34 DR. G. C. BOUKNE ON THE KANINID^ : 
apodeme seems to end just below the articular knob : in reality it is prolonged 
beyond it, and can be traced along the inner and hinder edge of the twisted 
lamella or arthrophragm that separates each articular cavity from the one 
next behind it. The arthrophragm, in fact, is a double fold of the cuticle or 
apodeme, continuous with the pleural apodeme ; the two edges of the fold 
run down rather to the inside of the hinder edge of the arthrophragm, and in 
the entire animal pass respectively into the arthrodial or joint membrane of 
tlielimb in front and the limb behind. At the lower end of the arthrophragm 
the groove forms the apparent boundary between the contiguous sterna, 
but here again there is no suture ; the sterna are really continuous, and 
are only separated by apodemes. All this is familiar, but as separate names^ 
epimera, sterna, arthrophragms, etc., are given to the various parts, I have 
thought it desirable, in view of the comparisons that are to follow, to emphasize 
the fact that these names only apj)ly to local areas and indurations separated 
from one another by more or less deep infoldings of a continuous cuticle. 
In the upper and lower parts of the arthrophragm its apodeme is shallow, and 
consequently the arthrophragmal partition extends only a little way into the 
cavity of the body ; but at about the middle of its course the apodeme 
becomes very deep and gives rise to a lamina running upwards, forwards, 
and inwards towards the middle line, short of which it expands into 
horizontally flattened plate witli jagged edges, which makes more or less 
intimate union with its fellow of the op)posite side. The laminae in question 
are the endosternites : they and their flattened summits form the walls and 
roof of the so-called sternal canal within which the thoi-acic nerve-ganglion 
chain lies. A sternal canal is found in the Macrura reptantia and in some 
Anomura, but, with the exception of the Raninidae, it does not occur in the 
Brachyura. PI. 4. fig. 11 also shows the epimeral apodemes or cndopleurites. 
As the articular cavities slant backwards, the endopleurites alternate in 
position with the endosternites, and, as is familiarly known, each endopleurite 
divides internally, sending a posterior branch to unite with the endosternite 
of the segment behind, an anterior branch to unite with the endosternite in 
front. The result of all these complicated arrangements is that the cavity of 
each limb communicates internally with two imperfectly delimited chambers, 
which may properly be called muscle-cavities as the abductor and adductor 
muscles of the limbs are attached to the several apodemic ingrowths that 
constitute their walls. In the Macrura, as the articular sockets of the limbs 
look nearly ventrally and the epimera are not much inclined inwards from 
the vertical, the muscle-cavities lie in nearly the same horizontal plane and 
alternate with one another. A consideration of the relative positions of the 
dorsal strophingia and ventral strophidia by which the coxa of the limb is 
hinged to the articular frame will show that the outermost and slightly 
dorsal of the two muscle-cavities encloses the abductor muscles whose action 
