1907.] Wheeler, The Polymorphism of Ants. 19 



number of specimens in alcohol. These were described in a paper pub- 

 lished during the autumn of 1901.^ In this paper the huge workers were 

 called macroergates, but for reasons presently to be given they may be more 

 aptly designated as mermithergates. 



Plate III, Figs. 37-40 represent, drawn to the same scale, the dorsal 

 view of the soldier and worker, which in Ph. commutata are not connected 

 by intermediates, and the dorsal and lateral views of a large mermithergate. 

 The length of normal w^orkers does not exceed 3 mm. INIany of them are 

 scarcely more than 2.5-2.8 mm., which was the length of Mayr's type speci- 

 mens from Florida. The largest mermithergates, however, measured fully 

 5 mm., while the smaller ones varied from 4 to 4.5 mm. Thus the volumes 

 of the normal workers and the extreme mermithergates, had they been of 

 the same shape, would be in the ratio of 27 to 225, but the gasters of the 

 latter were so enormously distended that the ratio must have been 27 to at 

 least 300. In other words, the large mermithergates were some twelve 

 times as large as the normal workers. They were even larger than the sol- 

 diers, which measure about 4 mm., though in this case, owing to the great 

 size of the head in the latter caste, the difference is less conspicuous. The 

 largest mermithergates differ from both soldiers and workers in usually 

 possessing three ocelli, of which the anterior is the largest, and resemble the 

 soldiers in the structure of the thorax (PI. Ill, Fig. 39). 



Examination of one of these extraordinary individuals, even with a good 

 pocket lens, reveals the cause of the great enlargement of the gaster. 

 Through its distended intersegmental membranes the coils of a parasitic 

 worm may be distinctly seen. My friend Professor T. H. ^Montgomery, to 

 whom I sent a few of the mermithergates, writes me that the parasite is a 

 species of Merml.s. Its exact location among the ant's viscera, i. e., whether 

 it occupies the lumen of the enormously distended ingluvies, or lies in the 

 body cavity outside of the alimentary tract, is not easily determined. Froni 

 careful dissection of a single large specimen (the one represented in PI. II, 

 Fig. 39) I concluded that the Mermis lies within the ingluvies, or crop. In 

 this specimen the head of the parasite extended forward through the post- 

 petiolar and into the petiolar segment, and thus occupied the attenuated 

 neck of the crop and the most favorable position for securing the ingurgitated 

 food of its host. The fat-body and reproductive organs seemed to have dis- 

 appeared completely and the walls of the enormously distended crop were 

 practically in contact with the walls of the gaster. The large mermithergate 

 shown in Figs. 39 and 40 contained only a single closely convoluted Mer- 

 mis, which was fully 50 mm. long, or ten times the length of the ant. One 



' The Parasitic Origin of Macn;6rgate.s among Ants. Amer. Naturalist, XXXV, Nov. 1901, 

 pp. 877-866, 1 fig. 



