IQO WILLIAM M. WHEELER. 



various forms of females to be met with among the Formicidse 

 are already foreshadowed in the small and very primitive group 

 of the CerapachyhiEe. Though we have no knowledge of the 

 females of several of the genera, we may recognize no less than 

 four different female forms : 



1. The female of Acantlwstichus, which, as Emery has shown 

 ('95), is decidedly Dichthadia-\\k.z, i. c., unmistakably like the 

 huge blind and wingless females of Doiylns and Eciton. This 

 female is considerably larger than the largest workers of the 

 colony as shown in Emery's figures of A. qnadratus which I 

 have reproduced in outline (Fig. 4, a, b and c}. 



2. Normal winged females like those of most genera of For- 

 micidae but more similar to the workers in size and structure. 

 These females are known to occur in the genera Lioponcra, Cera- 

 pachys and Sphinctomyrmex. 



3. The female of Cerapachys peringncyi from South Africa 

 (Fig. 3, a\ According to Emery ('95) this form is wingless and 

 not much larger than the worker (Fig. 3, b\ which it closely re- 

 sembles in structure. It may be designated as an ergatoid female 

 and is not unlike the ergatoids occasionally found in species of 

 Ponera (P. coarctata, P. opaciccps, etc.). 



4. The female represented by the above described Cerapachys 

 augustce from Texas. This form is wingless but in thoracic 

 structure resembles the winged females of the Ponerinse in gen- 

 eral. It is but little larger than the largest workers though pos- 

 sessing well-developed eyes and ocelli. 



This " morphological restlessness " in the structure of the 

 females of so small a group of genera as the Cerapachyinae is, 

 perhaps, significant as the phyletic source to which the different 

 female forms of all the subfamilies of ants are to be traced. We 

 may look upon the Dichthadia-Yike queens of the Dorylinae as a 

 further development of the conditions exhibited by Acant/io- 

 stichns, and the ergatoids, which crop out sporadically among the 

 Ponerinae and the higher subfamilies, may, perhaps, be regarded 

 as cases of reversion to females of the type of Cerapachys pcr- 

 ingucyi and the Mutillidae (Aptcrogyna, c. g.}. The pseudogynic 

 forms of the more specialized ants (Formica, Cainponotiis) re- 

 semble the female of C. augnstce. Finally, the winged females 



