FROGS OF COLOMBIA — COCHRAN AND GOJN 283 



distance equal to the interocular space, and terminate above the tip 

 of urostyle. In one of the largest specimens from Aracataca (ANSP 

 19755, 26.2 mm.), the dark stripe on each side from tip of snout to 

 groin is present but the rest of the dorsum and the top of the head 

 and snout are marked with vermiculations which show hardly a sign 

 of longitudinal arrangement but which do form an indistinct inter- 

 ocular bar, something never seen in Colombian specimens from other 

 localities. Two specimens (MZUM 90168) from Curumani are likewise 

 larger (24.5 and 28.3 mm. in head-body length) than usual, and one 

 is not so definitely striped as is usual for Colombian individuals. 



It seems probable that the above-mentioned specimens simply 

 show the influence of Hyla misera from the east. Rivero (1961, pp. 

 135, 136) points out that true misera ranges westward to the vicinity 

 of Sierra de Santa Marta where it seems to blend with microcephala. 

 Since we have not seen sufficient material from critical areas to 

 demonstrate intergradation, we presently consider all the material 

 from northeastern Colombia that we have seen as microcephala and 

 forego the use of the trinomial. 



Another series of specimens of Colombian microcephala showing 

 divergence from the typical individual is from Girardot, on the east 

 bank of the Rio Magdalena in Cundinamarca opposite the state 

 of Tolima. Specimens in a series (JRT 604-12) from this locality, 

 far up the Magdalena Valley, are smaller in size (19 to 21.9 mm.) 

 and also show a tendency for the dorsal pattern to become modified. 

 In all of these there is a marked tendency for the middorsal pair of 

 stripes to break up into blotches, and in two specimens the dorsal 

 pattern between the lateral stripes is merely a series of blotches, 

 with no middorsal stripes discernible. 



Remarks. — We have examined the type of Hyla underwoodi (BM 

 94.11.15.32-3 [1947.2.23.28]) and while we believe that it is certainly 

 close to if not identical with H. microcephala we do not believe we can 

 make a definite synonymic assignment without seeing ample fresh 

 material from the type localities of both forms. 



Specimens Examined 

 COLOMBIA 



Atlantico: Palmar de Varela, USNM 152143. 



BolIvar: Near Gambote, CJG 2444-5; Told Viejo, Rio Pechelin, CJG 



2459; Sincelejo-Tolu, CJG 2457 (20). 

 Choc6: Atrato, CNHM 74918 (2), 74919. 

 Cundinamarca: Girardot, JRT 604-12. 

 Magdalena: Aracataca, ANSP 19755-7; Curumani, MZUM 90168 (2), 



USNM 118247-8; Fundaci6n, MZUM 48281-2; El Banco, on Rio Magdalena, 



ANSP 25061. 

 Santander: 4.7 km. southeast of Bucaramanga, JRT 390-3. 



