102 University of California Publicatio7is in Zoology [Vol. 21 



its origin to one set of causes, and its subsequent development to quite 

 different factors, which produced the several subspecies now seen in 

 each group. 



There is food for thought in the subspecific variation seen in 

 Passerella, considered in relation to the complicated migratory move- 

 ments performed by the several subspecies. It is hard to reconcile 

 the observed conditions with any of the proposed theories regarding 

 the origin of bird migration. I am not sure that there is another 

 instance known among birds duplicating the migrations of the several 

 subspecies of the Unalaschcensis group of fox sparrows. There are 

 cases where one migratory subspecies passes completely over the range 

 of another, closely related but more sedentary in its habits, but the 

 yearly shifting of the Unalaschcensis subspecies is much more com- 

 plicated, as herein shown (see map, fig. H). 



Assuming that the north and south migrations of birds are habits 

 resulting from an originally short, involuntary retreat as advancing 

 Arctic conditions first produced seasonal changes where the semi- 

 tropical climate of pre-Glacial times had previously obtained, this 

 followed by a subsequent return as a milder climate again prevailed — 

 a theory commonly advanced in explanation of migration — it is still 

 hard to understand how the results we now see could have followed 

 in such case. It is conceivable, however, that the diverse migrations 

 of the different subspecies now seen on the Pacific coast developed 

 after the establishment of an originally simple north and south 

 movement similar to that now followed by iliaca. Such an hypothesis 

 would imply a change of topography and climate in the region occu- 

 pied, to which the southernmost birds gradually adapted themselves. 

 Whether or not the geological histor}^ of the Pacific coast of North 

 America lends support to this idea I do not know. The supposition 

 is of further interest in its bearing upon the question discussed else- 

 where in this paper (see pp. 93-96) as to the possibility of subspecific 

 characters being fixed by conditions in the winter as well as in the 

 summer habitat, for this would imply the establishment of the 

 migratory habit prior to the appearance of the physical characteristics 

 now seen in the different forms. 



As regards the subspecies of the Schistacea group, the assumption 

 is that these birds occupy their present high mountain summer habitats 

 as refuges to which they retreated with the advent of a warmer epoch, 

 prior to which they had occupied the adjacent valleys. This being 

 the case, why should not the winter migration be simply back to these 



