Dec, 1916.] Relationships of Apterygotan Insects. 275 



self, and as long as human nature remains as it is, we will have some 

 who are "lumpers," and some who are "splitters." Of recent ento- 

 mologists, however, those who are so " ultra-conservative " as to be 

 unwilling to admit of more than one Apterygotan order, the " Aptera," 

 are in a decided minorit}^ and one is led to suspect that they have 

 either not made a detailed study of the insects in question, or an 

 overweening desire for " unity and simplicity " has caused them to 

 cling to a system of classification but little removed from the ancient 

 Linnean one ! 



As time goes on, and as the structural details of the Apterygotan 

 insects have been more carefully worked out, it becomes more and 

 more evident that the general superficial resemblances between many 

 of the insects classed together in the older -groupings, are far out- 

 weighed by the differences in their morphological details. I would 

 therefore fully agree with Handlirsch and his followers (such as 

 Brues and Melander, Escherich, etc.) in their contention that the 

 heterogeneous collections of insects comprising the old " orders " 

 should be re-grouped into several distinct orders whose members are 

 more closely related. 



In the following key, the greater part of the insects referred to 

 belong to distinct orders although I would not maintain that this is 

 true of all of them, as will be discussed later. In accordance with 

 the quite widely accepted custom, I have designated the Apterygotan 

 orders by terms ending in " -ura " (e. g., Protura, Rhabdura, Thy- 

 sanura, etc.), rather than by Handlirsch's terms ending in " -oidea," 

 since the latter termination is preempted by groups of superfamily 

 rank (e. g., Muscoidea, Ichneumonoidea, etc.)- Handlirsch's method 

 of using a well-known family name to typify the group as a whole, 

 however, has much to recommend it, since it is self-explanatory, I 

 have therefore made use of a modification of this method in the 

 following key, merely for the sake of convenience, using the termina- 

 tion "oides" (rather than the preempted termination "oidea") in 

 connection with the name of a well-known family, thereby immediately 

 calling to mind the representatives of the group as a whole. 



1. Styli and segmented caudal filaments or terminal forceps present 2 



Styli and segmented caudal filaments or terminal forceps wanting 7 



2. Mouthparts not retracted into cavity of head 3 



Mouthparts retracted or concealed in cavity of head 5 



