa Barbara Co., Cal.," November 22, 1913) that shows a noteworthy trace of albinism. It has a pure white area on the forehead, extending from eye to eye, from the base of the bill backward to a maximum width of six milli- meters. On the left side it includes the lores and extends below the eye over the cheek. This white marking is of interest through its appearance of being an unusual extension of the two whitish spots between eyes and nostrils, present to a greater or less extent in prac- tically all fox sparrows. The series of breeding birds at hand sheds little light upon the dates of arrival and departure of this subspecies upon its nesting ground, for they are all midsummer specimens. The winter birds from Marin County were taken on dates ranging from September 23 to April 29. Specimens collected by Mailliard on September 23 were believed by him to have been raised possibly at the spot where they were taken (Mailliard, 1912, p. 63), a supposition that has not as yet been verified. One specimen at hand from Nicasio, Marin County (Carnegie Museum, no. 29698), bears the date of "June" 28. The label now attached to this skin is obviously not that of the original collector. In all probability a mistake was made in copying, so that this skin can not be admitted as evidence of the breeding of brevicauda at this place. Prom southern California by far the earliest fall migrant is Bowles' specimen from Little Pine Mountain, Santa Barbara County, taken August 30. Prom the Santa Monica Mountains there are speci- mens from October 4 to January 19, but there is little doubt that the birds arrive earlier and depart later than is indicated by these dates. 192 °] Swarth: Revision of Avian Genus Passer ella 169 There is one skin at hand from the Ojai Valley, Ventura County, April 18. From Santa Catalina Island a specimen collected April 19 affords the latest date for southern California. Specimens collected on the same island in mid-December attest the wintering of the bird there. Passerella iliaca canescens Swarth White Mountains Fox Sparrow Original description. — Passerella iliaca canescens Swarth, 1918, p. 163. Type specimen. — No. 28439, Calif. Mus. Vert. Zool. ; immature male (in nearly complete first. winter plumage) ; Wyman Creek at 8250 feet altitude, east slope of White Mountains, Inyo County,. California; August 15, 1917; collected by A. C. Shelton; original number 3549. Range. — In summer apparently confined to the White Mountains, in Inyo and Mono counties, California. Migration route and winter habitat unknown, save as the species is represented by a few speci- mens from Los Angeles and Santa Barbara counties and the Colorado River, California. Specimens examined. — 23 (see list, p. 205). Distinguishing charoxters.-^Oi the Schistacea group . (see p. 89). Distinguished from all other subspecies within this group, save schistacea,, by its diminutive bill ' (see fig. D). From schistacea, canescens may be differentiated by its decidedly more grayish color- ation. This is strikingly apparent when freshly molted birds of the two subspecies are compared, and it is also evident in the juvenal plumage. In worn midsummer birds these color differences naturally are obscured. Remarks. — The type series of canescens consists of twelve speci- mens from the Boreal Zone on the White Mountains, in Inyo and Mono counties, California. The series includes three adults in rather worn summer plumage, two adults undergoing the annual molt, but mostly in the new winter plumage, and six in juvenal plumage, some of them showing a few feathers of the first winter plumage. There are a few winter birds at hand from southern California that appear to belong to this subspecies, all taken at points some distance from the coast. One from Mount Wilson, November 4, 1894 (no. 34, Swarth coll.), appears to be typical canescens, extremely small-billed, of decidedly slaty coloration, and the breast spotting rather sparse and with no trace of the reddish brown peculiar to schistacea. Others from Blythe, Riverside County (on the Colorado River), San Antonio Canon, San Bernardino County, San Dimas and Palmer's canons, Los Angeles County, and Santa Barbara, though not so extreme in character, are apparently also to be referred to canescens. There is a notable difference between these birds and the brownish colored schistacea found migrating in the Sierra Nevada. 