A STUDY IN CARCINOLOGY. 67 
The l:ist pair of pereioi)0(1ir, though much reduced in size, is sufficiently 
liberally provided with hairy fringes to form an effective covering for the 
roof of the water-chambers enclosed between the other pereiopods and the 
thoracic flanks. The epimera forming these flanks, however, are not so 
deeply concave as in Notopoides, and the posterior branchial orifices, though 
distinct enough, are smaller and obviously of less functional importance. 
The dactylus of the last pair of pereiopods is small, elliptical, and clearly 
better adapted for natation than for digging. 
Before entering into a discussion of the transition from digging to 
swimming limbs it should be premised that in the pereiopods of all the 
llaninidse the articulation of the dactylus to the propodus is singularly like 
that of the tarsi of the whirligig-beetle Gyrinus, familiar to all naturalists 
from the illuminating description of Miall. Generally, in the Raniuidse, the 
carpus is moderately long, triangular in shape, with an extensive basal 
articulation for the propodus. The propodus is a short and broad oblong, 
more or less flattened, with a very small articulation for the dactylus close 
behind its postero-extei-nal angle. The dactylus is very variable in shape. 
Generally in the first two pereiopods it is lanceolate and obviously efficient 
in digging, but in the last two pereiopods something between crescentic and 
cleaver-shaped like an oriental " kukri." In either case it is so articiilated 
by a narrow pedicle to one end of the oblong propodus that it can be slid 
behind the latter as the sticks of a lady's fan slide over one another. A 
very similar form of limb with yet more elaboTate adaptive details is seen in 
Matuta which, like the Eaninidse, is at once a swimming and a burrowing 
crab. It is presumed that the lanceolate form of dactylus is more useful as 
a pick, the kukri-shaped dactylus as a shovel, but the latter may be service- 
able in swimming, just as a shovel may on emergency be used as a paddle. 
In Zanclifer the dactyli of all the pereiopods are narrow and sickle-shaped ; 
they seem ill adapted for swimming but effective instruments for digging. 
The criterion for deciding whether a limb is utilised for swimming or for 
digging consists in the length and mobility of its joints. The short strong 
limbs of Ranina showing restricted mobility at the joints are clearly fossorial 
and ill adapted for natation : the somewhat longer and more mobile limbs of 
Notopus and Notopoides are, however, serviceable for both purposes. In 
Notosceles (figs. 2 & 3) the dactylus of the second pereiopods is sickle-shaped 
as in Zanclifer, and it may be inferred that this pair of limbs is specialised 
for digging, but it can hardly be doubted that the broad kukri-shaped 
dactylus of the third pair, hinged fanwise to the outer angle of the trans- 
versely elongated propodus and provided with a marginal fringe of longsetse, 
is specially adapted for swimming. The reduced fourth pereiopods of this 
genus, with their small flat elliptical dactyli, can hardly be of any value in 
digging and are rather suggestive of steering paddles. 
5* 
