90 THE AMERICAN NATURALIST. [VoL. XXXVI. 
thereby indicating that they were perhaps members of the 
same colony. They soon distributed their progeny and provi- 
sions in three separate piles — one for the larvae, one for the 
pupæ, and one for the seeds. During the first few days of 
their captivity the ants were fed on house flies. These were 
not only eaten with avidity by the adult Pogonomyrmex, but cut 
into pieces and fed to the larvae in the same manner as I have 
described for the Ponerinz and some Myrmicinz ('00 and '00a). 
On one occasion nearly every larva in the nest could be seen 
munching a small piece of house fly. But a still more remark- 
able method of feeding was adopted after a few days, when the 
supply of insect food was exhausted. Then the ants were seen 
to bring seeds from their granary, crack them open with their 
strong mandibles, and, after consuming some of the softer por- 
tions themselves, to distribute the remainder among their larva. 
The latter could be seen under the lens cutting away with their 
mandibles and devouring the softer starchy portions of the seeds. 
The hard and useless hulls were afterwards carried away by the 
ants and placed on the refuse heap. These observations show 
that the Jarve of certain ants are not only able to subsist on solid 
food, but even on food of a vegetable nature. The adaptation of 
what were probably once exclusively carnivorous ants to a 
vegetable diet, although not yet complete, is, nevertheless, so far 
advanced that the larva already participates in the peculiar feed- 
ing habits of the adult insect. The P. zmberbiculus seem not 
to possess the power of feeding one another or their larvae by 
regurgitation. At any rate they were not seen to make use of 
this method in the artificial nests. 
These observations are quite in line with some which I 
made on artificial nests of the large “agricultural ant of Texas ” 
(P. barbatus Smith, var. molifaciens Buckley). In this case the 
workers carried the seeds, a few at a time, into the chamber 
containing the queen and her attendants. Here the ants, 
including the queen, gnawed away the soft portions of the 
seeds till they had satisfied their hunger. Thereupon the 
empty hulls were carried out. Even when the nest was sup- 
~ plied with honey or syrup, each ant helped herself from the 
v food supply, and neither fed other ants nor ed herself 

