STRTTCTTTEE AND RELATIONS OF CEETAIN COEALS. 
113 
work of tortuous branches of hard tissue, with cavities of two kinds, smaller and larger. 
In Millepora alcicornis the calcareous mesh work is comparatively open. In the 
Philippine species, in which the polyps are extremely minute, the ccenenchymal tissue 
is much more compact — so much so that instead of, as in M. alcicornis , the hard and 
soft parts appearing to form equally complicated meshworks, the soft tissues appear to 
occupy a series of tortuous canals bored in the compact coenenchym ; and in finely ground 
sections these canals becoming filled with debris , and thus as it were injected with 
opaque matter, stand out black in relief when the sections are viewed by transmitted 
light, and show the relations of the hard to the soft parts in a very clear manner. The 
tortuous tubes channelled in the hard coenenchym are seen to lead from the calicles 
in all directions, and anastomosing freely with one another to join the cavities of 
surrounding calicles. 
When a mass of Millepora hardened in alcohol is decalcified, a thin layer, consisting 
of almost the entire living substance of the coral, separates from the surface, and, as the 
only residue of the deeper tissues, a greenish gelatinous mass remains. The superficial 
layer thus separated consists of an irregular network of tortuous canals, which in the 
recent state occupied the canals in the hard coenenchym or were interwoven with its 
equally complex meshwork. The canals are filled with round cells, resembling the 
endodermic cells of Alcyonarians ; their wails are thickly beset with the large thread- 
cells, which have already been described and figured by Professor Agassiz. The cells are 
ovoid, and are provided with a long filament armed near the end with a spiral of spines. 
The thread-cells are developed in special large transparent sacs, and are seen in all 
stages of formation. Two kinds of thread-cells are present, the larger ones just described, 
and others of nearly the same form, but of not more than one fifth the length, which 
small thread-cells are confined to the tentacles, on the surfaces of which they are 
densely set. The larger thread-cells are especially abundant round each polyp, forming 
a zone around it. The canals are covered externally with a transparent tissue, pro- 
bably calciferous in function. In certain places in the Millepora obtained at Zambo- 
angan here and there canals larger than the others, chief stems as it were, are to 
be found running a long straight course on the surface for an inch or more, and giving 
off branches on either side. Canal offsets of the general meshwork join the polyp 
masses all around in a radiate manner. The larger polyps, as seen from above, 
have an irregular indented outline, whilst the smaller are simply circular or nearly so. 
The larger are less numerous than the small. Tentacular masses covered with the 
smaller thread-cells can be seen in the apertures of the polyps of both kinds. Their 
number could not be determined with certainty, since it is impossible to define the 
exact limits of the contracted lobulate masses and distinguish a mere lobe of a tentacle 
from an entire tentacle. In the larger polyps, however, there constantly appeared to be 
four. Professor Agassiz saw and figured four tentacles only in Millepora alcicornis- 
Pourtales says that the tentacles appeared to him “rather as tentacular masses 
studded with lasso-cells, five in number.” From the appearances which present 
