1915] Cocoons Among Ants 
ive) 
Se) 
oO 
Still another inference may be drawn from the presence or 
absence of cocoons among the various nest-spinning ants. As 
we have seen, the cocoons are always present, so far as known, 
in Camponotus, even in the nest-spinning species; in Polyrhachis 
the cocoon is rarely absent and then only in the species of the 
subgenus Cyrtomyrma, whereas in Cicophylla it is always 
absent. Very few species of Camponotus spin nests; a number 
of species of Polyrhachis have acquired this habit, and all the 
species of CGicophylla exhibit it in its highest manifestation. 
This is shown also by the hypertrophic development of the 
sericteries of young Cicophylla larvee, as observed Chun (1903) 
and Karawaiew (1906). Apparently, therefore, in this genus 
the spinning habit has been shifted back en bloc to an early 
larval stage and is no longer manifested for cocooning in late 
larval life; whereas in Camponotus and Polyrhachis (excepting 
the species of Cyrtomyrma) more mature larve are used as 
shuttles and the cocoon-spinning instincts have not been 
suppressed.* The employment of the larvae as instruments 
for spinning a silken nest in at least three very different genera 
is only one of several cases of convergent development among 
ants. While studying the Australian species during the 
autumn and winter of 1914, I detected another case of an 
equally adaptive convergence in the coloration of the cocoon. 
In the great majority of ants the cocoons are white, cream, or 
pale buff-colored. In some Ponerinze (Odontomachus, Stig- 
matomma, etc.) they are brown, but the cocoons of the species 
to which I refer are dark brown or black. Such cocoons I 
have found in three very different genera of Ponerine, namely, 
Diacamma, Rhytidoponera sens. str. and Leptogenys (subgenera 
Lobopelta and Odontopelta), and as nothing has been published 
on the habits of the species in question, I here subjoin a few 
of my field notes. 
*Wasmann (1905), in his rendering of Jacobson’s observations on Polyrhachis 
(Myrma) dives F. Sm., states that one of the larve used as a shuttle measured 
5mm. It must, therefore, have been much more nearly full-grown than the larve 
employed by CGicophylla for the same purpose. In this connection reference may 
be made to Technomyrmex bicolor Emery subsp. textor Forel of Java, which, accord- 
ing to Jacobson (Forel, 1909) inhabits, on the bark of trees, little nests 2 cm. long 
by 1 cm. in diameter, consisting of silk mixed with vegetable detritus. These 
nests are described as resembling those of Polyrhachis psena in miniature. If, as 
Forel remarks, these nests are really the work of the Technomyrmex, we shall have 
to regard this as a fourth genus which has independently acquired the habit of 
employing its larve in nidification. This would be of considerable interest, 
because Technomyrmex is a Dolichoderine ant, whereas all the nest-spinning species. 
above mentioned are Componotine. 
