ANNUAL MIGRATIONS 139 



argument, let us suppose that they were our birds. 

 We should still have no proof that the northward 

 migration depended on the state of the reproductive 

 organs even though they were in a physiologically 

 acitve condition. A bird possesses various ductless 

 glands that might respond to variations in day- 

 length. Examination of all the endocrine organs of 

 our juncos and crows lends no direct support to this 

 assumption, but an obvious test can be applied to 

 the gonad idea. The world over it is common 

 practice to caponize domestic poultry, i.e. to remove 

 the testes by a simple operation. It converts the 

 rooster into a better table product. If the gonads 

 are in truth the physiological seat of the recurring 

 stimulus to the migratory impulse, a castrated bird 

 should show no migratory inclination, no matter to 

 what experimental conditions it may have been 

 subjected. It should remain sedentary when liber- 

 ated, or merely w^ander. 



This and other modifications, were planned for 

 the 1929 experiments but shortage of birds neces- 

 sitated all-round curtailment and precluded the 

 opportunity. It is hoped, however, to be able to 

 repeat the undertaking on a yet larger scale in the 

 future, when the hypothesis will be more precisely 

 tested. 



