36 DR. G. C. BOIIUNE ON THE KANINID^ : 
edges stop some v/nj sliort of the middle line and are not prolonged into 
conspicuous endosternites. Towards their lower ends they do, indeed, give 
off triangular or thorn-shaped offsets (such an offset is shown in the inter- 
segmental artlirophragm x/xi in PL 4. fig. 12), but these are not expanded 
into flattened summits and do not come into contact with their fellows of the 
opposite side of the body. Therefore there is no sternal canal. The 
epimeral apodemes or endopleurites are well-developed in crabs and, as in 
Macrura, they bifurcate at their inner ends, the anterior bifurcation uniting 
with the intersegmental apodeme in front, the posterior bifurcation with the 
intersegmental apodeme next behind, as is shown in fig. 12. It should be 
noted that in all these respects Dromia has departed so far from its Macruran 
ancestry and has taken on so completely the characters of crabs, that it has 
been possible to make use of it as a typical example of this aspect of cancroid 
structure. 
If we now bring the Eaninidse into comparison, taking as our example- 
Ranina dentata, we find a curious admixture of crab-like and Macruran 
characters, but the latter predominate and the former must be called 
deceptively crab-like, for a detailed examination shows that they must have- 
originated independently, in relation to the habits of the animal, and not by 
way of inheritance from any other recognized crabs. 
In the first place, as may be seen by a comparison of PI. 5. fig-^. 13^ l-l, and 
15, the contour of the thorax of Ranina as seen from behind is more lik& 
that of a lobster than like that of a crab. In the crab the width of the 
thorax largely exceeds its height ; in the lobster the lieight clearly exceeds 
the width ; in Ranina the height is rather greater than the width. Looking 
at the sternal surface, Ranina has a broad ventral plastron and in so far 
resembles a crab, but the most cursory inspection shows tliat it is jiiade up. 
almost entirely of the greatly enlarged and flattened sternum of the tenlli 
segment. The more posterior thoracic sterna, particularly those of the 
twelfth and thirteenth segments, are narrow and keel-like ; indeed, they are- 
proportionately narrower than in the lobster. The broad sternum of th& 
tenth segment is flat and has no vertical depth ; that of the eleventh 
segment has the form of a pair of aliform expansior.s extending along the- 
posterior edges of the twelfth sternum and making up with it the ventral 
plastron. Posteriorly these aliform sclerites meet in the middle line to form, 
the short and somewhat thickened body of the eleventh sternum, tilted 
somewhat upward. The twelfth sternum is very narrow, o£ considei'able- 
vertical depth, and is curved upwards. The thirteenth and fourtenith sterna 
are still deeper vertically, and are so much tilted upwards that the cavities of 
the third pair of jiereiopods look nearly directly backwards and the cavities- 
of the fourth pair look backwards and upwards (PI. 5. fig. 14). The large 
articular cavities of the clielipcds are placed laterally, but those of the first. 
