1917.] Chapman, Distribution of Bird-life in Colombia. 157 



I conclude, therefore, that those individuals of subtropical species which I 

 believe inhabited the Subtropical Zone of Panama went out of existence 

 with their zone. 



The -student of living species, unlike the palaeontologist, has no means 

 of determining geologic time. When the Panama 'fault' occurred cannot 

 therefore be determined from zoological evidence alone. The absolute 

 identity of many of the birds inhabiting the two widely separated ends of 

 the zone implies that they have undergone no change since their ranges 

 were disconnected. But neither degree of variation nor stability afford a 

 measure of time. 



StUl one may believe that under the influence of isolation the more 

 plastic species would show some differentiation from one another and the 

 fact of the continued close resemblance of forms, which elsewhere vary 

 geographically, indicates that this 'fault' in the subtropical stratum took 

 place at a comparatively recent period. 



The facts in the case suggest that the subsidence which has occurred in 

 Panama, and made parts of its southern coasts the islands of the Gulf of 

 Panama, has also involved the littoral of Colombia. The trend of the 

 Western Andes and the existence of the Atrato valley, make it improbable 

 that this range was connected with the range on the Colombian-Panama 

 boundary. If this be true, we may ask how so many subtropical species 

 could cross from the Western Andes to eastern Panama, and so few go from 

 the Eastern Andes to the much nearer Santa Marta group. The Baudo, 

 or true coast range, through a more southern connection with the Western 

 Andes might, however, have formed the bridge between the main Andean 

 system and the mountains of eastern Panama. We should then have had 

 four, instead of three ranges of the Andes in Colombia. The evidence in 

 support of this theory is far from conclusive, but includes the apparent 

 necessity of a larger tropical area than now exists at the Pacific base of the 

 Andes for the development of the Colombian-Pacific Fauna, and the strong 

 probability, as shown by its fauna, that Gorgona Island was once a part of 

 the mainland. 



Our attempts to reach the summit of the Baudo range have unfortunately 

 failed; but such collections as have been made there by Mrs. Kerr appar- 

 ently show that some Central America species rare, or not found by us in 

 the Western Andes, were common in the coast range. Our data, however, 

 are far from satisfactory, and further exploration in this range would, in 

 my belief, result in the discovery of facts of much significance to the 

 zoogeographer. 



