150 tr. S. NATIONAL MUSEUM BULLETIN 206 



inner edge approaches its fellow until they nearly fuse across the back, and in 

 some cases they actually do become fused, with a few scattered light spots on 

 them which may indicate their usual limits. Sometimes there is an irregular 

 network of coarse dark spots down the back; sometimes an inverted V or its 

 separated elements may appear on the sacral region. The upper part of the 

 femur is usually irregularly spotted or reticulated; in only a few instances do 

 these spots suggest the regular dark cross bars characterizing some other members 

 of this group. Quite often the posterior surface of the femur reveals a rather 

 large, very irregular, elongate light spot surrounded and set off by narrow dark 

 reticulations. In other cases the posterior femur is finely marbled with small 

 light and dark spots in about equal proportions. The upper surface of the tibia 

 often has three irregular oval spots placed diagonally across it. Any of these dark 

 spots may lighten, leaving only a narrow dark outline which suggests a parallelo- 

 gram or an irregular circle. Numerous small dark dots scattered over the upper 

 surfaces still further obscure the pattern. The sides sometimes show linear rows 

 of dark dots more or less fused into a network, or often a scattering of fine dots 

 with no particular arrangement. The groin quite often has a very pale (yellowish 

 in life) area, confined sometimes to an elongate irregular spot encircled by darker, 

 heavier reticulations. The ventral surface appears to be immaculate in the 

 specimens at hand. A good many individuals show a dark brown stripe along 

 the canthus, with considerable spotting on the upper lip; other specimens have a 

 much lighter canthal stripe and the labial spots are reduced to a light marbhng. 

 Sometimes the canthal stripe is continued over the tympanum as a narrow dark 

 line, widening behind the ear and often ending in front of the forearm, but also 

 sometimes carried part way down the side of an irregular dark stripe. 



Remarks: It is perhaps rash to give a new name to another of the rubra group, 

 the most variable of any in Brazil, but no formerly proposed name seems to be 

 applicable to the form coming from within the Federal District. H. x-signata of 

 Spix from Bahia is apparently nearest to similis, after fuscovaria, but fresh Bahian 

 material is needed before an exact comparison can be made. 



This species has the peculiar habit of "swarming," as hundreds appeared at one 

 time on a tree outside the laboratory at the Instituto Oswaldo Cruz. The voice of 

 the adult is a high-pitched crack, crack, crack, crack, sounding somewhat like that 

 of a duck. Breeding takes place in March and April. The tadpoles metamor- 

 phose very rapidly, taking only 20 to 30 days to change into frogs, but they are 

 rather delicate, as they die from the effects of too much sun if the water in which 

 they live is not deep enough. 



While this species is exceedingly plentiful at Manguinhos, it is not known from 

 the southern part of the state of Rio de Janeiro. In fact, only two other examples 

 at all resembling this form are known outside of the Manguinhos material. One 

 of these is from Bom Successo, a few minutes' ride by automobile from Manguinhos 

 while the other, no. 96213, not in good condition but seeming to belong to the 

 present species, is from Amorim near the city of Rio de Janeiro. 



Specimens examined 

 BRAZIL: 



DisTRiCTO Federal: Amorim, USNM 96312, A. Lutz, January 1926. Bom 

 Successo, USNM 96203, A. Lutz, Oct. 25, 1928. Manguinhos, USNM 

 97317 (type of H. similis), Venancio, Feb. 25, 1935; USNM 97312-6, 97318-52, 

 97374-6 (paratypes of H. similis), Venancio, January- May 1935; USNM 

 96144-6, A. Lutz, January 1922. Rio de Janeiro, USNM 81119-21, A. Lutz, 

 1930. 



