14 University of Michigan 



decided the latter was the more probable. In trying to fit ihe 

 hooks on the hind margin of the prothorax it did not occur 

 to me that the apex of lo would rest on the dorsum of the 

 head (an opinion I had expressed to Dr. Walker in a letter) ; 

 I thought it would have to lie behind the head^ and I found 

 it was too long for this. If the dorsum of lo is placed on the 

 prothorax of the female with the hooks behind the posterior 

 margin, the apex of lo is too low to rest on the head. If the 

 apex is elevated sufificiently to rest on the female's occiput, the 

 dorsum of lo and the prothorax are widely separated, and the 

 apical dorsum of segment 9 may possibly be closely against the 

 middorsal thorax carina of the female. My idea is that the 

 hooks of 10 engage the occiput of the female, the dorsum of 

 the segment resting on top of the head. This position involves 

 no difficulties in regard to contact between parts of the male 

 abdomen and the thorax of the female, but has the disad- 

 vantage that the parts of the female grasped by the male are 

 as fixed as those of the latter. An attempt by the female to 

 free herself by a forward and downward movement of the 

 head could however be resisted by a forward and upward pull 

 of the male abdomen. The occiput of the female is for the 

 most part rounded but at each postero-lateral angle there is a 

 short projecting tooth, separated from the rear of the eye by 

 a notch. These teeth or notches are the right distance apart 

 to be grasped by the hooks of the male, and they are wholly 

 absent in the male. This is the chief reason why I incline 

 towards this theory rather than the other one. It does not 

 however account for the snout-like projection of the apex of 

 JO which seems unnecessarily long for either method, though, 

 according to your plan the long apex might allow a little play 

 of the female's head beneath it without permitting the latter 

 to escape." 



