THE AN ATOM'S OF THE HADREPORAR1A. 
593 
exact size and shape. This condition is probably due merely 
to alcoholic contraction, and does not imply that involution is 
the normal method of tentacular contraction. A similar 
invagination had taken place at the bases of the tentacles of 
Flabellum. They are covered with nematocysts, which are not 
so sharply defined into batteries as was the case in Flabellum. 
At a varying depth below the lip of the calyx (but generally 
at a lower point than is represented in fig. 13, which is consi- 
derably shortened in the longer axis) the external body wall 
perishes, owing probably to the various parasites that infest the 
external surface of most coral thecae and polyps; notably a 
sponge, which in some places eats its way right into the theca. 
The cavity marked / in fig. 15 is thus filled with sponge 
spicules. Below the point at which the body wall ends, is 
visible in some places a thin line of tissue indicated in fig. 15, g, 
which may or may not be a part of the polyp. The appearance 
of the periphery of the theca in such a section suggests very 
strongly that a secondary line of corallum has been deposited 
round the circumference to protect the canals from direct 
communication with the sea water and against the parasites. 
At the top of fig. 15 the semicircular outline of the canals 
seems to indicate such a formation. 
The mesenteries vary in number, and are, like the entosepta, 
generally of three orders. They are divisible into “ pairs ” as 
in the other forms described, and possess the same arrangement 
of longitudinal retractor muscles on their entocoelic faces, with 
the usual difference in the two directive pairs. The trend of 
these muscles is roughly indicated in fig. 13 ; but their minute- 
ness renders it impossible to recognise the arrangement of the 
protractor muscles, though they are just visible in microscopic 
sections. There appears to be but little contortion of the free 
edge of the mesentery ; and the traces of any organs resembling 
acontia are rare. This, however, may be due to deficiency of 
material, which has much hampered my investigation of this 
form. 
Both primary and secondary mesenteries appear to be united 
to the stomodseum for its whole length ; those of the third 
