A STUDY IK CAECINOLOGT. 51 
Eiich of tlie lateral wings is armed anteriorly with a prominent spine ; its 
enter border is nearly straight and makes nearly a right angle with the 
transverse ledge formed by the median lobes ; its inner border is tliickened 
and curves round the outer extremity of the mandibular sternum. The last- 
named structure is a projecting curved bar intimately fused to the median 
area of the antennary sternum, but marked off by a groove which in the 
middle line deepens to form a triangular fossa. Laterally the mandibular 
sternum seems to terminate in a pair of incurved prominences for articula- 
tion with the inner articular processes of the mandibles, but is clearly 
prolonged beyond these as a pair of diverging horns intimately fused to the 
postero-internal edges of the aliform external areas of the antennary sternum 
but projecting slightly beyond the latter. The mandibular epiniera are 
largely membranous, but their antero-internal ends are calcified to form the 
two plates marked in fig. 26 which form the roof of the anterior part of 
the exhalant branchial canals. A comparison of figs. 19^ 22, and 27 will, I 
think, convince the reader that in the whole make-up of the pre-oral region 
the Dromiacea have departed further from the Astacuran type than have the 
Raninidse, particularly in the reduction of the rostrum, the greater develop- 
ment of the suborbital lobe of the caiapace, the membranous condition of 
the inner moieties of the ocular peduncles, the size and relative positions 
of the antennulary and antennary fossae. Further, the modifications of the 
antennary and mandibular sterna are widely divergent in the two groups. 
It may be objected, and there would be some force in the objection, 
that the genus Dromia, on which I have relied for study of details, is a 
much modified and specialised genus of the Dromiacea, and that I should have 
directed my attention rather to the more primitive genera, Homolodroinia and 
Dicranodromia, on which Bouvier so lai-gely relies in establishing his theory 
of the Astacuran origin of cuiibs. Unfortunately examples of these rare 
and instructive forms were not at my disposal, but thej' have been described 
in sufficient detail by A. Milne Edwards and Bouvier (28), and a reference 
to pi. i. fig. 2 and pi. iii. fig. 2 of their admirable memoir will convince the 
reader that I am justified in extending the results of my study of the genus 
Dromia to the more primitive members of the group and in asserting that in 
the make-up of the pre-oral region, as in other characters previously dealt 
with, the Dormiacea have departed more widely from the Astacuran type 
than have the Raninidfe, and therefore cannot be enrolled in the ancestry of 
the latter. 
I submit that, by a detailed stud)' and comparison of the nervous system, 
the endophragmal system of the thorax, and the cephalic segments, I have 
proved the first part of my thesis and have shown that, whilst there is plenty 
of evidence for deriving each group separately from an Astacuran ancest«r, 
the HaninidtP, cannot be directly descended from a Dromiacean stock. The 
4* 
