126 THE RIDDLE OF MIGRATION 



in our chain that connects gonad changes with varia- 

 tions in day-lengths.i 



Microscopic examination was made of the repro- 



^ Trans-equatorial migrants appear to present an obstacle to 

 the theory here outlined. When they cross the equator in Sep- 

 tember or October they are encountering day-lengths that are 

 increasing. The effect of this on j uncos and similar northern 

 birds would be induced recrudescence of the gonads but there is 

 no such response in trans-equatorial migrants. Their organs 

 reach the minimum and remain there through the northern winter 

 months in spite of the fact that the birds are then actually enjoy- 

 ing the long days of the southern summer. If one supposes that 

 in these species the annual rhythm has become stabilized and is 

 not susceptible to interruption by exposure to adverse light con- 

 ditions for a single winter, then sojourn in the south would have 

 no effect. Yet compulsory retention in the south for a period of 

 years might prove a very different thing. Deciduous trees, for 

 instance, taken from the northern hemisphere to the tropics, 

 retain the habit of shedding their leaves during the northern 

 autumn months, but only for a few years. Thereafter the pe- 

 riodicity ceases. The leaves are cast at any time of the year. 

 The timing elements of the north are wanting in the tropics. 

 The annual rhythm, although persistent for a few years, finally 

 breaks down. Trans-equatorial migrants are exposed for part 

 of each year to the northern environment. This may be quite 

 sufficient to perpetuate the annual rhythm throughout their 

 life-time. There is some evidence (storks) that permanent 

 removal from the north may ultimately end in the adoption of a 

 rhythm in harmony with southern conditions, the exact opposite 

 to the rhythm of the north. 