170 University of California Publications m Zoology [Vol.21 The southern California winter specimens do not suffice to indicate the winter habitat of this subspecies. Apparently, however, canescens, like the other Sierran and Great Basin forms, avoids the vicinity of the coast, and must be looked for farther inland. The points of winter capture here recorded may indicate the northern and western limits at that season. It is a noteworthy fact that in all the bird collecting carried on by the Museum of Vertebrate Zoology in Owens Valley and in the Death Valley region, not a single migrating example of canescens was secured. As these places are directly south of the White Moun- tains, they would certainly be traversed by migrating canescens, unless — as seems probable — the birds pass over these regions in long,. direct nights toward and from the winter habitat. Accumulating data points more and more convincingly toward this latter mode of travel as prevalent among many of the subspecies of Passerella. Passerella iliaca monoensis Grinnell and Storer Mono Fox Sparrow Original description. — Passerella iliaca monoensis Grinnell and Storer, 1917, p. 165. Type specimen. — No. 26930, Calif. Mus. Vert. Zool. ; male adult; Mono Lake Post Office, altitude 6500 feet, Mono County, California; May 21, 1916 ; collected by Joseph Dixon ; original no. 4644. Range. — Breeds on the east slope of the Sierra Nevada in the vicinity of Mono Lake; possibly also in the Panamint Mountains. Occurs in winter in the western foothills of the Sierra Nevada and on the Pacific slope of southern California. Specimens examined. — 38 (see list, p. 206). Distinguishing characters. — Of the Schistacea group (see p. 89). A gray-colored fox sparrow, in bill character intermediate between canescens and mariposae (see fig. E). Of the pale coloration common to the canescens-monoensis-mariposae-stephensi series. In this line of subspecies monoensis occupies a position comparable to that of fulva in the schistacea-fulva-megarhynchus-brevicauda chain. Specimens of monoensis and fulva are sometimes very much alike in bill structure, but may be distinguished by coloration, fulva being a brownish colored bird, monoensis a grayish colored one. Remarks. — There is a fox sparrow at hand, an adult male in rather worn plumage, collected at Jackass Springs, Panamint Mountains, California, June 25, 1917 (no. 28430, Mus. Vert. Zool.) that I tenta- tively refer to monoensis. The capture of this specimen in midsum- mer, a bird which was in full song and in breeding condition, is fair 1920] Swarth: Revision of Avian Genus Passerella 171 Pig. BB. Map showing summer habitat of Passerella iliaea fvXva, P. i. mwri- posae, P. i. monoensis, P. i. stephensi, and P. i. brevioauda. The symbols indicate the location of breeding stations, as established by specimens examined by the author. 172 University of California Publications in Zoology [Vol.21 evidence that the species was nesting in the mountains where it was taken, though no more of the birds were seen. If I am correct in ascribing this specimen to monoensis, it may indicate that the sub- species occupies a rather extensive summer habitat on the summits of certain of the desert mountains. However this may be, the only breeding specimens at hand that appear to be unequivocally of the race monoensis (aside from the one exception above noted) are all from a relatively limited region about Mono Lake. Specimens from Mammoth, immediately to the southward, appear to be mariposae, as are the birds still farther south on the east slope of the Sierras (see p. 175). Furthermore, between Mono Lake and the Panamint Mountains lie the White Mountains with the very different subspecies canescens. It is hardly reasonable to suppose that the latter race would be thus interposed between the known habitat of monoensis at Mono Lake, and an outlying colony such as seems to be indicated by the Panamint Mountains specimen. Altogether, the capture of the latter is an extremely puzzling occurrence. While engaged in this study of Piasserella ) i took occasion to request Mr. Donald D. McLean to collect for the Museum specimens of fox sparrows from the vicinity of his home, near Goulterville, Mariposa County, California. There were but few of any of the subspecies at hand from the Sierra Nevada foothills, -and he had told me that there were many "thick-billed sparrows" in that region during the winter months. To my surprise, of the nine birds of this type that he sent in, collected during December and January, eight proved to be