AVIAN MIGRATION AND REPRODUCTION 683 



were begun to test the hypothesis. Two species have been used 

 extensively in our laboratory, the slate-colored junco and the white- 

 throated sparrow {Zonotrichia albicolUs). Both are Temperate Zone 

 species and migrate through Evanston, Illinois, in spring and fall. 

 Evanston is also within the wintering range of the junco. 



If summation was a factor, wintering birds subjected to different 

 daily doses of day length should show differences in time of response, 

 with those receiving the most light in a given period of time, respond- 

 ing first. The cumulative action of daily photoperiods had been con- 

 sidered in relation to the testis cycle of nonmigratory starlings (Burger, 

 1939a,b, 1940) and house sparrows {Passer domesticus) (Bartholo- 

 mew, 1949) and the molt cycle in the tropical weaver finches 

 {Euplectes, Vidua, Stegamira) (Brown and Rollo, 1940). The results 

 of these investigations showed clearly the importance of the daily 

 photoperiod and the existence of threshold values for stimulation and 

 inhibition. 



In the first series of experiments (Winn, 1950, Wolfson and Winn, 

 1948) birds were captured during the fall migration and were sub- 

 jected to the following constant daily periods of light and darkness 

 beginnina December 4: 9L-15D; 12L-12D; 15.5L-8.5D; 20L-4D; 

 24L. One group of birds was also maintained under natural day 

 lengths. Since this experiment, which, incidentally, was conducted 

 twice, was designed to determine the response to different daily dura- 

 tions of photoperiod in 24-hr cycles, no attempt was made to differen- 

 tiate between duration of light, duration of darkness, or proportion 

 of light to darkness as the effective daily stimulus. 



All groups developed the premigratory physiological state and in a 

 definite sequence as follows: 24L, 20L, and 15. 5L — 40 days; 12L — 

 80 days; natural (mean daily photoperiod 10 hr) — 120 days; 9L — 

 160 days. The response of the testis also occurred in a sequence. 

 Marked testicular growth was observed in each group as follows: 

 24L, 20L — December; 15.5L — January; 12L — March; natural — 

 April and May; 9L — November. 



The results of these experiments showed clearly that the duration 

 of the daily photoperiod (in a 24-hr cycle) determined the time when 

 the premigratory physiological state occurred and when the testis 

 reached complete spermatogenesis. In relation to Rowan's original 



