180 U. S. NATIONAL MUSEUM BULLETIN 206 



small tubercles on top of the head, faintly suggest H. catharinae 

 Boulenger, but the other characteristics of structure and pattern are 

 so utterly dissimilar that there can be no confusion when accurately 

 identified examples of both species are laid side by side. 



Suspecting that Hyla hilli Boulenger, from near Teresdpolis, was 

 synonymous with Hyla microps, from Nova Friburgo, a short distance 

 to the northeast in the same mountain range, I sent photographs of 

 the type of microps to the British Museum (Natural History) for com- 

 parison. Under date of December 17, 1936, H. W. Parker writes, "I 

 have compared the photographs of the type of Hyla microps with the 

 type of Hyla hilli and feel convinced of their identity; in fact the 

 photographs might almost be those of Boulenger 's species." 



Specimens examined 

 BRAZIL: 



EspfRiTO Santo: Rio Novo, NHMH 2814 (4), Erhardt, 1918-19. 



Rio de Janeiro: Bonito, Serra da Bocaina, USNM 96595-96600, 96708-9, 

 96711, A. Lutz. Nova Friburgo ZMB 7472 (type of H. microps). Teres6polis, 

 BM 1914.3.20.9 (type of H. hilli), Hill. 



Santa Catarina: Humboldt, USNM 66567-9, Fritsche, November 1918; USNM 

 101451-2; MZUM 58515 (7); ZSBS 67/1925, Erhardt. Itapocu, Jaragu^, 

 NHMH 2604, Erhardt, January-March 1907. Lages, BM 88.2.7.23-25 (4), 

 Michaelis. Rio Humboldt, BM 1923.6.1.92-99, Fritsche. Teres6polis, 

 NHMH 1825, Werner, 1916. 



8. geographica — group 



The frogs of the geographica group are at once recognizable among 

 other Brazilian hylids by their extremely attenuate limbs. The 

 upper arm, in particular, is unusually slender, and the tibia is narrow, 

 its width measuring but a small fraction (about one-eighth) of its 

 length, hence the contrast is striking between this and the rubra 

 group, in which the great width of the tibia is perhaps the best group 

 character. 



As members of these attenuate species apparently are not numerous 

 anjrwhere, only 24 adult individuals of the Rio de Janeiro subspecies 

 punctatissima having as yet come to my attention, extensive com- 

 parisons have not been practicable. The character distinguishing the 

 two subspecies, according to Parker (1933, p. 10), seems to be the 

 degree of webbing of the fingers, which in typical geographica geo- 

 graphica from Trinidad, Ecuador, and Peril are two-thirds webbed, 

 whereas the southern subspecies geographica punctatissima from 

 Brazil and Bolivia show a web only one-half the length of the fingers. 

 In most of the Eio de Janeiran frogs I have examined, the web seems 

 even less than one-half the finger length. 