mono- ensis. Coulterville is at the west base of the Sierra Nevada, almost due west of Mono Lake, thoughj of course, with the whole width of the mountain chain between the two points. There is a specimen of monoensis at hand taken in Yosemite Park, October 1, and four others from Horse Corral Meadow, in the mountains of Fresno County, secured between September 19 and 23, indicating that the subspecies spills over the crest of the Sierras on to the western slope in the migration from the summer home. If the main winter habitat should prove to be in the western foothills of the Sierra Nevada as is indi- cated by the Coulterville birds, it would imply a rather remarkable migration route — almost due east and west, rather than north and south. There are a few scattered winter birds at hand from points in Los Angeles County, including Santa Catalina and San Clemente islands, all taken west of the desert divides. Extreme dates of capture 1920] Swarth: Revision of Avian Genus Passerella 173 are September 24 (San Antonio Canon) and March 28 (San Clemente Island) ; other specimens have been taken during October, December, January, and February. Passerella iliaca mariposae Swarth Yosemite Fox Sparrow Original description. — Passerella iliaca mariposae Swarth, 1918, p. 161. Type specimen.— No. 25693, Calif. Mus. Vert. Zool. ; adult male ; ridge at 7000 feet, near Chinquapin, Yosemite Park, California; June 10, 1915 ; collected by J. Grinnell ; original number 3284. Range. — Summer visitant to the Canadian and high Transition zones in the northern and central Sierra Nevada. Breeding stations represented by specimens in hand range from central Siskiyou County (head of Little Shasta Eiver) at the north, south at least to the Yosemite region on the west slope of the Sierras, to Kearsarge Pass on the east slope. There are a few winter specimens at hand from scattered points in San Diego, Los Angeles and San Bernardino counties, but too few to indicate accurately the entire winter habitat or migration routes. Specimens eooamined. — 206 (see list, pp. 206-209). Distinguishing characters. — Of the Sehistacea group (see p. 89). With the gray coloration common to the canescens^monoensis-mariposae- stephensi series, and not to be distinguished in color and markings from the other members of this chain of subspecies. In bill structure mariposae lies between monoensis and stephe-nsi (see fig. E ) . In gen- eral size and length of tail in this line there is progressive increase from canescens to stephensi, and in these particulars as in size of bill, mariposae. occupies its appropriate intermediate position (see table 4). Remarks. — While there is at hand an abundance of summer speci- mens of mariposae, there is a notable dearth of winter collected bjrds. These facts, coupled with the reverse. condition. in megarhynchus, of which I have seen winter specimens. only, may be used by some as arguments proving the two to be identical, and only suffered to masquerade under different names through a misunderstanding of seasonal differences. In contravention to this idea it may be pointed out that there are many specimens in the available series of mariposae taken at summer stations just, prior to the birds' departure therefrom in the fall, and after they had assumed the winter plumage, these specimens sufficing to show the color differences distinguishing mari- posae from megarhynchus. Shape of bill is an excellent differentiating character between adults at any season. There are a few examples of mariposae at hand taken in winter at various points in southern California, but they are so few in 174 University of California Publications in Zoology [Vol. 21 number compared with the abundance of certain other subspecies associated with them that there seems no doubt that the places of capture of these scattered individuals do not represent the main winter home of the race. Then, too, some of these winter birds are not absolutely typical of mariposae, though most nearly resembling that form. Altogether, it is evident that the winter metropolis of mariposae is still to be discovered. In this connection it may be pointed out that we are similarly ignorant of the winter homes of the other Sierran fox sparrows, stephensi and monoensis, and of their neighboring relatives fulva and canescens. In each case there are scattering winter birds and migrants at hand, but nothing to show where the bulk of the populations winter. It is not to be supposed