AVIAN MIGRATION AND REPRODUCTION 709 



when the young can be reared at the time of maximum food supply. 

 If the breeding season is adapted to environmental conditions operat- 

 ing toward its close other factors must be postulated for the initiation 

 of the cycle. Although day length and temperature are acceptable for 

 higher latitudes, in the Tropics both "are too nearly constant to offer 

 a possible explanation" (Thomson, 1950, p. 182). 



There is little doubt that the early experimental studies which 

 employed increasing and decreasing day length to alter the gonadal 

 cycle and the fact that day length and temperature vary periodically in 

 higher latitudes have played a large part in the derivation of the above 

 conclusions. Our experimental findings, however, suggest that day 

 length should not be ruled out as a regulatory factor in the Tropics 

 simply because it is relatively constant. 



A pertinent question is whether tropical species respond to day 

 length. More experimental studies with tropical species are urgently 

 needed to answer this question, but the data from two series of ex- 

 periments (Brown and Rollo, 1940; Rollo and Domm, 1943) and 

 other observations suggest that the gonadal and molt cycles of some 

 tropical and equatorial birds can be altered by changes in day length. 

 Experiments were performed in the Chicago area with whydahs and 

 weavers {Steganura, Eiiplectes, and Vidua) which were imported 

 from Africa. The results demonstrated clearly that the daily photo- 

 period was a fundamental factor in determining the time of molt, the 

 type of plumage, and probably the activity of the gonads. 



Observations on the breeding season of birds transported from the 

 Tropics or Southern Hemisphere to northern latitudes (Baker and 

 Ranson, 1938) show that some tropical species change the time of 

 their breeding season to correspond with the seasons in the Northern 

 Hemisphere. For the tropical species that do not change it has been 

 suggested that they have an internal rhythm which is not readily al- 

 tered by environmental conditions. Another possible explanation is 

 that the day length requirements of these species is such that the 

 time of the breeding season in relation to the calendar year is not 

 altered. The most detailed observations on the effect of transport on a 

 tropical species are those of Orr (1945). These are especially signifi- 

 cant because the species used, Galapagos finches (Geospiza sp.) are 

 restricted to the Galapagos Islands. The individuals transported were 



