Todd-Carriker : Birds of Santa Marta Region, Colombia. 73 



valley of the Rio Cesar, and none from far up the Magdalena. We 

 must assume, therefore, either that they have been able to cross the 

 long stretch of humid forest country on the north coast, whereas 

 the species on the first list have not been able to do so, or else that 

 these conditions were not always present. But why, in the latter 

 event, should not all of the Arid Tropical forms have passed over, 

 supposing for the moment that they were able to adapt themselves to 

 a semi-arid habitat? 



Evidence is not wanting, however, to indicate that some of the Arid 

 Tropical forms have entered this region from the west and south. 

 There is reason to believe, for example, that Eupsychortyx leucopogon 

 leucotis is the parent from of E. leucopogon decoratus, and this in 

 turn of E. leucopogon littoralis, which latter has a very restricted 

 range, occupying the semi-arid region of the northeast coast, and is 

 replaced in the Goajira Peninsula by a distinct species, E. cristatus 

 cristatus. This happens to be the only case in which a peculiar form 

 has been developed in the Santa Marta region from an antecedent ' 

 inhabiting the lower Magdalena Valley, but there are several other 

 very suggestive cases in the category under consideration. Ortalis 

 garrula, Psittacula spcngeli, and Chlorostilbon Iiarbcrlinii, all of which 

 are replaced by a distinct species on the northeast coast, seem to have 

 entered from this direction, and doubtless others also, which are not 

 thus replaced. 



In the case of by far the larger number of semi-arid forms, how- 

 ever, we are not now able to say just where they entered the Santa 

 Marta region. The Rio Cesar Valley is known to be Arid Tropical, 

 at least for its upper part, and may have formerly sufficed to carry this 

 zone into the lower Magdalena Valley, whence its fauna may have 

 reached our region from the west or south. Unfortunately our knowl- 

 edge of this particular region is imperfect, but while it is most un- 

 likely that the Arid Tropical connection here is unbroken, as said by 

 Dr. Chapman, it may very well be that some at least of the forms 

 having such affinities have reached the northwest coast region in- 

 directly by this route, under different climatic conditions. 



We now come to a large class of species which are clearly Humid 

 Tropical in their faunal affinities. These occupy the forest regions 

 to the east of the Cienaga Grande and on the north coast respectively, 

 as alreadv shown, although some of them are able to exist also in 



