THE ANATOMY OF THE MADREPORARIA. 415 
usual structure; but while in the case of the independent 
polyps of Rhodopsammia this is supported on the peripheral 
lamelle of the mesenteries, it rests here on the echinulations 
of the coenenchyme in the manner characteristic of Madrepora, 
Seriatopora, &c.; a result of colonialism, the importance of 
which has been already: pointed out (“ Anat. Madr.,’’ iii, sud 
Turbinaria). The tentacles are simple, their ectoderm is not 
raised into batteries, and they are apparently both ectoccelic 
and entoclic. The septa are both ectocelic and entoceelic, 
and fusion occurs between certain of the cycles ceutrally, as is 
the case in Rhodopsammia. 
Owing to the fact that the calicles increase by fission, there 
is great irregularity both of the cycles of septa and of the 
number of pairs of mesenteries; the latter varying in my 
specimens from 17—80 pairs. In sections passing through 
the plane where fission commenced, and where consequently 
considerable absorption and re-deposition of coral has occurred 
(fig. 7), the tubes perforating the skeleton are very irregularly 
distributed. At the surface, however, they are arranged more 
systematically in the manner usual among the Perforata, and 
show very clearly the method of their formation, to which 
reference has been made in a previous paper. They are not 
the result of absorption of the corallum, but are formed in all 
coenenchymatous Perforata which I have as yet examined in 
the manner indicated diagrammatically in fig. 11, representing 
the peripheral edge of a transverse section, and showing the 
transformation of the hemicylindrical canals immediately 
beneath the body wall into the internal cylindrical canals by 
radial growth and transverse fusion of the echinulations. 
Their formation in accenenchymatous Perforata is probably 
due to a somewhat similar but less regular process. The non- 
appearance of “directive”? mesenteries in all the polyps which 
I have examined is possibly due also to the multiplication by 
fission; their absence has been previously noticed only in 
Lophohelia (Anat. Madr.,” iii), Mussa, and Euphyllia 
(Bourne, ‘ Quart. Journ. Mier. Sci.,’ xxviii), but no explana- 
tion has been advanced to account for it. 