that all five of these subspecies necessarily depart to the same place, but it seems possible that their modes of departure and arrival are very similar, and that the explanation of the puzzle in each case lies in the existence of a relatively restricted winter habitat, and one that has not been thoroughly scrutinized for these birds at that season. It may be, of course, that mariposae winters in some parts of the Sierran foothills, but there is little evidence of this either in the avail- able specimens, or in published literature. Belding (1890, p. 170), speaking of the Sierran bird (under the name megarhynchus) , says: "Not at Alta and Colfax November 17-21, nor at Red Bluff in warm winter of 1884-85. I never see it in the lower foothills of Calaveras County in winter, though P. imalaschcensis is common there at that time." That the detailed mapping of summer ranges of the forms of Passerella breeding in the Sierra Nevada will yield interesting and valuable results is indicated by the general facts already gathered. The habitats of stephensi and mariposae are each known to be inter- rupted along certain lines where thorough exploration has been prose- cuted, and as there are other parts of the Sierras presenting similar conditions it is evident that at these places also there must be breaks in distribution, and that each of these subspecies in summer is scat- tered into a number of colonies of varying extent. The Kings River Canon section of the Sierra Nevada may be cited in its bearing on the fox sparrow problem. This canon runs far back into the mountains, a broad, gently sloping valley, Transition zone clear to its uppermost confines and therefore not inhabited in summer by Passerella. The surrounding walls are high and steep, and drop from plateaus, zonally of high Transition, Canadian and higher. In 1920] Swarth: Revision of Avian Genus Passerella 175 this elevated region south of the Kings River Canon the present writer found stephensi in abundance in the late summer, evidently on its breeding ground. This is assumed to be the northernmost point reached by this subspecies in the Sierra Nevada, though no collecting has as yet been done on the high country immediately north of the Kings River Canon. This gorge evidently acts as a barrier to the continuous distribution of Passerella in this direction. Prom the head of the Canon (Transition Zone) steep slopes arise, carrying an extremely limited belt of Canadian Zone (where we saw no Passerella) and leading abruptly to the extensive strip of Hudsonian and Alpine- Arctic along the crest of the Sierras. These latter belts are not occupied by Passerella, hence they act as a barrier between colonies on the east and west slopes of the mountains. On the east side, in Kearsarge Pass, mariposae is abundant in summer, the southernmost point at which the subspecies has been found breeding. This station on the east side of the mountains is directly opposite that point on the west side where stephensi occurs at its northern limit, the two being separated by the intervening strip of Alpine-Arctic and Hudsonian. Thus mariposae occurs consider- ably farther south on the east slope of the Sierra Nevada than on the west slope. There are one or two specimens of Passerella at hand from Mammoth, Mono County, on the east slope and close to the habitat of monoensis. These appear to be referable to mariposae, and indicate the probable extent of territory occupied by this subspecies on the east slope of the Sierras, a territory cut off from the western colony for its entire length by the Alpine-Arctic and Hudsonian zone of the summit. It is difficult to understand why this isolated division should not be of the monoensis type. There is no obvious barrier between it and that subspecies, and reasoning d priori one would expect the features of the latter race to be developed over this entire eastern exposure where Passerella oecurs. Such is most emphatically not the case at the southern limits of mariposae, however, Kearsarge Pass specimens being unmistakably similar to mariposae of the Yosemite region, and impossible to confuse with their nearer neighbors, stephensi on the one hand and monoensis on the other. Farther north, in Eldorado County, there is what appears to be intergradation toward monoensis. Birds from the western part of this county are typical mariposae, but in the series from the immediate vicinity of Lake Tahoe and Mt. Tallac there are some individuals quite