18 THE CLASS OF I.\>KCT>. 



mere are modified to form the parts supporting the sting alone. 

 The external opening of the oviduct is always situated hrl ween 

 the eighth and ninth segments, while the anal opening lies at 

 the end of the eleventh ring. So that there are really, as 

 Lacaze-Dutbiers observes, three segments interpo>ed between 

 the genital and anal openings. 



The various modifications of the ovipositor and male organ 

 will be noticed under the different suborders. 



THE STRUCTURE OF THE HEAD. After studying the com- 

 position of the thorax and abdomen, where the constituent 

 parts of the elemental ring occur in their greatest simplicity, 

 we may attempt to unravel the intricate structure of the head. 

 We are to determine whether it is composed of one, or more, 

 segments, and if several, to ascertain how man}', and then to 

 learn what parts of the typical arthromere are most largely 

 developed as compared with the development of similar pails 

 in the thorax or abdomen. In this, perhaps the most difficult 

 problem the entomologist has to deal with, the study of the 

 head of the adult insect alone is only guesswork. AVe mu>t 

 trace its growth in the embryo. Though many writers consider 

 the head as consisting of but a single segment, the most emi- 

 nent entomologists have agreed that the head of insects is com- 

 posed of two or more segments. Savigiry led the way to these 

 discoveries in transcendental entomology by stating that the 

 appendages of the head are but modified limbs, and homol- 

 ogous with the legs. This view at once gave a clue to the 

 complicated structure of the head. If the antenna' and biting 

 organs are modified limbs, then there must be an elemental 

 segment present in some form, however slightly developed in 

 the mature insect, to which such limbs are attached. Hut the 

 best observers have differed as to the supposed number of such 

 theoretical segments. Burmeister believed that there were two 

 only; Carus and Audouin thought there were three; McLeav 

 and Newman four, and Straus-Durckheim reeognixed seven. 

 From the study of the semipupa of the Humble-bee (Bombus) 



support of the sting; e, the support of the stylet (i). R, the anus; O, the outlet of 

 the oviduct. The seventh, eighth, and ninth sternites are aborted. From Lacme- 

 l>uthiers. 



