AVIAN MIGRATION AND REPRODUCTION 685 



designed specifically to answer this question. Birds were caught 

 during the spring migration in April and May and subjected to 

 constant day lengths of 9, 12, 20, and 24 hr (photoperiods in a 24-hr 

 cycle), and to natural day lengths. The gonads were partly developed, 

 and the birds showed also heavy fat deposits indicative of the migra- 

 tory physiological state. 



The results of these experiments were generally in accord with 

 those of the previous series, but a number of remarkable differences 

 were found. The longer photoperiods of 20 and 24 hr induced a more 

 rapid and greater development of the gonads compared with the 

 birds exposed to 12-hr photoperiods and to natural day lengths. This 

 was especially true in some of the females. In the 9-hr group the 

 gonads regressed almost immediately. In the previous experiment it 

 was shown that a gonadal response could be induced with a constant 

 9-hr photoperiod, yet in this case regression occurred. In the present 

 experiment the testes were in an advanced condition correlated with 

 the natural day lengths of 14 hr or more at the start of the experiment. 

 When the day length was reduced to 9 hr, perhaps the daily increment 

 of gonadal stimulus was too small to meet the demands of the partly 

 enlarged testis. This interpretation is suggested also by the fact that 

 some of these birds initiated another gonadal cycle many months later, 

 although held at the 9-hr photoperiod. However, whether complete 

 spermatogenesis occurred was not determined. Was the reduction in 

 day length as such responsible for the regression in the 9-hr group? 

 Probably not, for in the 12-hr group, which experienced a reduction 

 in day length (from natural conditions) of approximately 2 hr at 

 the start of the experiment, gonadal growth continued, but at a slower 

 rate apparently than in the natural group. The extent of development 

 showed no marked differences, but one unexpected and highly signifi- 

 cant difference did occur. In the natural, 20-, and 24-hr groups the 

 gonads regressed (with one exception) after a few months of activity 

 as occurs in nature. In the 12-hr group complete activity of the 

 testes was maintained for about 9 months. 



Fat deposition continued in all the groups, but the fat deposits 

 disappeared first in the 9-hr group. The largest fat deposits were seen 

 in the 12-hr group. The deposits disappeared (with one exception) 

 in the natural, 20-, and 24-hr groups before they molted, but most 



