54 DK. G. C. BOUKNE ON THR KANINIDiE : 
relationship with the Raninidse. De H:ian laid great stress on this point of 
resemblance between the Loucosiidseand the Eaninidae, but a detail exiimina- 
tion shows that the result is arrived at in a different nmnner in the two 
families and must be regarded as a case of homoplasy rather than homogeny. 
Ill all the Raninidte the free end of the sternal plastron is formed by the 
more or less prominent lanceolate median projection of the tenth (chelipedal) 
sternum. The ninth sternum is narrow and projects upwards at an angle 
of 130*^ to 135° above the lanceolate ]irojection of the tenth sternum; The 
coxse of the third maxillipeds are ;irticulated to the hinder end*of the narrow 
ninth sternum, and therefore close to the median line, and they effectively 
exclude any part of the sternum from participation in the formation of the 
broad plastron (see fig. 27, ix). Further, in the Eaninidse, the connection 
between branchiostegite and plastron in front of the cheliped is effected by 
lateral outgrowths of the tenth sternum which meet the branchiostegnl 
margin, not by any marked ingrowths of the latter. In the Leucosiidaa 
(I have studied the arrangement in Ilia nucleus, Philyxia undecimspinosa, 
and Philyra Icevis) the ninth sternum is short antero-posteriorly, but broad 
and forms the anterior end of the plastron. The cosaa of the third inaxilli- 
peds are articulated to its outer ends and are therefore far apart. The tenth 
sternum does not give off any prominent outgrowths in front of the chelipeds, 
but is united to an inwardly directed process of the edge of the branchio- 
stegite on each side, this process being the most anterior of a series of 
similar downgrowths of the branchiostegal margin which pass between the 
coxae of the pereipods and effect the very close and elaborate fit of branchio- 
steoite to coxaj already referred to. It may be mentioned here that 
A. Milne Edwards and Bouvier (28) have described a shallow pterygostonnal 
gutter closed in below by the third maxillipeds in Cydodorippe. This 
gutter occupies the same position as the well-known pterygostomial inhalant 
canal of the Leucosiidse, and the similarity of the structural appointments is 
so oreat as to suggest the derivation of the Leucosiidaa from a Gyclodorippid 
rather than from a Raninid ancestor. 
Finally some importance is to be attached to an apparently minor feature. 
In all the Oxystomatous Crabs the floor of the exhalant branchial canal is 
formed by an expanded spoon-shaped operculum furnished hj the first 
maxilliped. In the Raninidje (see figs. 31, 45, 50, & 53) both the exopodite 
and endopodite are expanded and modified, and co-operate in forming the 
operculiform floor in question, but it is the e.xopodite which is the longest, 
most expanded, and the most effective agent in forming the operculiform 
floor of the anterior part of the exhalant canal. This modified exopodite 
never bears a flagellum. In the Leucosiidse, the Calappinse, and the 
Matutinae (Garstang is in error in stating that the exopodite is operculiform 
iwMatuta hanksii)\t is the endopodite of the first maxilliped that is elongated. 