as small-billed as selected examples of monoeiisis. So, too, as regards 176 University of California Publications in Zoology [Vol. 21 three skins from Douglas County, Nevada, southeast of Lake Tahoe. Two of these are about as small-billed as monoensis, while the third is an average mariposae. Monoensis is not a strongly marked form, and its characters, such as they are, obviously show intermediateness between mariposae and canescens, between which races monoensis is situated geographically. This fact may have some bearing on the relative restriction of the region in which typical monoensis occurs. Doubtless also, a careful working out in greater detail than has yet been done of the breeding ranges of the several subspecies concerned will go far toward explain- ing facts the meaning of which now seems obscure. There is little data at hand to indicate precise dates of arrival and departure of the subspecies mariposae at the summer and winter homes. In Kearsarge Pass, on May 27, male fox sparrows were in full song, evidently settled at their nesting grounds, and on May 29 a set of three well incubated eggs was taken. There is a female mari- posae at hand (no. 29090, Mus. Vert. Zool.), which had laid part of its set, taken May 15 at the head of Little Shasta River, Siskiyou County, the northernmost point of record for the subspecies. Of fall dates in the Sierra Nevada, there is a large series of mariposae at hand taken at Cisco and Blue Canon, Placer County, from August to October, the latest from Blue Canon, October 14. There is one bird from Tuolumne River, Yosemite Park, October 1. There is one speci- men at hand taken at Yermo, on the Mohave Desert, May 28. In view of the above nesting dates this last must be regarded as an extremely belated migrant. This bird is a typical example of mariposae; the point of capture, of course, is many miles from the nearest point in this subspecies' summer habitat. Two specimens from Mt. "Wilson, Los Angeles County, taken April 1 and 4, respectively, may indicate more nearly the usual limit of its stay in the spring in the south. Passerella iliaca stephensi Anthony Stephens Pox Sparrow Original description. — Passerella iliaca stephensi Anthony, 1895, p. 348. Type specimen. — No. 15387, Carnegie Institute Museum; male adult; San Jacinto Mountains, California, at 8000 feet; July 14, 3895 (seep. 181). Range. — Summer visitant to upper Transition and Canadian zones in the southern Sierra Nevada and the Sierras of southern California. There are several isolated tracts occupied by this subspecies, namely : in the Sierra Nevada, from Hume and Horse Corral Meadow, Fresno 1920] Swarth: Revision of Avian Genus Fasserella 177 County, south through Tulare County ; on the summit of Mount Pmos, Ventura County ; in the San Gabriel Mountains, Los Angeles County; the San Bernardino Mountains, San Bernardino County ; and the San Jacinto Mountains, Riverside County. The winter home is unknown. Specimens examined. — 186 (see list, pp. 209-211). a. Fasserella iliaca stephensi Fasserella iliaca brevicaua'a n Passerella iliaca monoensis DISTRIBUTION MAP MUSEUM OF VERTEBRATE ZOOLOGY UNIVtRSITY OP CAI.ir Pig. CO. Map showing 'winter occurrence of Passerella iliaca stephensi, P. i. brevioauda, and P. i. monoensis in California, as established by specimens exam- ined by the author. Note the restriction of brevicauaa to the near vicinity of the coast. The locality stations for stephensi denote the capture of but four specimens, apparently stragglers from the still unknown main winter habitat. Distinguishing characters. — Of the Schistacea group (see p. 89). Most nearly like mariposae, from which it is distinguished by slightly greater general size and notably large bill (see fig. E). The enor- mous bill distinguishes stephensi from all other fox sparrows except breviccmda. Prom the latter it is distinguished by rather longer, 178 University of California Publications in Zoology [Vol.21 more slenderly pointed bill (compared with the blunt, stubby bill of brevicauda) (see fig. F), by more grayish coloration, and by longer tail (see table 4). Remarks. — In all its characters stephensi represents the cul- mination of the line of development beginning with oanescens and continuing through monoensis and mariposae. Gray coloration is common to all these forms, as compared with the more brownish hue of the other subspecies of the Schistacea group, and there is no appar- ent difference in this respect between