PELICAN ISLAND 193 
usual colony had taken possession of the island.” 
(Auk, xiv, 1897, p. 284.) 
The migration of the island-nesting Terns in the 
tropics is apparently no less regular. Scott states 
that the Noddy arrived in the Tortugas “on April 
20th in large numbers, but remained only two days; 
after inspecting their breeding grounds, all departed 
to return about a week later in greatly increased 
numbers, when breeding was almost at once com- 
menced.” (Auk, vii, 1890, p. 306.) 
These insular colonies, however, not only throw 
much light on certain existing phases of bird migra- 
tion, but they also furnish us with a clew to the 
origin of migration itself. This is especially true of 
those species whose lives are passed in the tropics 
or subtropics, and which we are accustomed to class 
“permanent residents,’ but 
rf 
as nonmigratory or as 
which are as regularly migratory, in the real mean- 
ing of the word, as if they summered within the 
arctic circle and wintered south of the equator. 
Their movements are apparently in no way in- 
fluenced by climate nor, at this season, are they 
governed by the food supply, but prompted solely 
by the annually recurring physiological change 
which fits both sexes for reproduction, they repair 
to a certain islet, perhaps in the heart of their range, 
with the one object of finding a suitable nesting site 
in which their eggs may be laid and young reared in 
safety ; and this object accomplished, they desert the 
locality, where they may be unknown until the fol- 
lowing spring. 
Divested, therefore, of the complications which 
ensue when in studying the migration of birds the 
14 
