1 88 WILLIAM M. WHEELER. 



Forel, on the other hand, advances the following reasons for 

 regarding the Cerapachyinae as true Ponerinae. While they un- 

 doubtedly exhibit traits which ally them with the Dorylinae, their 

 habitus is, nevertheless, decidedly Ponerine. The little that is 

 known of their habits certainly indicates that they live in small, 

 stationary colonies like the Ponerinae, instead of populous, nomadic 

 colonies like the Dorylinae. The queens, moreover, are so nearly 

 of the same size as the workers as to preclude anything like the 

 great fecundity of the queens of Dorylus and Eciton. The Cera- 

 pachyinae, too, have short legs of such a structure as to indicate 

 a slow gait and more sedentary habits. The workers of the 

 Cerapachyinae have ordinary facetted eyes, whereas those of the 

 Dorylinae are absent or ocelliform, while the atrophied eyes of the 

 Ponerinae have a very different structure. 1 The conditions of the 



FIG. 5. Apterogyna olivieri. Female, after Emery. 



pedicel in the males, females and workers of the Cerapachyinae 

 are correlated as they are in the Ponerinae, and do not exhibit the 

 differences seen in the Dorylinae between the worker on the one 

 hand and the male and female on the other. The wingless con- 



1 The distinction to which Forel calls attention is worthy of histological study. I 

 am inclined to think, however, that it may be a distinction without a difference. I 

 have recently sectioned a number of pupa: of Eciton schmitti and find that the ocelli- 

 form lateral eyes are really very much atrophied compound eyes, too much atrophied, 

 in fact, to be at all functional as visual organs. The retinal hypodermic, which is 

 somewhat thickened under the convex lens, shows indistinct but unmistakable traces of 

 ommatidia. The optic nerve is very short and not connected with the brain. It ends 

 freely in a blunt point a short distance from the ommatidial layer. This is interesting 

 as proving tJi at the visual fibers must arise in tlie retina and grow towards the brain 

 and not in the reverse direction from cells in that portion of the brain known as the 

 optic ganglion. If there is a distinction between the abortive eyes of the DorylinEe 

 and Ponerinae it would seem to be that in the former subfamily the ommatidia disap- 

 pear both by fusion with one another and by reduction in number, while in the latter 

 the number of ommatidia is gradually reduced without fusion. 



