FROGS OF SOUTHEASTERN BRAZIL — COCHRAN 59 



after the quadrangle, and various spots and another obscure longitu- 

 dinal line on the flanks; the transverse borders of the legs are more 

 frequently two or three." 



Variations. — Two adult specimens were received from Miranda- 

 Ribeiro with the one which serves as the basis for the above redescrip- 

 tion. Of these, the male, USNM 101724, has the skin highly spinose 

 from the thickly scattered tubercles. Structurally it is not different 

 from the described specimen, but its color is much darker, the whole 

 dorsal surface having turned to seal brown or chocolate to such an 

 extent that the dorsal quadrilateral is only faintly discernible. Fem- 

 oral bands, however, are plainly distinguishable. A few small 

 brown spots occiu* along the sides of the belly extending into the 

 region of the axilla and groin. The other specimen, USNM 101723, 

 a female, shows the brighter, more complicated pattern described by 

 Miranda-Ribeiro. The legs in particular are brilliantly crossbanded, 

 including even the feet. Its skin is tliick above but quite smooth 

 except for some granules behind the ear and along the side. A 

 fourth specimen, FMNH 9006, also a topotype, taken by K. P. 

 Schmidt, agrees in every essential with the specimens described above. 



It will be interesting to learn the extent of stability in the sexually 

 dimorphic color pattern on a large series of individuals of this species. 

 Apparently males and females of the related species H. venulosa are 

 patterned alike, while the skin texture of the male does not differ 

 materially from that of the female in any of the specimens examined. 

 In H. mesophaea, the patterns of both sexes are also alike, while only 

 one out of the four males in the series has a roughened skin, MP 268, 

 from Espii-ito Santo, although this is perhaps due to the season of 

 capture, since skin rugosities may be present only during the breed- 

 ing period. 



Remarks. — Barbour and Dunn (1921) described from the main stream 

 of the Amazon between Mandos and Teffe a frog which they named 

 Paludicola imitator. Parker (1927b) pointed out that this species is a 

 Hyla. Therefore in Brazil we have Hyla imitator (Barbour and Dunn) 

 from the Amazon region, and Hyla imitatrix Miranda-Ribeiro from 

 Rio de Janiero. 



The recommendations under Article 36 of the International Rules 

 of Zoological Nomenclature caution against introducing similar 

 specific names ending in -tor and -trix into a single genus. But the 

 recommendation states that "once introduced, such names are not to 

 be rejected on that account." 



Specimens examined 

 BRAZIL: 



Rio de Janeiro: Teres6polis, USNM 101722-4, Miranda-Ribeiro; FMNH 

 9006, K. P. Schmidt, July 2, 1926; MHNP 31/45-46, Miranda-Ribeiro. 



