Carpenter — Relationshi2)S heticeen ChiHsoa of Arthropoda. 327 



And an examination of the relationships of those classes and their 

 orders shows that the foimer alternative has much evidence in its 

 favour. 



The morphological studies of Kingsley ('88) andPocock('93A), and 

 the embryological researches of Heymons ('01), have established beyond 

 any reasonable doubt that the " Class Myriapoda " must be abandoned, 

 the Centipedes (Class Chilopoda) being more nearly related to the 

 Insects than to the Millipedes (Class Diplopoda). The Centipedes 

 agi'ee with the Insects in the simple segmentation of the body, in the 

 lateral position of the spiracles, in the anastomosing tracheal tubes, 

 and in the posterior position of the genital openings ; while the 

 Millipedes exhibit for the most part a fusion of the segments in 

 couples, so that each apparent segment carries two pairs of legs, the 

 spiracles are ventral in position, the air-tubes are unbranched and do 

 not anastomose, and the genital openings are far forward on the third 

 body segment. 



If, then, it is believed that Insects and Centipedes on the one hand, 

 and Millipedes on the other, have diverged from some common 

 ancestral stock, it is natural to inquire whether any living form can 

 suggest approximately what that stock may have been like. The 

 only animals that combine some of the divergent characters of Insects, 

 Centipedes, and Millipedes, are the Scolopendrellidse, now usually 

 regarded as a distinct class, called, on account of their annectant 

 characters, the Symphyla (Eyder, '80). These small, frail, somewhat 

 degenerate creatures, show the series of similar, simple limb-bearing 

 segments characteristic of Centipedes, the forwardly-situated genital 

 aperture as found in Millipedes, and a number of body-segments 

 identical with that occurring in Insects. Their chief point of 

 specialization is the curious inequality and displacement of the 

 tergites. l^o surprise need be felt that some students of their structure, 

 like Packard ('98), regard them as representing the ancestral stock of 

 Insects; others, as Grassi ('85), that of Centipedes and MilHpedes. But 

 if we are willing to accept the view, admitted as possible by Lang 

 ('91), that most living Centipedes and Millipedes have become what 

 they are by an increase from the number of primitive seg-ments, there 

 is no reason why we should not, with Haase ('86) and Pocock ('93a), 

 regard the Symphyla as approximate to the common ancestor of the 

 Insecta and the Chilopoda on the one hand, and of the Diplopoda 

 (including the Pauropoda) on the other. Haase particularly suggests 

 that the common ancestors of the three great Tracheate classes had as 

 many segments as Scolopendrella. 



