4 GO Mr. L. A. Borradaile on the 



of typical genera Boas came to the conclusion tliat the 

 existing arrangement was unnatural in that it contrasted the 

 Brachyura and Anomura — single branches of the Decapoda 

 — with a heterogeneous assemblage (the Macrura) consisting 

 of the whole of the rest of the tree, some branches of which 

 are more nearly related to the Bracliyura and Anomura than 

 they are to the rest of tlie ]\Iacrura. He accordingly proposed 

 to divide the order into two suborders — the Reptantia, 

 containing the Crabs, Anoraurous forms, Thalassinidea, 

 Homaridea (Nephropsidea), Scyllaridea, and Eryonidea; and 

 the Natantia, containing the Penseidea (including Steno- 

 pidje) and the Caridea *. The names of these suborders 

 indicate the main difference by which, on the whole, they are 

 separated ; a more accurate diagnosis will be found below. 



4. There can be little question of the correctness of Boas's 

 view that the members of his Reptantia form a natural 

 group. The characters that they hold in common are too 

 numerous and too specialized to admit of doubt on this point. 

 But it by no means follows that the same is true of the 

 Natantia. Unless it can be shown that the former group 

 arose from the Decapod stem before the separation of the 

 forms which constitute the latter, Boas's classification will be 

 open to the objection that he raised against Latreille's — that 

 is to say, it will be based, not on the divergence of two groups, 

 but on the elevation of a branch to the same rank as the 

 parent-stem. This, as it happens, is precisely the impression 

 conveyed by the tree figured by Boas on p. 27 of his paper. 

 The following considerations, however, serve to show that 

 Boas's tree is wrong and his classification true. 



A search for the most primitive group of the Decapoda 

 leads, beyond all doubt, to the Penseidea. This is seen, (1) in 

 their primitive life-history, recalling that of the Euphausiacea ; 

 (2) in the peculiar copuiatory armature of the male, which 

 suggests tiie same relationship ; (3) in the small number of 

 special features, unshared by other Decapoda, which the 

 group posgesses, and the number of characters that they have 

 in common with one or other of the remaining groups — 

 thus, with the lower Reptantia they share the shape of the 

 first three pairs of legs, which are fairly alike except some- 

 times in size, and all chelate t, whereas those of the Caridea 

 often differ much and their third pair is never chelate, the 

 structure of the maxillipeds, which lack special modifications 

 found in the Caridea, and the absence of the Caridean bend 

 iu the abdomen, while they share with the Caridea all those 



* Eukyphotes of Boas. 



t Except in the aberrant Sergestidae. 



