360 THE FOUNDING OP COLONIES BY ATTA SEXDENS. 



ber of the patches, are clearly explainable by this process of manur- 

 ing. The occasional use of crushed eggs as a manure has never been 

 detected by me in any microscopic examination of the fungus garden 

 nor seen by direct observation, although in an indirect sense it is 

 evident that the eggs ma}^ be looked upon as a source of enrichment. 

 According to my observations the manuring of the fungus begins 

 very shortly after the flight from the mother colony, and it continues 

 until the establishment of the garden is completed. 



If an active Atta female is watched for several successive hours 

 it is seen that her work is methodically arranged. The inspection 

 of the apartments and the cleansing and leveling of the floor require 

 comparatively little time after the initial work is done, but are never- 

 theless attended to at stated intervals. Then comes the care of the 

 fungus gardens, a task requiring more time, followed by a third and 

 supremely important duty, the care of the young brood. Eggs 

 and larvae require to be diligently licked, placed in clusters, or spread 

 out separatelj^, put in contact with the fungus or separated from it, 

 and later on laid in the hollow center of the fungus garden, or cer- 

 tain individuals of the brood removed from it. By the use of a 

 plaster box or a Goeldi nesting case, Qgg lajdng and the process of 

 manuring can both be readily observed by means of a hand lens, or 

 photographs can be taken (see figs. 18 and 19). The ant raises itself 

 upon the middle and hind limbs and receives the &gg in its mandi- 

 bles. After a lengthy examination with its feelers the egg is placed 

 with the others in the center of the fungus garden. This procedure 

 is not always followed. By ver}^ careful observation, and especially 

 when the ant can be watched in profile, so that the motions of the 

 mouth parts are j)lainly observable, it is possible to discover that the 

 Qgg^ after being carefully tasted, is not actually deposited, but taken 

 up again, tasted anew, and then suddenly disappears in the mouth. 

 The mandibles are not, however, brought into action, but the ant 

 remains for some seconds perfectly motionless, the head held above 

 the fungus garden, the feelers moving lightly, indicating a state of 

 satisfaction. The mandibles and tongue now move rapidly, and the 

 front feet are drawn through the mouth parts in a peculiar way. It 

 is evident that this period of poise and activity indicates that the ant 

 is devouring or sucking the contents of the Qgg, which is pressed 

 between its mouth parts. Usually it is not immediately eaten, but 

 quite a period of tasting and an apparent state of indecision on the 

 part of the ant seems to take place. Much more infrequently those 

 eggs are eaten which have been actually deposited — at least this is 

 the case under normal circumstances, when, for example, the fungus 

 garden is already developed. If, however, this source of supply is 

 lacking, the before-mentioned occurrence is more common, and a 

 gradual decrease in the number of eggs already deposited can then be 



