92 University of Calif ornia Puhlicaiions in Zoology [Vol. 21 



species are separated by physical conditions, the range of some one 

 subspecies is, in many instances, again broken up into several discon- 

 nected areas, these separated as widely as, or even more widely than, 

 those of perfectly distinct forms. Variation is regular and definite ; 

 it follows certain lines of development in certain directions; in every 

 case where boundaries between subspecies exist and an obvious change 

 in the characters of the birds takes place, there is a gap in distribu- 

 tion. Separation into subspecies, however, does not occur in every 

 case where there is such a gap in distribution. 



To summarize : We find in the fox sparrows of the northwest coast, 

 a definite line of variation, with, at intervals, the abrupt accentuation 

 of certain features serving to typify the several subspecies, this con- 

 dition occurring over a region where circumstances permit the con- 

 tinuous distribution of these birds. In the mountains of western 

 North America we find, in the fox sparrows of the Schistacea group, 

 exactly similar lines of variation, again abruptly accentuated at 

 intervals into what we designate as subspecies, but occurring over 

 a region where the birds exist in more or less widely separated 

 aggregations. 



It would seem that the factors governing variation in this group 

 of birds acted to some extent independently^ of the circumstances by 

 which colonies are isolated at various points within the general range. 

 Information is lacking as to the exact nature of the boundaries between 

 several of the forms of the Schistacea group, but we do know of wide 

 gaps in distribution within the range of several of these subspecies. 

 It is a remarkable fact, and one that is worth dwelling upon, that 

 where there is a continuous trend of variation there is little change 

 within the geographical limits of each form. To take as a concrete 

 example such a trend as extends from northeast to southwest in 

 the canescens-nw-noensis-mariposae-stephensi line of subspecies, once 

 stephe7isi is reached, after a brief step from mariposae, there is no 

 further change. Stephensi from the northern extreme of its range is 

 practically identical with stephensi from the southern extreme, despite 

 the disconnected nature of its habitat, and the enormous gaps between 

 the several colonies. This, to a greater or less extent, is what appears 

 to occur in each subspecies. There are narrow areas of intergradation 

 with much more extensive areas of stability of characters, each change, 

 however, from one form to another being part of a general trend. 



With the above points in mind we can now turn to a general 

 consideration of the genus. In the Iliaca group are iliaca and alti- 



