1907.] Wheeler, The Polymorphism of Ants. ]1 



instahilis, it will be necessary to return to the worker semipiipfe on which 

 the Orasema occasionally deposits her eggs. In such cases the young 

 larvpe of the parasite must be very inadequately fed and proljably soon die 

 and fall off, leaving their hosts in a depleted condition but still able to pass 

 on to the pupal stage. Now in all the nests infested with Orasema, and 

 only in these, I have found a number of peculiar puppe like those represented 

 in Plate II, Figs. 24-26 and Plate V, Fig. 65 s, and differing from the normal 

 worker pupse (Fig. 27) in several important characters. They are smaller, 

 of a waxy white color, with more decided intersegmental constrictions and 

 are broad behind and very narrow anteriorly. The head is remarkable for 

 the very small size of the brain and eyes and the situation of the latter on 

 distinct stalks. The mandibles, too, are abortive. The resemblance of the 

 head to depleted female pupse, like the one represented in Fig. 16, is very 

 striking. The thorax is extremely slender and the gaster has a high fold on 

 each side, and in balsam preparations (PI. II, Fig. 26 and PI. V, Fig. 67 i) 

 is seen to contain a number of large urate masses in the corpus adiposiivi. 

 In many specimens the gaster is concave ventrally, with its tip turned 

 upward and forward. Although these singular pupse are carefully cleaned 

 by the workers and kept with the normal individuals, they never succeed in 

 hatching. After lying in the chambers for many days without even acquir- 

 ing a deeper color of the body or pigment in the eyes, they are either carried 

 to the refuse heap or eaten by the workers. I am convinced that these 

 extraordinary pupse, which may be called phthiser gates, have arisen from 

 worker semipupse that have had part of their juices sucked out by Orasema 

 larvae, so that only enough formative material was left to produce pupse with 

 very defective head and thorax and hence quite unable to develop as far as 

 the imaginal instar. It is interesting to note that these microcephalic, 

 microphthalmic, and stenonotal characters represent merely greater diminu- 

 tion of the similar characters of the normal workers as compared with the 

 more macrocephalic, macrophthalmic, and eurynotal soldiers and females. 

 The theoretical bearings of these conditions will be considered in the latter 

 part of this article. The depleted semipupse of the instahilis females and 

 males, which like the phthisergates are incapable of further development, 

 may be called phthisogynes and pJtthisaners re.sj)ectively. 



In this connection the question naturally suggests itself: are the inter- 

 mediates between the instahilis workers and the soldiers due to similar 

 depletion in their semipupal stages? In other words, do the intermediates 

 arise from soldier semipupse that have been partially exhausted by Orasema 

 larvse prematurely torn from their hosts by the workers ? I am inclined to 

 answer this question in the negative, for reasons to be given in the se([ucl. 



After finding 0. viridis so common in the nests of Ph. instahilis I was 



