FROGS OF SOUTHEASTERN BRAZIL — COCHRAN 21 



smaller parotoid in the Sao Paulo toad, a purely individual difference, 

 in all probability 



In the half-dozen young toads from S§,o Paulo the skin is normally 

 roughened wdth tubercles, the parotoids are normally large and no 

 postfemoral light line is visible, although the dark doi'sal spots appear 

 definitely to arrange themselves in a line on either side of the light 

 middorsal area. An example from Guapi, near Teres6polis, in the 

 State of Rio de Janeiro, USNM 97664, has the same dorsal and post- 

 femoral coloring as the female, USNM 97772, from Sao Paulo, and 

 its skin is also not very rough. Three other specimens from Guapi 

 show a weak dorsal pattern and no postfemoral lines, while the re- 

 maining five are quite uniform in color on the entire dorsal and post- 

 femoral areas, with skin usually somewhat rough and occasionally 

 spiny. A half-grown toad from Manguinhos shows a dorsal pattern; 

 three adults from Caxias a short distance away are unicolor, although 

 one has a light middorsal line {B. dorsalis Spix). A series of eight 

 adults labeled simply "Rio de Janeiro" vary in the same way, five 

 without dorsal pattern (B. scaler Spix) and three with vestiges of a 

 pattern or a strongly developed one (B. ornatv^ Spix), the skin ranging 

 from nearly smooth to spiny. 



The two adults from Juiz da F6ra, in Minas Gerais, as weU as one 

 of the two from San Salvador, in Bahia, show some yellow spots on the 

 femur and sides, somewhat as illustrated by B. stellatus Spix. The 

 second adult from San Salvador and the one from Pernambuco already 

 mentioned are ashy in color, with scarcelj^ any trace of fighter spots 

 on these regions. A very large series of toads from all the States of 

 eastern and southern Brazil must be examined before the local degree 

 of variation can be properly analyzed. 



No specimens are at hand from Gavea, near the city of Rio de 

 Janeiro, which is the type locahty of B. crucifer mayi Miranda- 

 Ribeiro. Of the several specimens from other nearby localities, none 

 shows the dark spots invading the middorsal region as shown in the 

 figure of mayi. Onl}^ a large series of toads from Gavea can settle 

 the fixity of this color character. 



The same need for more material is felt in regard to establishing 

 fully the variety B. crucifer henseli Lutz from SSo Bento, in Santa 

 Catarina, and the variety B. c. roseana Miranda-Ribeiro, for which no 

 definite type locality'' was stated, but which is from Rio D'Oiuro. 



Although the type of Girard's B. gracilis from Rio de Janeiro is no 

 longer in existence, the descriptions and figures which he gave in his 

 amended notes of 1858 make it clear that his specimen belonged to 

 the species named B. crucifer by Wied in 1821. 



