142 THE RIDDLE OF MIGRATION 



shown that this tissue (in the junco and several 

 other passerine species) can be brought to known 

 stages of development by manipulating the external 

 stimulus — duration of day-length. It has further 

 been shown that the behaviour of birds so treated 

 actually varies upon liberation with the varying 

 state of the reproductive organs. 



The grounds for adopting the latter assumption 

 are partly based on the knowledge that the hormones 

 of the gonads control sex-behaviour in vertebrates 

 and since, in accordance with our definition, migra- 

 tion may be legitimately looked upon as a particular 

 phase of breeding behaviour it, too, should come 

 under control of this regulatory mechanism. Be- 

 haviour induced by internal secretions is instinctive. 

 All that we know of typical migration lends support 

 to the view that it falls into this category. 



In considering the evolution of migrations it has 

 been pointed out that we are discussing events of 

 the distant past and are left to draw inferences to a 

 considerable extent. We know, however, that 

 migrations are in some cases being built up at the 

 present time and by analysing the factors involved 

 in such instances, we can identify at least some of 

 them and be reasonably sure that similar factors 

 must have been operative in the past. Living 

 organisms as we know them today existed millions 



