15 U. S. NATIONAL MUSEUM BULLETIN 2 06 



tint; limbs light burnt umber above, with a wide crossbar on the tibia; 

 tips of toes a little darker than the dark soles of the feet. 



Variations. — A mutilated individual with the same data, USNM 

 96539, has bleached to an almost uniform drab, otherwise it agrees 

 with the described specimen in size and details of structure so far as 

 can be seen. An additional cotype, USNM 96412, came from Sacco 

 Sao Francisco near Nictheroy. It has likewise bleached and softened, 

 but its specific characters can quite readily be made out. 



A series of frogs was taken at Covanca near Jacar6pagud, in the 

 Federal District about an hoiu-'s drive by automobile west of the city 

 of Rio de Janeiro. They were hopping about in the dead leaves with 

 Leptodactylus marmoraius near a small impounded pool among dense 

 vegetation. The two largest females are both distended with eggs. 

 The smallest frog in the lot is 11 mm. long, and fuUy matamorphosed. 

 The specimens are very similar in color, with the dark lateral stripes 

 prominently setting off the paler dorsal tone which ranges from olive 

 to drab-gray, and on which a dark interorbital triangle, a middorsal 

 X and a sacral A sometimes make their appearance. The wood- 

 brown femiu- has a definite diagonal umber crossbar dividing it into 

 two light patches, while the tibia has a simUar dark bar running 

 straight across it and dividing it exactly in the middle. The tarsus 

 and foot are also crossbarred. The ventral surface is very sharply 

 set off from the dark lateral stripe, and is pearl-gray in alcohol. The 

 side of the head below the brown lateral stripe is also gray. Anteriorly 

 the femm- sometimes has a row of brown spots, or a fairly distinct 

 brown stripe, flecked with paler dots; posteriorly the brown color is 

 not sharply separated from the ventral tone but continues onto the 

 lower surface as a powdering of small brown dots. A more or less 

 distinct short brown stripe occurs on the proximal part of the upper 

 arm. 



Remarks. — Nothing is known of the development of eggs or tad- 

 poles, although many half grown young have been found. The col- 

 lector Joaquim Venancio believes that they breed all the year round, 

 as he has always found the yomig at any season. They frequent the 

 margins of lakes and pools of stagnant water. The voice is like that 

 of a cricket, pip-pip-pip-xnp, but difficult to distinguish. In life the 

 two stripes along the sides appear to be black, sharply set off from 

 the color of the belly which is white anteriorly and yellow posteriorly. 

 The species Physalaemus olfersi which it so closely resembles in color 

 is red below. 



I am indebted to Dr. E. R. Dunn for suggesting the identity of the 

 Rio de Janeu'o form with Phyllobates hrunneus from Mato Grosso, 

 Brazil, and from Colombia, Venezuela, and British Guiana. 



