1907.] Wheeler, The Polymorphism of Ants. 39 



tion to sterilize the earth which must have contained some predacious insect. 

 At any rate, I could find no traces of the larvae when I carefully examined 

 the earth several days later. 



This brief but interesting glimpse of the larval life of Xenodusa led me 

 to examine all the incerta nests I could find in the hope of obtaining more of 

 the parasites. Although I failed to secure any more of the larvae, I was 

 rewarded July 14 by finding a mature Xenodusa in a small colony on the 

 western hill. This colony had a depauperate appearance and comprised 

 only about 50 workers, two dealated females and a few worker pupae. It 

 contained no pseudogynes. 



These observations, made, as I have stated, within an area of less than 

 a square mile, prove that the larvae of X. cava are reared by F. incerta during 

 the latter part of June and the first weeks of July, and that the adult beetle 

 may be found as early as the middle of the latter month. During August, 

 however, the Xenodusce are found only with C. novceboracensis. That they 

 pass the winter with ants of this species is indicated by Blanchard's note 

 cited above. It is probable therefore that A'^. cava is heteroecious like the 

 European Atemeles, as Wasmann surmised,^ but instead of a Myrmicine 

 and a Camponotine host, both hosts in the American species are Campono- 

 tine ants. 



The occurrence of pseudogynes late in the autumn in so many of the 

 incerta nests agrees well with Wasmann's theory that these anomalous forms 

 are the result of Lomechusine parasitism, but there are several facts con- 

 nected with this occurrence which seem not to be in full accord with his 

 theory. A colony like the one above described, which contained nearly 

 80% of pseudogynes together with several normal winged females, was not 

 encountered by Wasmann in his study of the sanguinea colonies. He never 

 found more than 20% and usually the number did not exceed 5%. He 

 says, moreover: ^ 



"According to this theory the pseudogynes arise from converted female 

 larvae; from which follows, first, one ought to find no recently hatched nor- 

 mal females, or at least very few of them, in the same colony with recently 

 developed pseudogynes, since the female larvae of the respective generation 

 have been changed into workers as far as possible; second, the number of 

 pseudogynes in a single colony must not surpass the number of normal 

 females which they represent. I do not hesitate to subscribe to both of these 

 conclusions. The former, at least according to my own observations, is 

 actually substantiated, since I cannot remember ever to have found recently 

 developed females associated with recently developed pseudogynes. At 



» Zur Biologie der Lomechusa-Gruppe. Deutsche ent. Zeitschr., 1897, Heft II, p. 275. 

 8 Die Ergatogynen Formen, loc. cit., p. 635. 



