40 DR. G. C. BOURNE ON THE RANINID^ : 
beyond the intersegmental endopleurites xiii/xiv, unite with the inner edges 
of the lower pnrt oE the endopleurites xii/xiii. The sella turcica o£ Ranina 
is therefore of large extent, larger than in so ty[)ical a crab as Carcinus 
mienas. Among the various Brachyura that I have studied, it most nearly 
resembles in form and extent that of Calappa. But, as mentioned above, the 
sella turcica is not found in all crabs. Fig. 15 represents a posterior view of 
the thorax of Dromia vulgaris. The sterna of the twelfth, thirteenth, and 
fourteenth segments can easily be recognised by the strophidia on their outer 
extremities ; the large pair of ventro-later.al projections in front of them 
belong to the eleventh sternum ; their strophidia, being on their anterior 
face, do not appear in the drawing. The fourteenth (last thoracic) sternum 
does not differ much in shape from that of Banina and offers resemblances 
to that of the lobster. The thirteenth and twelfth sterna are irregularly 
shaped transverse bars, of which the outer portions are calcified and 
distinctly separated by apodeines, but tlieir admedian portions lose their 
distinctness and pass into a median membranous pouch, the walls of which 
are little if at all calcified. There is no trace of a median longitudinal 
apodeme in this membranous representative of the penultimate and ante- 
penultimate thoracic sterna, consequently there is no sella turcica. The 
arthrophragms of the intersegment xiv/xv are seen converging forward from 
the anterior edges of tlie V-shaped arms of the last thoracic sternum, and in 
other crabs they unite above with the epimera of their own segment, in front 
with the arthrophragms of the intersegment in front, but in the absence of 
the ventral apodeme they fall far short of nKeting in the middle line to form 
a sella turcica. In this respect, then, Dvomia has progressed less from the 
Macruran towards the Brachyuran condition than Ranina, and this might be 
fastened upon as an argument in favour of the derivation of the Eaninidse as 
of all other crabs from a Dromiacean ancestor, but the transversely elongated 
twelfth and thirteenth sterna of Dromia are quite Brachyuran in character 
and very different from the narrow and keel-shaped corresponding sterna 
in the Raninida3. 
On the balance the evidence afforded by the structure of the hinder part 
of the thorax is against the Dromiacean ancestry of the Raninidse. 
Another crab in which the sella turcica is absent is Pldlyra hevis. In view 
of de Haan's emphatic opinion that the Lencosiidse are clearly united to the 
Raninidse through Lyreidus, an opinion which has found favour with subse- 
quent authors, the m:irked differences in the structure of the thorax demand 
attention. A detailed comparison of the endophragmal skeletons of the 
Raninidse and Leucosiidaj is best reserved for the discussion of the possible 
derivation of the latter from the former family in another part of the paper, 
but it may be conveniently pointed out in this place that in Philyra the 
plastron is broadest at its posterior end ; the intra-sternal apodemes are nearly 
