230 THE ENTOMOLOGIST. 



be secondary, and draws a comparison between the small head, 

 lack of antennae, and prolongation of the head ganglion into the 

 prothorax as found in the Protura, and a similar combination in 

 many insect larvae. He does not attach much importance to the 

 mouth-parts for deciding the systematic position, but believes 

 that the anamorphosis and the large number of abdominal 

 segments must be primitive, and also the abdominal appendages 

 and the position of the genital opening, which separate them at 

 once from the CoUembola. He proposes to consider them as a 

 class Myrientomata, between the Myriapoda and the Insecta. 



Prell [10] regards the anamorphosis as the most important 

 and primitive character, while the number of abdominal seg- 

 ments, the position of the genital opening, and the abdominal 

 appendages, although also primitive, are of less importance. 

 The lack of antennae and cerci, on the other hand, are secondary, 

 as also is the reduction of the tracheal system in the Aceren- 

 tomidae. He proposes the following arrangement : — 



Class Insecta. 



Subclass Anamerentoma (with anamorphosis). 



Order Protura (with twelve abdominal segments ; no antennae, &c.) 

 Subclass Holomerentoma (with no anamorphosis). 

 All the other Insecta. 

 This classification seems to be the most sound yet put 

 forward, as it points out the undoubted affinities which the 

 Protura have with the Insecta, but at the same time recognizes 

 the fact that they differ from the other orders of Insecta to a 

 greater extent than these do among themselves. 



Prell does not think that the Protura are to be considered as 

 in anyway the direct ancestors of the Holomerentoma, but as a 

 group which branched off from the ancestors of the Insecta at a 

 very early date, long before the origin of wings, and which has 

 since become largely modified and specialized. 



The further subdivision within the order presents no great 

 difficulties at present. The species fall naturally into two 

 families, containing three genera, with the chief characteristics 

 as follows : — 



Family I. — Acerentomid^. Tracheal system absent. Second and 

 third abdominal appendages consisting of a single 

 joint. Dorsal pleurites of the eighth abdominal 

 segment with comb-like posterior margin. 

 Genus 1. — Acerentomon. Labrum produced anteriorly into a long 

 process. Maxillary palps l-jointed. Labial palps 

 3-jointed. Mandible terminating in a very long 

 pointed style. 

 Three species. — A. doderoi, Silvestri [18] ; A. micro- 

 rhinus, Berlese [5] ; A. affinis, Bagnall.* 



* The description of this species, hitherto [1] only mentioned by name, 

 will, I believe, appear in the Ent. Mo. Mag. for August, 1913. 



