700: THE AMERICAN NATURALIST. [VoL. XXXIV. 
“In Nototrema testitudineum, as in Nototrema oviferum, the 
young undergo their complete metamorphosis in the maternal 
pouch; while in Vototrema marsupiatum they leave the pouch 
in the tadpole state." The tadpoles of JVozotrema marsupiatum, 
and those of another species, /Vozotrema plumbeum, are set free 
in water, according to Werner ('98). 
Yet another frog of the tropical Andes, Amphignathodon 
giintheri, is said to have a dorsal pouch, which is probably a 
receptacle for eggs. 
The female of a small Venezuelan species, Vototrema pyg- 
maum, bears from four to seven young in a dorsal pouch, which 
splits open when the development of the tadpoles is complete. 
The pressure of the active young starts the splitting of the pouch 
at the slit-shaped posterior opening, and the whole pouch is 
laid open in the middle line. The pocket is perhaps formed 
from lateral folds after the eggs are placed on the back, and is 
possibly not closed again after the young escape, but is formed 
anew at the next breeding season; or the pouch is perhaps 
never formed a second time. 
Observations on a number of tadpoles of Rhinoderma dar- 
winit have disclosed some of the facts of the development of 
this interesting frog. It was first discovered on the voyage of 
the Beagle, in the shady forest of the province of Valdiera in 
Chile. It was formerly supposed to be oviparous, but, in reality, 
the tadpoles develop in an enormously expanded gular pouch, 
which covers nearly the entire ventral side of the male frog. 
One account tells of fifty embryos in the pouches of five 
adults, and a much later account describes certain features of 
eleven tadpoles of one pouch. The tadpoles from one adult 
are not equally developed. : 
The youngest tadpoles described were thought to be just 
hatched. They measured 4 mm. to the tail; the alimentary 
canal was not differentiated, but contained much yolk. In all 
later specimens the alimentary canal was short, coiled, and of 
a yellow color, due to the presence of yolk. No external gills 
were observed, nor horny mouth-parts. Internal gills are not 
mentioned, although the anterior legs in young stages are con- 
cealed beneath a fold of skin (suggesting the operculum) : 
