322 Proceedings of the Royal Irish Academy. 



these form an assemblage of characters quite unique in the Animal 

 Kingdom. And any attempt to explain their appearance in the various 

 classes of Arthropods as the result of convergent evolution must raise 

 far more difficulties than it can solve. 



But the principal view maintained in this essay is one which was 

 suggested nearly sixty years ago by Huxley ('58), and which has been 

 abeady published in outline by the present writer ('99) — that Crus- 

 taceans, Arachnids, and Insects agree closely in the primitive number 

 of their segments. It has been generally believed that the fixed and 

 definite number of segments found in the Malacostraca has been derived 

 by reduction from the numerous segments of Branchiopodan ancestors ; 

 that the definite segmentation of Insects has arisen by condensation 

 from some primitive richly-segmented Myriapod. But, as Huxley 

 wrote in 1858, "I venture to think it a matter of no small moment if 

 it can be proved that a Lobster, a Cockroach, and a Scorpion are 

 composed of the same primitive number of somites." If this be the 

 fact, we have well-nigh demonstrative proof that the classes to which 

 these three animals belong are truly akin to each other, and that their 

 allies with very many segments represent abnormal developments. It 

 is almost impossible that a reduction to exactly the same number of 

 segments in three classes of similarly-formed animals could have been 

 independently produced. 



It is proposed, therefore, to compare the orders of the various 

 classes of Arthropods so far as it may be necessary to arrive at a con- 

 ception of the most primitive members of each class, with especial 

 reference to their segmentation. Then the various classes as a whole 

 can be profitably compared and their affinities discussed. The writer 

 would express his special indebtedness to a most suggestive but 

 strangely neglected paper by Hansen ('93). If students of the 

 Arthropoda would follow his example, and compare diligently Arthro- 

 pods with other Arthropods, before comparing them with specialized 

 Annelids, our phylogenetic studies might advance with greater 

 assurance and less controversy than at present. 



Nature af the most primitive Insects. 



Any lengthened discussion of the relationship between the various 

 Orders of Insects is needless in view of the almost universal agree- 

 ment among entomologists in regarding the Thysanura as the most 

 primitive of living groups. Bernard (Hutton, &c., '97) has, indeed, 

 recently revived the suggestion that the caterpillars of Lepidoptera 

 and Sawflies are to be considered as representing the ancestral stock 



