BRACHYURA. 



69 



We close these remarks on the Brachyura with some observations on 

 the classification of De Haan. The exalted merit of his labours seems 

 to forbid criticism on a matter of arrangement. Yet classification is 

 of the highest importance; since it should exhibit the progress and 

 condition of the science, and present in a single view its general 

 truths. The defects in his system have arisen from a wrong prin- 

 ciple, as we think, in its very foundation ; and the results of the prin- 

 ciple are seen, not only in many of the larger divisions, but also in 

 laying down his genera. The objectionable point referred to, is his 

 giving paramount importance in classification to the maxillipeds, the 

 true value of which has been already explained. In the Leucosoidea 

 they have a strongly characteristic form, and this group is readily 

 characterized by reference to them. Yet even in this they may be 

 treated with undue importance. The true distinction of the Leuco- 

 soidea is exhibited not in the inner maxillipeds, but in the character 

 of the efferent branchial channel : the peculiarity in this important 

 function is the fundamental difference separating these from other 

 Brachyura. The elongate form of the inner branch of the first maxil- 

 lipeds is rather an indication of the difference, than correctly the 

 great point of difference. 



The graiid divisions of the Brachyura in De Haan's system are as 

 follows: A. BRACHYGNATHA, including I. Cancroidea; II. Maiacea; 

 III. Dromiacea; IV. Trichidea: B. OXYSTOMATA, including I. Dorip- 

 pidea; II. Calappidea; III. Matutoidea; IV. Leucosidea: and the 

 Cancroidea are subdivided into Corystes, Cancer, Portunus, Ocypoda, 

 and Grapsus groups. 



In the primary divisions, the Oxystomata (curtailed of the Corystes 

 group, placed among them by Edwards, McLeay, and other authors) 

 make a natural section among the Brachyura : they are the Leuco- 

 soidea of the system adopted. The character of the efferent channel 

 affords a strongly-marked division. But in making out a natural clas- 

 sification, it is necessary to inquire whether there may not be other 

 distinctions equally important ; whether, among those species that are 

 alike in the efferent channel, there may not be points of difference 

 fully as essential, thus requiring the institution of other groups of like 

 importance with that of the Leucosoidea. The arrangement which has 

 been proposed contains our views on this point. The Maia, Cancer, and 

 Grapsus types, have each important characteristics, based upon points 



18 