these several southern subspecies. There is, however, gradual increase in general size from canescens along the line indicated to stephensi, with notable increase in size of Fig. DD. Bills of two subspecies of the Schistacea group (races that are distinguished mainly by differences of size in this member), showing how variation of adult and immature may be mistaken for variation differentiating subspecies. a and 6 are, respectively, adult and undeveloped immature of stephensi, c and d, of mariposae; o is more nearly like c than like a. a. Passerella iliaca stephensi, adult male; no. 20505, Mus. Vert. Zool. ; Taylor Meadow, Sierra Nevada, Tulare County, California; July 19, 1911. o. Passerella iliaca stephensi, immature male; no. 27388, Mus. Vert. Zool.; Hume, Fresno County, California; August 18, 1916. c. Passerella iliaca mariposae, adult male; no. 25693. Mus. Vert. Zool.; Chin- quapin, Yosemite National Park, California; June 10, 1915. A. Passerella iliaca mariposae, immature male; no. 23195, Mus. Vert. Zool.; Blue Canon, Placer County, California; August 28, 1912. bill, and there is also, to a lesser extent, some increase in proportionate length of tail. Stephensi has a strikingly disconnected range, occurring in a number of widely separated high mountain localities. In the southern Sierra Nevada the subspecies is found in high Transition and Can- adian from southern Fresno County (at points just south of Kings Eiver), south through Tulare County. The localities of occurrence are all on the west slope of the range. Fox sparrows at this latitude do not ascend into the Hudsonian and Alpine-Arctic zones, which constitute so large a part of the summit of the Sierra Nevada. This strip along the divide is an effective barrier to the general dispersal of these birds, and in the case of stephensi it has apparently restricted the subspecies to the western slope. 1920] Swarth: Revision of Avian Genus Passerella 179 South of the Sierra Nevada the next breeding station of stephensi is a limited area on the summit of Mount Pinos, some eighty miles southwest of the southernmost station in the Sierras. Then, still farther south, there are widely disconnected areas on the San Gabriel Mountains, the San Bernardino Mountains, and the San Jacirito Mountains, separated from each other by many miles of lower zones. In the San Jacinto Mountains points of occurrence are mostly from 8000 to 9000 feet, though a few birds were seen in one part of the mountains as low as 6000 feet. In the San Bernardino Mountains the subspecies occurs at about the same elevation, mostly from 7400 to 9000 feet; in the San Gabriel Mountains it is known only from the higher slopes of Mount Waterman from 7000 feet upward to the summit (7752 feet altitude) ; and on Mount Pinos (altitude 8826 feet) from 8000 feet upward. In the Sierra Nevada the vertical range is from 7000 to 10,000 feet. It is a notable fact that, despite the isolation of these several colonies and the demonstrated plasticity of the species throughout its range, there are no discernible differences between specimens of stephensi from the several regions occupied. Considering the varia- tion existing between the many closely connected subspecies, and the way in which each represents a step from one extreme to another, it might be expected that such widely parted groups as those into which stephensi is divided would illustrate separate and distinguishable stages as they advanced farther and farther from the starting point, but such is not the case. In the series at hand I am unable to appre- ciate any tangible features serving to distinguish birds even from the extremes of the subspecies ' range. The most that can be said regard- ing such differences is that in the series from Hume, Fresno County (the northernmost point at which stephensi has been collected), there are certain small-billed individuals that may be taken to illustrate intergradation with the nearby form mariposae, but in the same series there are other specimens with quite as large bills as the maximum from any other place. In other words, while stephensi is the culmin- ation of a series of steps from canescens and through monoensis and mariposae, the maximum of change is reached as soon as the habitat of the subspecies is entered at the north, where mariposae and stephensi come together. There is no further accentuation of character in the extensive but disconnected habitat of stephensi as it extends southward. Stephens