238 U.S. NATIONAL MUSEUM BULLETIN 2 88 



lower part of eye, and through the tympanum and terminates in the 

 groin. The edge of the upper lip is pale, as is the side and venter 

 below the lateral stripe. 



Remarks. — It bothers our zoogeographic sense somewhat to use 

 the name "squalirostris" for this single specimen, for the Serra da 

 Bocaina and the Maracaibo Basin have little in common faunistic- 

 ally; however, our specimen is certainly close to, if not identical 

 with, the Brazilian specimens of the species. 



The single specimen at hand seems to differ structurally from the 

 Brazilian specimens only in that the snout is a little shorter and the 

 tympanum a little larger and set a little closer to the eye. It is im- 

 possible to judge the significance of these differences, as they are 

 based on comparison of a single specimen with a population about 

 which we know little concerning geographic variation. 



There seems to be no doubt that squalirostris is allied to parkeri 

 Gaige, evelynae Schmidt, and lindneri Miiller and Hellmich and, 

 through them, to the rubra group. However, we do not feel that this 

 justifies putting these species in the rubra group as at present under- 

 stood, for it seems probable that such forms as rubra and hayi are 

 much more closely related, one to the other, than squalirostris and 

 parkeri are to either of them. 



Specimens Examined 



COLOMBIA 



Norte de Santander: 20 km. west of Cucuta, MLS 122. 

 ARGENTINA: Maschwitz, Buenos Aires, MCZ 35421-2. 

 BRAZIL: Serra da Bocaina (boundary of Rio de Janeiro and Sao Paulo), 



USNM 96719-20 (cotypes of squalirostris), 96608-13; Bonita, Sao Paulo, 



USNM 96721. 



12. rubra group 



At the present time this is one of the most difficult groups among 

 all the South American tree frogs. H. rubra, sensu latu, is one of the 

 most widespread and one of the most variable of all South American 

 frogs. Every local population differs in some respects from other 

 local populations, in some cases quite markedly. We have tried to 

 consider all of these geographic variations as simply populations of 

 rubra and have restricted nomenclatorial recognition to distinct, 

 sympatric populations that show an obvious relation to rubra. The 

 group as a whole is represented by small to medium-sized Hyla with 

 reduced webbing between the fingers; vomerine teeth in two small 

 rounded patches; the absence of patagia, ulnar, or tarsal ridges; and 

 by the presence of distinct markings on the posterior surfaces of the 

 thighs. The three species now known from Colombia may be distin- 

 guished by the following key: 



