1915] Cocoons Among Ants 301 
of the mangrove and a species of Wurmia. Sometimes two 
leaves have their edges spun together with a silken tissue, or a 
silken bladder with a short tubular entrance is constructed on the 
lower surface of a single leaf. The outer surface of the silk is 
thickly covered with minute vegetable particles. There are 
only 25 to 50 ants in a colony and the pupe are always naked. 
This is interesting, because all the other nest-spinning species of 
Polyrhachis, of which at least 20 are mentioned in the literature, 
are described as having the pup enclosed in cocoons just as 
they are in the species that nest in the ground or in logs. It is 
evident, therefore, that the larve of the subgenus Cyrtomyrma, 
like those of Gicophylla, though actively sericiparous in the 
construction of the nest, no longer spin a pupal envelope. 
All the species of Gtcophylla and Polyrhachis are confined to 
the Old World tropics. The third genus, Camponotus, is cos- 
mopolitan and comprises several hundred species, nearly all of 
which nest in the ground or in dead wood. Only a small num- 
ber of species and these, so far as known, peculiar to tropical 
America, build silken nests. Several years ago Goeldi (Forel 
1905) observed that the Brazilian C. senex employs its larvee in 
the construction of a somewhat globular silken nest on trees, 
and the var. fextor Forel of the same ant is known to make a 
similar nest in Central America. I am able to add a second 
species, a beautiful little, opaque, red ant, C. (IM yrmobrachys) 
formictformis Forel, to the list of forms with the same habit. Dec. 
28 and 30, 1911, my wife discovered two nests of this ant while 
we were walking in the neighborhood of Escuintla, Guatemala. 
One was between two broad liana leaves, which had been con- 
verted into an elliptical nest about four inches long and three 
inches broad by having their edges bound together with a broad 
film of white silk (Fig. 1). In this nest the silk was quite clean, 
like that of Gicophylla, 2. e., without admixture of any foreign 
material. The other nest, of about the same size, was of a dif- 
ferent appearance. It consisted of silk spun around the radiating 
petioles of a cluster of pinnate leaves on the branch of a legu- 
minous tree. No leaves were included in the walls of the nest 
so that very large quantities of silk had to be produced, and 
this silk was of a gray color and covered with numerous particles 
of extraneous vegetable matter. Each of the nests contained a 
flourishing colony of several hundred very active ants, which 
