A NUTRITIONAL STUDY OF INSECTS 63 



colony is grown upon the liquid excrement of the female, which 

 touches small pieces of the mycelimii to the tip of its gaster and 

 then replaces them in the fungus garden. Some of the eggs are 

 sacrificed as manure for the fungi. 



Parrott and Gloyer ('16) are of the opinion that the- fungus- 

 growing habit of the Attii may have come about through the col- 

 lection of caterpillar feces in which many spores accidentally 

 eaten by the larvae are still ahve. They do not give any expla- 

 nation of the habit of collecting the feces, however. \\'Tieeler 

 ('07) points out that the gardens are usually composed of a sub- 

 stratum consisting largely of fecal material in the case of the am- 

 brosia beetles and termites, and that the habit is pronounced in 

 the lower genera of leaf-cutting ants and visible in all cases close- 

 ly studied. It therefore seems probable that the food of the 

 Attii ''may have been originally grown on fecal substances" 

 Von Ihering ('94) beheves that the habit may have phylogenet- 

 ically originated with ants using moldy seeds stored as food. I 

 should suggest the possibility that the ants may originally have 

 fed upon fungous myceha developing on caterpillar feces from 

 spores unkilled in their passage through the digestive tract. 

 Such droppings might finally be carried into the nest where the 

 fecal substratum and the moisture of the nest would soon allow 

 the growth of a valuable crop of fungous food. Thus the fungus- 

 growing ants developed their habits as a direct response to a val- 

 uable food supply. On the other hand, the termites developed 

 their habits as a means of making use of an umnanageable food 

 supply. 



The examples cited above indicate that the use of microor- 

 ganisms as food-" is widespread among ijisects and is a direct re- 

 sponse to the high food value of the fungous cells. The feeding 

 habits may be grouped into three classes as follows : 



1. Ingestion of microoganisms with substratum, i.e., Dro- 

 sophila, Musca, Sciara, worker termites. 



2. Feeding directly on microorganisms, i.e., mites, tree 

 crickets, many adult Diptera, etc. 



-^ See page 68 for case of mosquito larvae feeding on fungi. 



THE JOURNAL DF EXPERIMENTAL ZOOLOGY, VOL. 28, NO. 1 



