66 



SCIENCE 



[N. S. Vol. XLI. No. 1045 



of the coast region as far north as Alaska. 

 This valuable material has been collected 

 largely with a view to the study of geographic 

 variation. Throughout that time, and indeed 

 for a much longer period, the author of the 

 present report has been active in describing 

 and subdividing species of Pacific Coast birds 

 and mammals. 



A considerable proportion of the vertebrates 

 of this region are represented in different 

 parts of their range by different local races, 

 for which the term " subspecies " has gained 

 general acceptance. Indeed, the process of 

 " splitting " in these groups has been carried 

 to such lengths that a large majority of the 

 birds and about three fourths of the mammals 

 listed in the paper here considered are desig- 

 nated by trinomials. Outside of taxonomic 

 circles the feeling is sometimes expressed that 

 these trinomials stand for more or less ficti- 

 tious entities, the product of minds in which 

 the passion for detecting differences has al- 

 most reached the stage of paranoia. In quite 

 a different spirit is Bateson's recent advice 

 to the systematists to " subdivide their mate- 

 rial into as many species as they can induce 

 any responsible society or journal to publish," 

 since " the collective species is a mere abstrac- 

 tion, convenient indeed for librarians and 

 beginners, but an insidious misrepresentation 

 of natural truth." ^ Whether these ultimate 

 subdivisions are termed species or subspecies 

 is, of course, a matter of secondary importance. 

 The main thing is that they should be de- 

 scribed and named. 



It is probably no mere accident that several 

 of the leading exponents of the " isolation " 

 theory of specific differentiation have done 

 much of their field work on the Pacific Coast 

 of North America. Here the subdivision of 

 the earth's surface by means of natural barriers 

 is carried to an extreme probably not else- 

 where found within the limits of the United 

 States. It is true that in many cases the areas 

 thus marked off differ very greatly in their 

 climatic conditions, as witness the abrupt 

 change which we encounter in crossing the 



1 Bateson, ' ' Problems of Genetics, ' ' Yale Uni- 

 versity Press, 1913. 



mountains from the Mojave Desert to the 

 orange belt of southern California. Any 

 specific differences which are met with on the 

 opposite sides of such a barrier might be 

 attributed to environmental differences acting 

 directly or indirectly. Such cases do not, of 

 course, prove anything as to the efficacy of 

 isolation per se in giving rise to divergent 

 descent lines. 



In the lower Colorado River, however, Grin- 

 nell finds what he regards as a critical in- 

 stance. Here is a river, bordered on each side 

 by a vast expanse of desert, uniform in its 

 character for great distances, whether to the 

 right or left. Considered as a physical envi- 

 ronment, the California side of the river is 

 identical with the Arizona side. Tet of the 23 

 species of rodents collected in the valley of 

 the Colorado by the museum expedition of 

 1910 Grinnell and his party found 8 which 

 were absolutely restricted to one or the other 

 side of the river. These last were all strictly 

 desert-dwelling forms which probably never 

 visit the water's edge. On the other hand, 

 the inhabitants of the lower reaches of the 

 river bottom were found to be in every in- 

 stance common to the two banks. 



The case upon which the greatest stress is 

 laid is that of two species of ground-squirrel, 

 belonging to the genus Ammospermophilus. 

 Twenty-four specimens of A. harrisi harrisi 

 were captured at scattered points on the Ari- 

 zona side of the river, while seventeen speci- 

 mens of the closely related A. leucurus leucurus 

 were taken on the California side, the two 

 occupying the same " ecologic niche " in their 

 respective territories. In no case was a single 

 individual found on the " wrong " side of the 

 river. These two species were seen at points 

 only about 850 feet apart in a direct line. 

 Commenting on this case, Grinnell remarks: 



The sharp separation of the ranges of [such] 

 nearly related vertebrates by a barrier of such 

 narrow width is, to the best of the writer's 

 knowledge, not known elsewhere in North Amer- 

 ica. 



The author makes the assumption common 

 to both Lamarckiaiis and Natural Selectionists 

 that morphological differences must, in some 



