FROGS OF SOUTHEASTERN BRAZIL — COCHRAN 147 



shaped, olive mark extending from the axilla onto the center of the 

 chest. 



Color in life. — Some color notes on a half-grown frog from Recreio 

 dos Bandeirantes, collected on February 10, 1935, were taken from the 

 living specimen two days after its capture. Ground color gray, with 

 an irregular, broad band of dark color on each side extending from 

 ocular region to near the groin. A narrow dark-edged crossband 

 between the eyes. Arms irregularly marked with darker colors; one 

 longitudinal spot along the anterior part of the upper arm and some 

 larger ones on the forearm. Femur with large, rounded, partly con- 

 fluent yellow spots on the upper sides, on a dark brown ground. 

 Tibia with three large but rather indistinct crossbands. Foot gray, 

 with darker blotches. Disks of toes and fingers dark at the base. 

 Sides of the trunlv marked in the anterior half with a dark blotch; 

 an oblique dark band extending upon the posterior half and termi- 

 nating in a point. Rest of posterior half yellowish, except for a black 

 triangle in front of groin. Underside light, immaculate. Eye large, 

 iris gold-bronze, pupil rhomboid, with a slight notch above and below; 

 a prolongation of the rhomboidal pupil extended in a dark brown line 

 horizontally. A colored sketch of this individual was made at the 

 same time. On February 23, 1935, the frog, still alive, was compared 

 to this painting made 11 days earlier. The color change in the frog 

 had not been great in this interval; the back was a slightly more in- 

 tense yellow ochre, otherwise it had remained as it was when first 

 collected. 



Variations. — The bodily proportions are quite constant in the 22 

 examples of this species from the Recreio dos Bandeirantes. Some- 

 times the dark interocular bar is poorly developed. 



Remarks. — The specimens, USNM 118992-5, from Baixada Flu- 

 minense at Caxias are included with considerable doubt, since the 

 probable ranges of variation in the femoral length and in the foot 

 length do not overlap those from Recreio dos Bandeirantes. A larger 

 series from Caxias, however, may prove that those from Recreio dos 

 Bandeirantes are actually intergrading with the latter, and it seems 

 unwise to base any final conclusions about the Caxias frogs on so few 

 specimens. 



Specimens examined 

 BRAZIL: 

 DiSTRicTo Federal: Baixada Fluminense, at Caxias, Rio de Janeiro, USNM 

 118992-5, B. Lutz, November 1939. Recreio dos Bandeirantes, USNM 

 96388, 97587-97604, 97630-2, B. Lutz, Cochran, and Venancio, April 1935; 

 MZUM 104134 (2), Bailey, 1941. 

 Rio de Janeiro: Barro Branco, MZUM 104151, Bailey, 1941. 

 Sao Paulo: Boracea, MHNP 50-252, Bokermann, November 1948. 



