464 Mr Gardiner , Notes on Variation, 
as new under the name of F. variabile 1 — and Blastotrochus nutrix 1 2 . 
The latter genus is one which I can consider as in no way different 
from Flabellum. It was founded by Milne-Ed wards and Hamie 3 
on a misconception, the universal belief at that day and in 
Semper’s time being that the coral skeleton was formed by the 
endoderm, whereas it is really a deposit outside the ectoderm and 
hence morphologically completely outside the coral polyps. The 
wall of the coral of Flabellum is epitheca and not theca, so that 
hence there is no polyp tissue external to it, from which budding 
could occur. 
The Cape of Good Hope specimens belong to F. rubrum, but 
have about 2 per cent, of their number intermediate towards 
F. stokesi and F. nutrix. The Maidive specimens mostly belong to 
Semper’s F. stokesi , but some to the same author’s other species. 
The first three cycles of septa in all are equal and fuse in the 
axial fossa by trabeculae from their edges. The fourth cycle is 
smaller, but commonly some of its septa fuse to the others by 
trabeculae, this character being generally more marked in F. stokesi. 
The Maidive specimens of each of the so-called species as com- 
pared with the Cape specimens show an increased length of the 
basal scar. The breaking off of the corallum takes place along 
one of the lines of growth, and probably the growth is greater 
with increase of temperature, better food-supply, etc., so that even 
if all coral lites broke off along the same growth line — an unlikely 
supposition — the scar would in different regions vary in size. 
There is absolutely no difference in the polyp anatomies of either 
of the two species, but the presence of wings along the end of 
the calicle in one and of spines in the other is a fairly definite 
difference. The British Museum forms show, however, inter- 
mediates in this character and also in the size of the scar. Out 
of about 600 the total number of intermediates (in the strictest 
sense) was 17, of which 4 alone were intermediate in both 
characters. It is hence evident that F. rubrum and F. stokesi re- 
present variations of the same species. 
Doubt only remains as to the so-called Blastrotrochus nutrix. 
I had only one specimen from the Maidive, which was absolutely 
identical with Semper’s description of the species. It was quite 
similar in the polyp anatomy to the other forms, but in the 
corallum closely approached to F. rubrum. The so-called buds 
are approached in many of the Cape specimens of F. rubrum. 
Neither in these nor in the Maidive specimen is there any 
connection with the tissues of the large polyp, and it is doubtful 
whether they may not have been largely derived from ova, rather 
1 Loc. cit. pp. 245 — 251. 
2 Loc. cit. pp. 237 — 241. 
3 Ann. des Sc. nat., 3 C ser., t. ix., p. 281 (1848), and Cor. n. p. 87 (1857). 
