CLASSiriCATlON OF THE MAIOID CEUSTACBA. 641 



tion may be traced from Micippa with well defined orbits, to tbe 

 remarkable g&nus, JPicroGerus,iu. wbicli the true orbits are as little 

 developed as iu many Inachinse. 



In cases such as these, I believe it is often better to preserve 

 tlie natural sequence of the genera, though in so doing one must 

 slightly overstep the literal definition, than, by too strict an 

 adherence to the definition of the group, to separate forms which 

 in all characters save one may be nearly allied. Nature imposes 

 no artificial limits ; and not even an arbitrary distinction will in 

 all cases avail to separate kindred forms *. 



Legion II. Parthenopinea. (Parthenopinea, Dana at auctorum). 

 Basal antennal joint very small, and embedded with the next joint in 

 tlie narrow hiatus between the front and inner suborbital angle ; the 

 infraocular space being mainly occupied by the lower wall of the 

 orbit. 



Family IV. Parthenopid^, Characters of the section; — This group 

 corresponds in the main with M. -Edwards's Parthenopiens ; but the cha~ 

 racters are modified to include several genera which agree with those 

 known to Milne-Edwards in tlie structure of the orbits and antennae^ but 

 ditFer in the carapace and anterior legs. Moreover I follow De Haan in 

 excluding iJ?<ryraome (which really belongs to the Maiidse) and including 

 (Ethra (which is placed by Milne-Edwards in a separate section of the 

 Canceriens — Canceriens cryptopodes). 



As already stated, the Parthenopinea are very distinct as a 

 group from the rest of the Oxyrhyncha. Perhaps their nearest 

 affinities in that direction are with Inachus through Inaclioides. 

 The triangulate form of the carapace, with its strongly marked 

 depressions separating the different regions, is the same, and the 

 slender basal antennal joint. In Inaclioides the rostrum is simple, 

 as in Partlienope and Lamirus. 



In the plates that accompany this paper I have figured 

 what may be regarded as typical examples of the principal 

 modifications in the structure of the orbital and antennal region 

 throughout the Oxyrhyncha, wherein may be traced the gradual 



^'' Dana, for example, separated the Cancroidea into two parallel groups 

 (Cancridas and Eripliiidas), characterized respectively by the presence or absence 

 of a ridge on the endostome defining the efferent branchial channel; but in 

 Xanthodius, a genus since described by Stimpson, this ridge is rudimentary, and 

 this genus may be referred either to the vicinity of CMorodius in the former, or 

 Ozius in the latter group. Similar intermediate forms occur between Actaea 

 and Actcpodes, genera belonging respectively to the xDarallel series Xanthine and 

 Chlorodiina3 in the family Cancindte, 



