72 CRUSTACEA. 



inward at base, are so peculiar, and so evident an exhibition of a re- 

 lation to the Hippidea, that we naturally give it a prominent place 

 among the characteristics of the Corystoid Crustacea. We have re- 

 marked in a former paragraph upon its being a mark of degradation. 

 We should, hence, exclude Platyonychus, and the allied, from direct 

 association with Corystes, notwithstanding the similarity in the outer 

 maxillipeds. Indeed, in all their characters they are so closely like 

 many Cancroid species, that we find no means of distinction. If then 

 they are Cancroid in character, and not Corystoid, they must be ar- 

 ranged, either in the Portunus group as a separate subdivision of it, or 

 they should form a distinct division among the Cancroidea. We incline 

 to make them a distinct division near the Portunido3. The genus Piri- 

 mela is placed by De Haan in the Corystes group, for the same reason as 

 Platyonychus, although essentially Cancroid in character. The genus 

 Cancer has almost equal title to a place there, and on like grounds. 



The importance allowed to the outer maxillipeds has led to other 

 unnatural associations among his Cancroidea. The .Cancer group 

 contains species that have the general habit, branchia3, and other 

 characters of the Grapsidas. I refer to the Gonoplax family, which 

 is rightly placed with the Grapsus group by Edwards. They have 

 the fourth joint of the outer maxillipeds articulated with the inner 

 apex of the third joint, and this is the only character that would ally 

 them with the Cancer division, rather than with Grapsus. 



The distinction between Ocypoda and Grapsus, depending on 

 whether the fourth joint of the outer maxillipeds is articulated with 

 the summit, or with the outer apex of the third joint, is exceedingly 

 difficult of application, and does not in all cases, lead to natural asso- 

 ciations. Cardisoma and Uca are arranged by De Haan in the Ocypod 

 group, and Gecarcinus in the Grapsus group. 



The difficulties from relying so implicitly on the maxillipeds are still 

 more strongly seen in the generic distinctions as given by De Haan. 

 There are cases in which the distinctions are good; but they are 

 used to such an extent as to be in the main bad. The genus Xantho 

 is said to have the third joint of the outer maxillipeds subquadrate, 

 a character which would exclude species in which it is oblong : and 

 it embraces species that are not true Xanthos, if judged by the charac- 

 ter of the antennae, organs of higher importance than the maxillipeds. 

 The genus Eudora, containing the Rupellia tenax of Edwards, is so 

 characterized as to include a Xantho; that is, the character of the 

 orbit of the Kupellia, which has no similar example except in the 



