AVIAN MIGRATION AND REPRODUCTION 681 



and a discussion of this problem is beyond the scope of the present 

 paper. It is important, however, to recognize that a marked change in 

 physiological state precedes the onset of migration in the spring and 

 that this'state can be induced artificially in the laboratory by means 

 of changes in day length. This state can easily be recognized in live 

 birds by means of changes in body weight and observations of the 

 fat depots. In all the experiments that will be described, the occur- 

 rence of this state is used as an indication of a readiness to migrate, 

 but, with one exception, the birds were not released to determine 

 their behavior. The reproductive organs were studied when birds were 

 sacrificed. Thus, three manifestations of the premigratory state, in- 

 crease in body weight, marked deposition of fat, and gonadal recru- 

 descence, were studied. Data were obtained also on other phases of 

 the annual gonadal cycle. That the premigratory physiological state 

 is a good indicator of a readiness to migrate has been demonstrated 

 by studies of migratory restlessness (Zugunruhe) in caged birds (Eys- 

 ter, 1954; Farner, 1955, review; Weise, 1956). 



DAY LENGTH AND TROPICAL AND TRANSEQUATORIAL 



MIGRANTS 



The experiments of Rowan and Wolfson pointed clearly to the 

 increasing day lengths of winter and spring as the environmental 

 stimulus for spring migration, but there was a serious weakness in this 

 aspect of Rowan's theory. Birds that wintered in the North Temperate 

 Zone or northern sub-Tropical Zone would experience substantial in- 

 creases in day length after December 21, but what about birds which 

 wintered in the tropics, or on the equator, or in the Southern Hemi- 

 sphere? In the northern Tropics the increases in day length would be 

 slight; in the equatorial region there would be practically no change; 

 in the Southern Hemisphere, the birds would arrive in November 

 when the days were increasing in length, and after December 21 the 

 days would decrease in length until spring migration was underway 

 in March and April. Increasing day length, obviously, was not a 

 feature of the environment on the wintering grounds of all migratory 

 birds. 



To overcome this weakness of Rowan's theory, Bissonnette (1937), 



