340 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1919. 
the migratory habit. In this connection it is interesting to quote 
Whitlock (Migration of Birds, final paragraph) : 
Every species, nay, every little clan of birds has its own migratory history, 
resembling as a whole the story of the common flight, but on the other hand 
differing in many points in its minor details. 
Before, then, the migration of any species can be studied as a 
whole, a detailed knowledge of its geographical distribution will be 
necessary, and in grappling with this question we are at once con- 
fronted with the question of subspecies or geographical representa- 
tives. 
A subspecies is an incipient species and is evolved ab initio from 
exactly the same causes as a species. The causes of variation in 
species or subspecies may be roughly summarized as follows: We 
may attribute variation in size to the quantity or quality of food, 
variation in structure to some essential habit developed in the daily 
search for food (it is hoped to show at some future date that length 
of wing is not dependent on length of migration, but on daily habit), 
special decorative development to courtships necessitating nuptial 
display, the thickness or extent of the feathered regions to climate, 
and variation in color to climate or local surroundings or food. A 
high temperature, a dry atmosphere, and a bright light seem to pro- 
diice that bleached effect usual in desert forms. A temperate climate, 
moist air, and a dull sky tend to dark plumage. Alpine and Arctic 
forms display more white than is noticeable in the same bird from 
farther south or from lower altitudes ; though we see in the case of 
Corims comix capellanus the brilliant glare of the Persian Gulf hav- 
ing the same effect on plumage as the glare from the Arctic snows has 
on many northern forms. 
It is curious that it is the influence of the breeding quarters which 
causes differentiation, the winter quarters and regions traversed on 
migration having little effect on coloration or structure. Wide-rang- 
ing and common species show the most variation, so long as their 
breeding area is not restricted, as in some of the Polar breeding 
species. It therefore seems likely that it is the nursery which tends 
to differentiation. This is most remarkable in such birds as Cuculus 
canorus and Micropus apus, whose nursery life scarcely extends to a 
quarter of the whole year, and among which several well-marked 
geographical forms exist which in some cases share a common winter 
quarters. 
But it seems by no means proved that the breeding quarters of a 
species is necessarily its real home, though it is undeniable that the 
present breeding quarters of a species produces the homing influence 
on spring migration. Seebohm (Geog. Dist. of the Charadriidse) has 
already pointed out that it is possible that the present winter quar- 
