ANNUAL MIGRATIONS 119 



of migration outlined above has been the outcome. 

 It may be right or wrong but has the merit of being 

 amenable to experimental analysis. It would be 

 a quite simple matter, for instance, to control day- 

 lengths artificially, to observe if such manipulation 

 has any effect on the development of the gonads, to 

 examine these under the microscope for changes in 

 the interstitial tissue and, if changes were noted, to 

 liberate the birds and observe possible modifications 

 of behaviour that might be correlated with histo- 

 logical changes. It might even be possible to 

 induce reversed migration experimentally and 

 thereby procure tangible evidence that migration is 

 an inherited custom. 



In 1924 our first aviaries were erected. The bird 

 used was the junco {Junco hyemalis connectens) a 

 regular migrant, readily trapped; not pugnacious, 

 hardy and thriving in captivity. The last, and 

 most successful, series of experiments with juncos 

 was conducted in the fall and winter of 1927-8. 

 The birds were housed in a large and well designed 

 aviary the erection of which was made possible 

 through research grants from the Royal Society of 

 London. The winter turned out to be particularly 

 severe, thereby adding considerably to the value of 

 the results obtained. 



The general plan of the experiments was as fol- 



