1917.] Chapman, Distribution of Bird-life in Colombia. ' 155 



species, indeed, are so eminently terrestrial that they rarely fly more than 

 a few yards, and a continuous flight of several hundred miles would for them 

 be impossible. 



It is true that birds populate remote oceanic islands, but we do not find 

 among island forms such sedentaxy species as Formicarius rufipedus, Gral- 

 laricula flamrostris, and Siptornis erythrops, etc., which in but slightly differ- 

 entiated races are common to the Subtropical Zones of Colombia and Costa 

 Rica. Nor do we find species establishing themselves in regions which are 

 already occupied. Early arrivals on oceanic islands encounter no opposi- 

 tion, but the mainland offers no such favorable opportunities for settlement. 

 The available space is taken and the emigration of even a score of species 

 from one mainland home to settle in another mainland home, at a distance 

 of several hundred miles, is an unknown phenomenon in the distribution of 

 bird-life. 



^ In this connection it is quite to the point for us to compare the Subtropi- 

 cal Zone bird-life of the Eastern Andes with that of the Santa Marta moun- 

 tains. The subtropical portions of these mountains are separated by not 

 more than forty miles; nevertheless, of the one hundred and ninety-eight 

 species received by us from the Subtropical Zone of the Eastern Andes only 

 fifty-odd have been recorded as present or represented in the Santa Marta 

 group. On the other hand, as we have seen, some sixty species of 

 Colombian subtropical birds are present or represented in Costa Rica, 

 though here their ranges are separated by four hundred miles, or ten times 

 as great a distance as that lying between the northern end of the Eastern 

 Andes and the subtropical Santa Marta slopes. Geologists are, however, 

 agreed that the Santa Marta mass is of independent origin and has never 

 been connected with the Andean system. Zoogeographically, its life zones 

 above the Tropical Zone are, therefore, as much islands as though they stood 

 out in the Caribbean Sea. The absence from these zones of many common, 

 widely distributed species is doubtless to be accounted for by the fact that 

 they have never been connected with the corresponding zones in the Andean 

 system. 



If, therefore, so small a proportion of East Andean subtropical species 

 have crossed the forty miles lying between that range and the subtropical 

 slopes of Santa Marta, we certainly cannot account for the presence in a 

 continental area, distant four hundred miles, of a larger proportion of West 

 Andean subtropical species under the assumption that they have reached 

 Costa Rica fortuitously or by emigration. 



To my mind, the existence respectively in Colombia, Costa Rica and 

 eastern Panama of these two ends and a fragment of the Subtropical Zone, 

 is conclusive proof of a former physical connection of the areas concerned, 



