386 



inner phalangeal tubercle on all toes (less web than in fUzingeri, much less 

 than in longirostris) ; first and second fingers equal; inner metatarsal tubercle 

 well developed; a small outer one; skin finely granular above; a fold over tym- 

 panum; a slight dorso-lateral fold; smooth below; vomerine teeth in two trans- 

 verse groups, well behind choanae and extending as far as their inner edges; 

 brown, a darker brown dorsal marking, beginning as an interorbital bar 

 (snout lighter) and extending to rump as an hourglass-shaped marking; faint 

 barring on upper lip, and on anterior face of thigh and on tibia; belly white; 

 under side of throat and hmt)S dusky; length 47 mm. 



Variation. — Young specimens show a white line on upper lip 

 and a black mark from snout through tympanum over arm to 

 middle of side and on to the white of the belly. 



Remarks.— Th\^ is close to longirostris but differs in longer 

 snout, less web on toes, absence of light throat stripe, larger 

 disks on outer fingers, and color of young. It occurs with 

 longirostris at Suretka, Almirante, and La Loma. From fitzingeri 

 with which it occurs at Suretka, Almirante, and Monte verde it 

 differs in longer snout, less web on toes, longer legs, larger 

 finger disks, absence of throat stripe, absence of light spotting 

 on hind side of thighs, and color of young. 



No other species can be confused with this one, which has a 

 rather restricted range (fitzingeri and longirostris are widely- 

 ranging forms). 



It has previously been called by me E. diastema and speci- 

 mens from collections have been distributed under that name 

 to various museums. I have noted B. M. N. H. no. 1925- 

 3-10-5, from Almirante; Berlin no. 28608 from Almirante, and 

 U. S. N. M. no. 73222 from Suretka. 



It has some similarity to nnbilus (Kscazu, Costa Rica; type, 

 B. M. N. H. no. 1902-5-13-29) in color and shortness of webs, 

 but nubilus is really much closer to fitzingeri, and, if anything, 

 is merely a short-webbed race of that form from upper Costa 

 Rica. Specimens from Kscazu, La Palma, and Guapiles may 

 represent it, [dthough I should hesitate to give a final opinion. 

 The type of nubilus agrees with fitzingeri in every character save 

 shortness of webs, and dorsal color. The dorsal marking of 

 nubilus is not that of talamancae. It has the throat mark, the 

 wartiness, the spotting on the hind side of the thigh, the shorter 

 legs, the shorter snout, the smaller finger disks, all fitzingeri 

 characters. 



