362 THE FOUNDING OF COLONIES BY ATTA SEXDENS. 



the feeding of larvae with eggs an equal number, and in an after- 

 noon during two hours eight cases of egg laying and four of feeding 

 these to larva?, I am of the opinion that up to the time of the 

 appearance of the workers eggs form the exclusive food of both the 

 mother-ant and her brood. I have never detected a case of the 

 Atta female supplying the larva? with either the mycelial threads 

 or the kohlrabi of the fungus {Rosites) . Furthermore, in opposi- 

 tion to Von Ihering's observations, I have never seen the mother-ant 

 herself feeding upon the kohlrabi. These structures appear on the 

 fungus-garden after about a month, but I have been impressed with 

 the indifference toward them uniformly displayed by the Atta female. 

 In order to test this, I have several times supplied a mother-ant which 

 had lost her supply of fungus with a portion thickly beset with 

 kohlrabi formations. This she has applied to the purpose of culti- 

 vation without showing any interest in the kohlrabi. The latter 

 remained untouched for a week or more and were finally hidden 

 under a growth of mycelial threads. The most conclusive proof 

 that the fungus plays no appreciable part as a food supply during 

 the earlier portion of the breeding period and up to the time of the 

 appearance of the workers is the fact that the female Atta is able 

 to raise her brood in the absence of this fungus, although in dimin- 

 ished numbers. I have, indeed, never seen such an occurrence under 

 natural conditions and only once under the artificial culture condi- 

 tions. A migrating female which left the nest on the 12th of March 

 had on the iTth of March been unable to secure any mycelial growth 

 with her transported fungus pellet, which remained blackish in 

 color. On the 18th of March the female secured from another ant 

 a piece of fungus from a fungus-garden. It was immediately cul- 

 tivated, began to grow, but early in April proved a failure. From 

 that time on the work was carried on without any fungus. The 

 number of larvae and pupae were fewer than in other colonies of 

 equal age, but on the 25th of April there were two workers of average 

 size, and on April 30 seven vigorous workers. The devouring of 

 eggs by isolated female ants is made evident in an indirect way by 

 the disappearance of eggs already laid, although the process has 

 jiever been observed by me. In regard to the feeding of the brood 

 hj a solitary maternal ant, Janet and Forel, as I rmderstand it, have 

 advanced the opinion that the larvae are fed with a nutritive liquid 

 derived from eggs devoured by her and which she secretes in the 

 communal stomach (jobot). In respect to Atta^ at least, this is not 

 the case. The eggs are h^re directly fed to the larvae. Comparative 

 investigations must determine whether or not this method of feeding 

 takes place with other ants. It is at least noteworthy that in the 

 case of Atta the larvae are not later on fed by the Avorkers with their 



