REPORT ON CORALS—HYDROCORALLINA. 25 
boundary often very indistinct. The cells form in some places much thicker layers than 
in others. Towards the most internal part of the ccenosare they become entirely lost, 
their place being taken on the surface of the thin-walled cyst-like innermost elements of 
the network by a thin layer of structureless protoplasm (Pl. XIV. fig. 8, B). 
Narrow strings of this ectodermic protoplasm (Pl. XIV. fig. 8, S) cross over here and 
there between adjacent vessels of the innermost part of the network, being possibly the 
remnants of effete vessels, The ectoderm covering the parts of the network near the 
surface is much thickened and modified, a large proportion of its cellular elements being 
there found converted into the parent-cells of nematocysts, and being thus inflated and 
occupied by nematocysts in all stages of development. 
At the actual surface, the ectodermal cells undergo still greater modification, forming 
a superficial layer of elongate prismatic transparent cells, which-shows on the surface a 
series of irregularly hexagonal areas corresponding with the summits of the cells. These 
cells contain oval nuclei and nematocysts of both kinds in various stages of development. 
The most superficial film showing the hexagonal areas separates often in osmic acid pre- 
parations as an exceedingly thin membrane, as is shown on the right-hand side of Plate 
XIV. fig. 8. There is some uncertainty as to the exact structure of the superficial layer 
of the ectoderm. The figures represent what, after a careful investigation, was con- 
cluded to be the arrangement existing. The layer is seen well only in preparations from 
specimens hardened in osmic acid. The lateral boundaries of the prismatic cells were 
never seen well defined, but the polygonal areas, corresponding with their summits, were 
seen well in various preparations. It could not be demonstrated with certainty that 
this layer extends uniformly all over the external surface of the ccenosare. It is 
extremely transparent, and difficult to trace in preparations viewed from the surface, 
over the cavities caused by the removal of the hard parts by decalcification. 
The exact arrangement of the superficial layer in its connection with the mouths of 
the calicles and zooids is also somewhat uncertain. The layer certainly is prolonged into 
the calicular cavities, and contains the mass of large oval thread-cells which surrounds each 
zooid. In most preparations the zooids are far retracted through the action of reagents, 
and the mouths of the calicles are closed above by a layer of tissue which shows a radiate 
striation or slight plaiting around a very small circular central orifice, which orifice leads 
down a short tubular cavity formed by the superficial layer drawn everywhere inwards to 
the retracted zooids of the calicle. The layer of tissue thus contracted over the calicle 
nearly or sometimes completely closes it, and thus usually the zooid can be seen in prepara- 
tions in which the ccenosarc is viewed from the surface only by focusing the objective into 
the depths of the tissue. The orifices of the pores of retracted zooids were unfortunately 
not carefully examined in the living condition of the coral; hence, it is uncertain 
whether the superficial tissue contracts in this manner in the living condition so as to 
close the orifice of the pore, and protect the zooid, or whether such extreme contraction 
(ZOOL. CHALL. EXP.—PART vil.—1880.) G 4 
