1869.] DR. J.S. BOWRRBANK ON SILICEO-FIBROUS SPONGES. 79 
seum in about the same degree of preservation as the type one; but 
in consequence, probably, of not having been so much washed to 
make it look pretty, it abounds in the beautiful and characteristic 
spinulo-trifurcate hexradiate stellate retentive spicula. 
The fibre in the skeleton is abundantly but irregularly tuber- 
culated, as represented in fig. 1, Plate III., from a section of 
the type specimen from Barbadves in the British Museum, x 108 
linear. 
The tuberculation of the fibre is remarkable and very character- 
istic ; when viewed with a power of about 700 linear, their apices are 
always more or less papillous ; in some the papillee are numerous and 
well produced, while in others they are in an incipient condition. 
Fig. 13, Plate III., represents two of the tubercles on the side of a 
portion of skeleton-fibre with their terminal papillee, x 666 linear. 
Beside the large primary fibres, there is a secondary series of 
skeleton-fibres, which are evidently auxiliary to the larger system. 
They occur especially in the large interstitial spaces of the sponge, 
their office being apparently that of filling up those vacant spaces 
when no longer necessary in the economy of the animal, and to sus- 
tain therein the multiplied folds of the interstitial membrane ; their 
office in this respect is the same as that of the large rectangulated 
hexradiate spicula (Pl. III. fig. 2) which occur so frequently in the 
interspaces of the skeletons of the siliceo-fibrous sponges, and their 
mode of development very closely resembles that of those spicula. 
In an early stage of their growth they very closely simulate the form 
of the spicula; but instead of being freely developed amidst the 
membranous tissues, they are always based on the primary skeleton- 
fibres. A single small fibre pullulates from some part of one of the 
larger skeleton-ones, and is projected in a straight line into the vacant 
space: if it meets with none other in its progress, at some distance 
from its origin four lateral branches are thrown out at right angles 
to the axial fibre and to each other, and the axial fibre conti- 
nues its progress in a straight line. If it meets no other fibre in 
its progress, the distal ends of the axial fibre and of the lateral 
ones become clavated, and all parts of the shaft and radii profusely 
spinous, and the whole constitutes a perfect simulation, in form, 
of a rectangulated hexradiate spiculum. But, on the contrary, 
should the axial or the radial branches meet with other such fibres, 
they immediately inosculate, and the previously straight radii are 
contorted in various directions to meet the necessities of the situa- 
tion; and, as is frequently the case, where many of these fibres 
are projected from different bases into the same space, they unite 
and form one mass of small contorted fibres, while there is good 
reason, from the gradual increase in size of the basal porticns of 
some of them, to believe that they are ultimately developed into 
the size and form of the primary skeleton-ones. 
The primary skeleton-fibre averages =, inch in diameter; the 
auxiliary fibres vary from 5755 to => inch in diameter. How 
ever closely they may simulate the form of true hexradiate spi- 
cula, they may always be distinguished from them by their attach- 
