304 Mr. H. J. Carter oz 
cellous structure of bone), so that, in the absence of the latter, 
as just stated, none of them presents the least sign of a 
wall. The only part which appears solid or different from the 
rest in structure is the tabula or transverse septum of the 
tubular excavations (fig. 5,c). Indeed, here as elsewhere 
in the Hydractiniide, it is the intricate anastomosis of the 
minute branches of the ccenosare which leads to the formation 
of the peculiar, cancellated, coenosarcal skeleton, which, again, 
is as indicative of the structure of the Hydractiniide as it is 
opposed to that of the Foraminifera. Not unfrequently the 
tabula is surmounted centrally by a kind of style, which, in 
some species of Stromatopora, seems to be indicated by the 
presence of a white point in the centre of the transparent calc- 
spar filling the rest of the calicle. 
Finally, if the elkhorn-like branch of Millepora alcicornis 
be split in two through the centre parallel with its flat sur- 
face, the plane of each part thus exposed will present concen- 
tric lines of lamination in the cancellous structure, which, ex- 
tending from side to side parallel with the plane, indicate the 
progressive formation of the flat branch upwards from the 
point at which it commenced to expand. These lines of 
lamination, however, are, in accordance with the rest of the 
structure, not indicated by distinct laminee, but by the position 
of the rows of apertures in the ccenosarcal skeleton or co- 
rallum. 
Thus we have every thing structural in the corallum of 
Millepora alcicornis that is to be found in Stromatopora, 
excepting the stedliform systems of venation. 
What, then, were those “ stelliform systems’ which, in my 
paper on Hydractinia, &c. (‘ Annals,’ 1877, vol. xix. p. 68, 
and pl. viii. figs. 19 &c.), I have likened to the superficial 
stelliform excretory canal-systems of some sponges, and sug- 
gested a like function? J] need hardly answer the question, 
after what has been above stated, especially when they are to 
be seen on the upper surface of every layer of a Stromatopora 
as it is split off from the entire specimen. ‘They were not 
water-vascular excretory systems as in sponges, but tubular 
venations of the coenosarc on the surface, probably filled with 
“‘entodermic cells,” as will appear hereafter, preparing the 
way proliferously for the new layer; and, although in no 
instance that I know of, yet there may be an existing species 
of Millepora in which the grooved venation, instead of mean- 
dering generally over the surface in large branches and sending 
off smaller ones, which by subdivision become still smaller, 
and ultimately intermingle with each other (like the capil- 
