%6 CALIFORNIA ACADEMY OF SCIENCES  [Proc. 47H Serr. 
August, and sometimes a few days sooner; and every straggler 
invariably withdraws by the 20th: while their congeners, all 
of them, stay till the beginning of October, many of them all 
through that month, and some occasionally to the beginning 
of November. This early retreat is mysterious and wonderful, 
since that time is often the sweetest season in the year. But, 
what is more extraordinary, they begin to retire still earlier in 
the more southerly parts of Andalusia, where they can be 
nowise influenced by any defect of heat, or, as one might sup- 
pose, defect of food. Are they regulated in their motions 
with us by a failure of food, or by a propensity to moulting, or 
by a disposition to rest after so rapid a life, or by what? This 
is one of those incidents in natural history that not only baffles 
our researches, but almost eludes our guesses!’”* 
In the Western Hemisphere, Cooper’s Shearwater of the 
Eastern Pacific, the Scarlet Tyrant of Argentina,’ and the 
Louisiana Water-Thrush of the South Carolina highlands? are 
typical examples of this early exodus-migration. 
Immediate failure of food is obviously not the cause of early 
exodus-migration in warm temperate regions, for in the later 
movements in these regions the transients from higher lati- 
tudes, in certain species, far outnumber the breeding represent- 
atives that had previously taken their departure, evincing that 
the food-supply had suffered no diminution. In resident 
species, the replacing of the breeding birds of a locality by 
individuals of the same species from higher life zones bears 
directly upon this point. . 
In the return-migration, it is also apparent that immediate 
failure of food is not the cause of the evacuation of the warmer 
areas of either hemisphere. Sooty Shearwaters might linger 
in force in the North Temperate Zone with the Black-vented 
Shearwaters if present supply of food alone was concerned.* 
Viewed in the light of the foregoing facts, it is evident that 
failure of food is not the immediate cause of a large part of 
migration. 
1JIn this relation, see migration reports Bull. Brit. Orn. Club on the early exodus 
movements of the Swift. 
* Cf. Hudson, Argentine Ornithology, v. 1, p. 154. 
3 Auk, v. 9, p. 34. 
#It is sometimes urged as a cause of return-migration that a special food is re- 
quired for nestlings, which can be obtained only at the breeding stations. The fallacy 
of this argument is particularly obvious when applied to transequatorial migration. 
