128 
ME. H. N. MOSELEY ON THE 
lies a layer of apparently homogeneous membrane, which appears to form everywhere a 
wall to the vessels and canals. The cavities of the vascular network are lined by, and 
in many places nearly tilled with, cellular elements of two kinds — pigmented cells and 
small transparent globules. The pigmented cells (Plate 3. fig. 17) closely resemble those 
of other hydroids. They are spheroidal in form, with transparent wall and contents 
composed of irregular granules, which are of a bright gamboge-yellow colour. It is 
these cells which give the bright yellow tinge to the tips of the living corallum. The 
cell-contents in these cells are frequently to be seen divided into two, each half having its 
own nucleus, or sometimes more rarely into four (Plate 3. fig. 17, b , c). The more super- 
ficial part of the vascular network of the hydrophyton is in most places almost crammed 
full of these pigmented cells, and they are abundantly present also within the somatic 
cavities of the zooids. They become less abundant towards the deeper parts of the 
living layer, and in certain of the deepest ramifications of the network are entirely 
absent, their place being taken by transparent globules. In some parts of the 
corallum large quantities of the pigmented cells are met with which are coloured dark 
brown instead of yellow. These belong probably to the older parts of the coral, which 
have in the living condition a brown appearance, it being only the growing tips which 
are bright yellow. Such, however, was not ascertained to be the case. 
At the under surface of the living layer of the hydrophyton the vascular network 
has in connexion with it, or is prolonged into, a network of extremely transparent thin- 
walled vessels, many of which terminate in blind extremities, as shown in Plate 3. 
fig. 16, B. These vessels are distended with small exceedingly transparent and highly 
refractile globules, without any admixture of pigmented cells. These transparent globules 
are found scattered amongst the pigmented cells throughout the vessels of the hydro- 
phyton, but occur in masses only as above described. No clue to the function of these 
transparent globules, nor explanation of their being thus agglomerated in the deeper 
parts of the living layer, was obtained ; the masses of them probably point to a fatty 
degeneration of the effete deep regions of the network of the hydrophyton. 
Thread-cells. 
The thread-cells are of two kinds. They are shown, carefully drawn to measurement, 
in Plate 2. figs. 1 and 2. One kind is that which appears to be confined to Hydrozoa, 
and not to occur at all in Anthozoa, viz., that in which a bladder-like enlargement of 
the thread occurs at that part of it which is immediately next the mouth of the cell, 
the bladder being armed near its summit by three spines set in one whorl. The three 
spines in this form of thread-cell in Millepora are remarkably long, and directed at right 
angles to the axis of the thread, instead of recurved, as usual. These thread-cells vary 
very much in size. The one figured is one of the largest observed, being of about two 
thirds of the length of the ovoid thread-cells. The larger examples of these three- 
spined thread-cells are of comparatively rare occurrence, only a few being present in 
some of the tentacles, and being more commonly present in the tentacles of the mouthed 
zooids. The smaller thread-cells of this form have not more than ^ of the length of the 
