706 REPRODUCTION AND MIGRATION IN BIRDS 



In this connection, we have data from our laboratory which show 

 that a few birds, held under constant long days during the spring, 

 summer, and fall, initiated another gonadal cycle spontaneously in 

 the fall (Winn, 1950; Wolfson, 1955). In these instances, the pre- 

 paratory phase was completed under long days. In another instance, 

 one bird exposed to 20-hr photoperiods beginning April 6 did not 

 undergo regression and remained in continuous reproductive activity 

 for a little more than a year. Although these few birds are exceptions, 

 they indicate the tentative nature of our present interpretations. 



Day Length and the Timing of Spring Migration in Equatorial 

 and Transequatorial Migrants 



On the basis of the experimental results and the interpretations 

 presented here, the timing of spring migratory behavior in migrants 

 wintering on the equator, in the Tropics, or in the Southern Hemi- 

 sphere can be explained. Experimental work must be done with equa- 

 torial and transequatorial migrants to test the hypothesis presented, 

 but the data available so far for North Temperate species support it. 

 The relatively constant day lengths of the equatorial region can no 

 longer be regarded a priori as nonregulatory. Constant photoperiods 

 of 12 hr have been shown to be effective, and the duration of the 

 photoperiod, moreover, regulates the rate of response. Hence, birds 

 wintering in the equatorial region could be responding to the relatively 

 constant day lengths of 12 hr. The birds that cross the equator and 

 winter in the sub-Tropics or temperate regions of the Southern 

 Hemisphere are exposed to gradually increasing day lengths after they 

 arrive in late October or November and to gradually decreasing day 

 lengths after December 21, which reach a value of about 12 hr on 

 March 21. Therefore, the birds are exposed during their entire stay on 

 the wintering grounds to long days which, although they increase 

 gradually to a maximum and then decrease gradually to 12 hr, would 

 remain at an effective photoperiodic level, judging from our experi- 

 mental work with juncos. 



The main problem in equatorial and transequatorial migrants, as I 

 see it, is not the effect of the day lengths on the wintering grounds, but 

 the relation between day length and the preparatory phase and the 

 initiation of the progressive phase. At the outset, it seems highly prob- 



