325 



being absolutely impossible for any one man, or even for a few men, to 

 cover all the ground in a satisfactory manner. 



The key presented herewith is based upon species in the collection 

 of this Laboratory with one exception, and gives a synopsis of the 

 characters which I have used in separating them. There are in the 

 series examples of the following families : Bomhyliidce, Mydaidcu, 

 Asilidcr, and Therez'idcc. The pupse of all these families, as far as I 

 nm aware, bear a strong resemblance to those of Tahanidcc, but the 

 pupae of the latter differ noticeably, as far as I have seen, in having no 

 long thornlike cephalic processes, the protuberances being short and 

 not heavily chitinized. The use of the cephalic armature by the 

 Asilidcr and their allies in digging their way out of the ground necessi- 

 tates that those organs be strong, as the species often must burrow 

 through rather hard dry soil, while the Tabanidcc, being for the most 

 part species which live in water or in moist situations, do not require 

 such powerful organs to assist in their emergence from their pupal 

 habitat. The abdomen in all the species included in the key is armed 

 upon each segment with a median transverse series of thorns, or hairs 

 and thorns in alternation. This feature is somewhat similar to that 

 presented by the pupse of the Tahanidce, but the apical segment differs 

 considerably in the two groups, while the species of Tahanus have 

 usually the transverse thorns or hairs in a double row, the anterior one 

 consisting of shorter and stronger thorns than the posterior one. 



The pupae of the only two species of Cyrtidcc that I have seen, one 

 of which is described in the present paper, dififer very considerably 

 from those of the group included in the kev. and also from the Taban- 

 idcr, in having the head and abdomen without thorns or bristles, and 

 the abdominal spiracles reduced in number, there being not more than 

 five pairs. 



Key to Pup^ 



1. Head with but two thorns, abdominal spiradcs conspicuously ele- 



vated, thornlike Psilocephala liwmorrhoidalis (p. 334). 



— Head with more than 2 thorns, abdominal spii-acles rcniform, not 

 conspicuously elevated 2 



2. Doi'sal abdominal segments with a transverse series of short flattened 



thorns on middle of each segment, and alternating long, slendei-. 

 slightly curled hairs ; sometimes the short thorns are bent iipward 

 at right angles at bases and apices* (Bomhyliida;) 3 



*In all cases that I have seen these short thorns have the appearance of beinfr 

 attached to the abdomen, rather than of being a part of it as is the case in the other 

 pupse dealt with here. 



