1917.] Chapman, Distribution of Bird-life in Colombia. 91 



a more detailed knowledge of the Andean Ice Age, and of the variation in 

 altitude with latitude, of Andean life-zones, we can only speculate on the 

 extent to which the zones were affected by the rise and fall of a glacial period. 



We may with some hesitation suggest that s'o pronounced a boreal tj^pe 

 as Otocoris is a recent bit of glacial drift stranded on the Savanna of Bogota. 

 But on the other hand, with far more confidence, we may believe that the 

 undoubted northward extension of South Temperate Zone species along the 

 Andes, with increasing altitude, to the mountain crests of Colombia, has 

 been coincident with the retreat of the glaciers; and the often close rela- 

 tion of these birds with their sea-level, ancestral form supports this view. 

 Whether or not this be true there can be no question of the southern origin 

 of most of the species of the Paramo and Temperate Zones. 



The trend of life in the Tropical Zone is less susceptible of determination. 

 One cannot say that life does not radiate from an equatorial center and 

 flow north, and south, to the limits of the Tropical Zone; though the north- 

 ward current in America is now not only stronger but reaches farther. 



In the Subtropical Zone with its extremities reaching into Mexico, the 

 evidence also indicates a current setting toward the north. If, however, 

 this northward bound current is of post-glacial origin, it apparently follows 

 that the former subtropical bridge, which carried the numerous subtropical 

 species now found in Costa Rica to that country from Colombia (as suggested 

 beyond), has disappeared since the Glacial Period. 



It is in this connection that we find our best illustration of the biogeo- 

 graphic effects of the two other modifying factors, — subsidence and erosion. 



The retreat of the glaciers to higher altitudes with the resulting upward 

 extension of life-zone boundaries, is accountable for the formation of Paramo 

 Zone islands separated by Temperate Zone areas. Again Temperate Zone 

 islands have apparently been formed by erosion of the mountain crests 

 which at one time connected them. This appears to have happened in the 

 Western Andes where the close relation now existing between the life of the 

 Temperate Zone of the Andes west of Popayan and that of the ParamiUo at 

 the northern end of the range, suggests the former continuity of the Temper- 

 ate Zone on the crest of that range. 



Apparent proof of subsidence, doubtless accompanied by erosion, is 

 found in what I have termed the 'Panama fault' in the Subtropical Zone 

 which, after terminating at the northern end of the Western Andes, reap- 

 pears again on the crest of the higher mountains of eastern Panama and of 

 western Panama, though in the intervening areas it is widely separated by 

 the Tropical Zone. The evidence on which this theory of the former con- 

 tinuity of the Subtropical Zone from Colombia to Costa Rica rests, is pre- 

 sented beyond. 