fox sparrow is a summer visitant only upon its breeding grounds, but there are neither specimens nor notes available indicating 180 University of California Publications in Zoology [Vol.21 dates of arrival in its summer home. All the breeding birds at hand were taken long after they had settled down to the duties of nesting. As regards the time of departure at the summer's end, this can be fairly definitely fixed. In the San Jacinto Mountains I revisited in September, 1914, certain sections where the Stephens sparrow had been found in abundance in the summer of 1908. On September 9 one was secured near Strawberry Valley, somewhat below the breeding level in. this range. On September 11, after most careful search, four were seen in Tahquitz Valley, in the heart of the nesting ground, and on September 15, one other was noted, the last observed, though I remained in the region some days longer. In the San Bernardino Mountains Stephens sparrows "were seen in small numbers at Bluff Lake as late as September 3" (Grinnell, 1908, p. 99). At Hume, Fresno County (altitude 5300 feet), Stephens sparrows were abundant during the middle of August; 1916, evidently on their nesting grounds from the number in juvenal plumage, and numerous specimens were collected from August 17 to 23. On Sep- tember 25 and 26, this point was revisited, and while fox sparrows were again numerous they proved to be practically all examples of more northern subspecies on their southward migration. One stephensi was collected, on September 26. The main winter home of the Stephens fox sparrow is unknown, and it seems evident that when the birds leave their summer habitat it is an abrupt departure, and possibly to a relatively great distance. They do not descend the mountains to the foothills and adjacent valleys, which are occupied in winter by other subspecies, or they would surely have been detected in such places. Stephensi has been reported in winter from Santa Catalina Island (Oberholser, 1900, p. 233), but the specimens upon which the statement was based prove to be of another subspecies, orevioauSd (see p. 167). In partial contradiction to my general statement regarding the winter habitat, I must admit to having before me four fox sparrows, taken in midwinter and to all appearances unequivocal examples of stephensi. The data for these skins are as follows: No. 8106, coll. F. S. Daggett, male, Franklin Canon, Santa Monica Mountains, Los Angeles County, California, November 24, 1915 ; no. 1424, coll. J. E. Law, female, Hollywood, Los Angeles County, California, February 22, 1912; no. 1762, coll. "W. M. Pierce, male, Palmer's Canon, Los Angeles County, California, January 11, 1919; coll. A. Brooks, male, Santa Barbara, California, January 11, 1913. As far as they go these 192 °] Swarth: Revision of Avian Genus Passerella 181 birds must be conceded to be winter collected examples of stephensi, but it must be recalled that they are but four specimens out of several hundred fox sparrows collected in the same general region, and the conclusion seems unavoidable that they represent unusual occurrences of some sort. When the normal winter home of stephensi is discovered that place undoubtedly will be found to harbor the birds in numbers. Through the courtesy of the authorities of the Carnegie Museum, I have been permitted to examine the type specimen of stephensi, collected by A. W. Anthony, and labelled as taken in the San Jacinto Mountains at 8000 feet altitude. This elevation in these mountains can mean, in this connection, but one place, Tahquitz Valley. In this valley the Stephens fox sparrow is an abundant summer visitant, while the opposite slope of San Jacinto Peak at that altitude is far too precipitous to afford suitable surroundings for the bird. Tahquitz Valley may thus be taken as the type locality of stephensi, fixing it a little more exactly than the general term San Jacinto Mountains. 182 University of California Publications in Zoology E VoL - 21 a t) S 3 13 ,£! 5 — 0.13 o .3 •-* 8 'm oo ^ i_ — ' ■<• a .J 'S3° 18* S « a ffl v e.S „ K >■ w ■ fe ■■ 2 S m i 1-3" ^ ss O CT> 5 ■a' °° is o o •Si" & -a . f£r ~ JB ^ ft, oi "3 °5 2? U5 -. > 00 I'S 9 a « is™ II 1» :§£« 0. J3 S *6 ** +3 od g od « « JS s s fe rt I ..I w a oo m oo ww. » . a ^ co -3 co ft- w s ^ 1920] Swarth: Revision of Avian Genus. Passerella 183 *5 st 3 -a I T I T T Cl CI O o ea ■* M co s S si ss i ■s K S A A A A a ■a s •a 8 5 2 S ■a J3 iSo S o o saJoa « • ° d •5 . w ^ ja " *o„ I?