Canon Tristram on the Polar Origin of Life. 209 



1. That all birds breed at the northernmost limit of 



their range. 



2. That those which penetrate furthest north for nidi- 



fication^ whether species or individuals, usually 

 retire furthest south. 



3. That all northward migration is for the purpose of 



nidification ; the southward being for food or 

 warmth. 



4. That the lines of migration are very different in the 



case of different species, and often intersect each 

 other. 



5. That birds which breed in the tropics do not migrate, 



unless in the case of birds which ascend the moun- 

 tains for nidification, and descend to the plains in 

 winter, as various Thrushes in the Andes, and num- 

 berless species in the Himalayas. 

 If it be asked, why, on the doctrine of heredity, should 

 not all birds migrate for nidification ? I would reply that 

 these tropical genera are the descendants of those which left 

 the Poles at a much earlier period than others, before their 

 structure or habits had become in any degree adapted to the 

 decreasing temperature, and therefore steadily retreated as 

 the cold increased. Take, for example, the great family of 

 the Humming-birds, one of the most differentiated in our 

 avifauna. The ancestors must have left the Pole by the 

 western shore-line of N. America. They may have been 

 confined originally to the portion of the old Arctic continent 

 nearest Bering^s Straits, or the parties which took the other 

 routes have perished and left no trace behind. Working 

 southwards they threw out colonies, especially to the east- 

 ward, and peopled the Antilles, where, as further south, the 

 sedentary parties soon became difterentiated. Still following 

 the retreating warmth, they settled in tropical South America, 

 and being a forest- and mountain-loving race, they clung espe- 

 cially to either side of the Andes. The vast eastern plain-region 

 of the Amazons was then probably too hot for them. Some re- 

 mained at different altitudes of the mountain-ranges, segre- 

 gated in little groups, venturing neither to face the heat of the 



