401 



Barro Colorado specimens and for Brazilian ones. I am not sure 

 that the alhornarginata from Central America do not deserve 

 racial recognition. The glandular dorso-lateral fold is much less 

 developed in the northern specimens (seen from Barro Colorado 

 Island, Trinidad River, Panama; San Carlos, Costa Rica; Ma- 

 chuco, Maselina Cr., Nicaragua). Nor do I know whether the 

 Brazilian specimens have the red on webs and thighs. 



At any rate, the general type is represented in Central America 

 by a beast differing but slightly from the Brazilian, and by a 

 smaller beast, which has gone into the mountain brook habitat 

 and has very specialized tadpoles, as described by me in 1924. 