* a y • § o « J -S £ ~ O « ,2 •S S -2 Is o u .. « ^ -° * « as rl k a? i a. B » I -S « - S - ■so -;=!, |o" iaca brevicau a Bolly Mt., 12.0 on S3 1 r •1|A O N t~ OS It; co --. 3 ^ — —' ^ i A ^ A "3 X J « ih sis Ja 4 a© ^S'S 1 I §^ s ^ a .2 US DD ^ a -S™ ■^ * -h o - o i oi 0, -[ ■S .2 !2 U io on 2 -H -H S w * w I -5 A I5A a o> O o • oi O s °. MS s A *- • S §|A •3 en 3 &S CM 2 i3n < £ m A 1A A A °2 A AAA AAA 184 University of California Publications in Zoology [Vol. 21 LIST OF SPECIMENS EXAMINED Passerella iliaca iliaca Collection No. Sex Age Locality Date Mus. Vert. Zool. 4586 C? ad. Andreavsky, Yukon River, Alaska June 3, 1895 4587 d 1 ad. St. Michael, Alaska May 20, 1896 6087 9 Taunton, Mass. April 13, 1887 6088 rf 1 Eastford, Conn. April 12, 1876 11283 ..:... San Geronimo, Marin Co., Calif. Deo. 31 1901 • 7256 c? San Geronimo, Marin Co., Calif. Feb. 17 1879 rf 1 Jan. 17 1897 9 Los Angeles, Los Angeles Co., Calif. Jan. 17 1897 9 & Millard's Canon, Los Angeles Co., Calif. Santa Monica Mts., Los Angeles Co., Calif. Feb. 11 Dec. 6 1901 H. S. Swarth 771 1897 937 9 Los Angeles, Los Angeles Co., Calif. Nov. 15 1898 1023 9 Santa Monica Mts., Los AngeleB Co., Calif. Jan. 13 1899 G. Willett 1669 a* Stony Ford, Colusa Co., Calif. Mar. 5 1915 2043 C? ' San Antonio Canon, Los Angeles Co., Calif. Jan. 5 1917 Passerella iliaca sinuosa Collection No. Sex Age Locality Date Mus. Vert. Zool. 1535-1550 llcTad. 59 ad. Hinchinbrook Id., Prince William Sd., Alaska June 25 to July 2, 1908 1551-1565 2cTad. 59 ad. 5c?juv. 39juv. La Touche Id., Prince William Sd., Alaska July 19 to Aug. 8, 1908 1566-1590 9cTad. 10 9 ad. I9im. 4cTjuv. i9iuv. .Montague Id., Prince William Sd., Alaska July 6 to July 31, 1908 1591 C? ad. Hawkin's Id., Prince William Sd., Alaska June 22, 1908 1592 r? ad. A. B. Howell 36 d 751 d 1984 9 4961 d" 4964 d 4966 9 6568 cT 6569 ef 1 6570 c? S. G. Jewett 134 694 9 695 d 1 1045 cf 1116 cT 1490 revicauda, but there are such intergrades connecting each one with still another subspecies, sehistacea. [218] UNIV. CALIF, PUBL. ZOOL. VOL. 21 [ SWARTH ] PLATE 4 PLATE 5 Pig. 2. Meadow on the mountain side at Port Snettisham, southeastern Alaska. The Townsend fox sparrow {Passerella iliaoa townsendi) was found here in fair abundance, mostly in the densest underbrush. Photograph taken August 25, 1909. Pig. 3. Tahquitz Valley (8000 feet altitude), San Jacinto Mountains, Califor- nia. The trees in the middle distance are mostly Jeffrey pine and silver fir, those on the higher slopes in the background, lodgepole pine and limber pine. The brush covering the hillside in the foreground is chinquapin. Stephens fox sparrow is an abundant summer visitant to this valley, frequenting for the most part the densely matted ehaparral formed by the ehiuquapin. Tahquitz Valley is the type locality of Passerella iliaca stephensi Anthony. Photograph taken September 11, 1914. [220] UNIV. CALIF. PUBL. ZOOL, VOL. 21 I SWARTH ] PLATE 5 Fig. 2 Fig. 3 PLATE 6 Pig. 4. Fort Tejon, Kern County, California. This locality is in the Upper Sonoran zone, the buildings set amid hills whose slopes are covered with dense chaparral. In the more level places there are many live oaks, and along the streams, sycamores and thickets of willows. These are the characteristic sur- roundings sought by the more northern subspecies of fox spaxrow for their winter home in southern California. Fort Tejon is the type locality of Passerella megar- hynohms Baird, the type specimen, a winter bird, having been taken there some time prior to 1858. Fort Tejon is the type locality for other species of birds also, as well as for certain mammals and reptiles, hence is a point of special interest as regards the early history of zoology in California. The old fort, . itself, is the adobe building at the left, under the oak trees. Photograph taken July 24, 1918. [222] UNIV, CALIF. PUBL. ZOOL, VOL. 21 [ SWARTH ] PLATE 6 PLATE 7 Fig. 5. Little Onion Valley (altitude 7500 feet), Sierra Nevada, on the Inyo County side, near Kearsarge Pass, California. The scattered conifers are mostly silver fir; the small trees along the stream are willow, alder and birch; the chaparral on the hillsides is mostly manzanita and sage. The Yosemite fox sparrow (Passerella iliaoa mariposae) is an abundant summer visitant to this place. A nest was found in a manzanita bush shown in the foreground of this picture. Photograph taken June 3, 1912. [224] UNIV, CALIF, PUBL. ZOOL, VOL, 21 [ SWARTH l PLATE 7