hearing on the origin of Insects. 151 



After prolonged study of numerous dissections and 

 preparations, I have arrived at the conclusion that the 

 mandibles of Blattidce are compound structures, each 

 made up of three (or four) such joints as are to be seen in 

 Macliilis ; and I believe that the process of modification 

 by wliich the head and mandibles of such an insect as 

 Machilis have become converted into those of Blutta may 

 have been somewhat of this nature ; the basal joints have 

 gradually shortened and coalesced with one another until 

 little more was left of them than the ball-shaped condyle; 

 pari passu with this change the Avails of the head have 

 gradually completed themselves behind the shortening and 

 retiring basal joints of the jaws, so as eventually to form a 

 *gena;' and finally, a 'ginglymus' articulation was formed 

 by a process sent off from the front of the head to meet 

 a cupped process of the base of the mandible* of the 

 same side. 



3§. Are the mandibles of insects and myriopods, like 

 the jaws of Peripatus, modifications of walking-legs ? I 

 think not. In the cockroaches, a notch at the extremity 

 of the mandibles on the inner side sharply divides the 

 crushing and cutting portions of those appendages from 

 one another ; in the embryos a curvilinear sutural mark 

 extends from the bottom of this notch, and separates the 

 two parts off from one another still more definitely; this 

 is seen better marked in all the species of Lepisma (see 

 Lubbock's figure of the mandibles of L. sacckarina), a 

 form in many respects intermediate between Blatta and 

 Machilis, and in some of the species of which the apical 

 portion of the mandible closely resembles that of Blatta. 

 In Machilis, the notch becomes a deep fork, widely sepa- 

 rating the two pai'ts, and from its bottom there extends 

 basewards a distinct inflection of the integument. This 

 peculiar feature in the structure of the apical joint seems 

 to me only explicable on the hypothesis that the mandibles 

 of these insects and of myriopods have resulted from the 

 direct modification of such a biramous appendage as is 

 seen in the earliest {Naiiplins) condition of many crus- 

 taceans, the two or three basal joints attached to the head 

 representing the protopodite, and the molar and incisive 



* The 'ginglymus' is still incomplete in the ripe emlnyos with which 

 I have worked, and 1 do not think it is com])l('ted till after birth, probably 

 not till the tirst extra-ovuhir moult has taken place and removed all 

 traces of segmentation in the mandibles. The embryonic development of 

 the mandibles in lilatta repeats the historic development perfectly. 



