Amphibians and Reptiles of Santa Marta 17 



Pueblo V''iejo, who had been employed by him as hunters during all his 

 activities on the north slopes of the Sierras. Contrary to Mr. Brown's 

 statements that he himself collected all of his specimens, I have unquestion- 

 able first-hand information that both in Bonda and in the Sierras, at least 

 at Pueblo Viejo, San Miguel, and the paramos, he did little or no shooting, 

 employing native hunters for that purpose, while he prepared the specimens. 

 As in the case of Smith's data, I would hesitate to use any distributional 

 records based solely on Brown's collections, unless supported by other data. 

 Mr. Brown worked both sides of the Sierras, at Pueblo Viejo (then 

 San Antonio), Chirua, Santa Cruz, San Miguel, and Alacotama on the 

 north slopes, and San Sebastian, Pueblo Viejo, and El Mammon on the 

 south slopes, as well as the intervening paramos of Macotama, Chiruqua, 

 and Chinchicua. Notwithstanding the fact that he passed through Dibulla 

 and Riohacha and traversed the road from the latter place to Valle de 

 Upar, no collecting was done in any of these regions, so that the first defi- 

 nite information secured concerning their bird fauna was by the writer at 

 Dibulla and Riohacha in 1914 and along the road from Riohacha to Valle 

 de Upar and thence around to Fundacion in 1920, when attached to the 

 University of Michigan expedition. 



It is to be understood that I have no reason for making these comments 

 upon the work of Brown and Smith other than the desire that our knowl- 

 edge of the fauna of the region should be free from the confusion caused 

 by inexact records. 



Geographical Situation: The geographical boundaries used for the 

 region discussed in this paper (cf. map) are as follows: On the north and 

 northwest the Carribean Sea; on the northeast and east the Goajira Penin- 

 sula, or. more exactly speaking, the Rio Rancheria (called Calancala on its 

 lower reaches), up to where it turns abruptly northwest to enter the Sierras; 

 from this point on the Rio Rancheria on a straight line to San Juan de 

 Cesar, on the Rio Cesar; thence down the Rio Cesar to where it divides 

 into two channels (Cafio Corredor and Cano Matanza) ; from this point 

 in a more or less straight line to Alto de las Minas, passing through Cam- 

 perucho, and thence to Ariguani ; from there the trail to Fundacion, the 

 terminus of the railway from Santa Marta; and from Fundacion a line to 

 the south end of the Cienega Grande and thence across to Playa Caiman 

 and the sea. The line running across from the Rio Cesar to Cienega 

 Grande is a purely arbitrary one, since it does not designate the boundary 

 of any known habitat. However, the upper Rio Cesar and Rio Rancheria 

 represent an actual faunal and physic^raphical boundary between the 

 Sierra Nevada and the eastern Andes, and I have used no records from 

 localities on the slopes or foothills of the eastern Andes, such as Villanueva 

 and Manuare (included by Simmons). The lower Rio Rancheria is also 

 an arbitrary boundary line, since it traverses the arid Goajira Peninsula 

 and cuts through a distinct faunal area which is well represented around 

 Riohacha. Thus we have in the mountain mass an absolutely isolated area 

 containing its own distinctive habitats and faunal characteristics, into which 



