350 E. A. MINCHIN. 
out very easily in the transparent preparations. As the tubes 
are rounded, the spicules placed more towards the side are, 
of course, somewhat foreshortened. At first I used to draw 
the whole spicule, but I found it just as useful, and much less 
laborious, to draw only the straight posterior rays with 
the camera lucida, and to add the lateral rays freehand. It 
is for this reason that I state that the precise form of a given 
spicule is not to be considered as accurately represented in 
the figures. In fig. 108, Pl. 21, I have carefully drawn with a 
camera lucida the arrangement and form of the triradiate 
systems in one of the diverticula. 
The sponge has erect oscular tubes, often of great size 
(Pl. 21, figs. 105, 107), from which grow out very numerous 
diverticula, ramifying in their turn (text-fig. 4). The diverti- 
cula are at first, and for a long time, blind at their ends (text- 
figs. 1—4), but become sooner or later perforated at their 
blind extremities to form oscula (text-fig. 5, 0.), thus pro- 
ducing smaller oscular tubes which may be conveniently dis- 
tinguished as secondary from the large erect primary oscular 
tubes. 
Beginning the description with the primary oscular tube, a 
very uniform and regular arrangement of the triradiate 
systems is found (text-figs. 1, 2, and 3,0. 7.; the arrows 
point in the direction of the oscular opening). Lach triradi- 
ate is disposed with its straight unpaired ray pointing away 
from the osculum, and its two lateral rays extending ina 
direction as near to the horizontal as their form permits, that 
is to say, with their points slanting slightly upwards. In 
those systems which develop a fourth or gastral ray, and so 
become quadriradiates, it follows that the gastral ray curves 
forward to point towards the oscular opening, that is to say, 
that the gastral rays bend over and point in thé direction of 
the water current, which flows up the oscular tube towards 
the terminal aperature. The suggestion that the gastral rays 
owe their curvature to the influence upon them of the water- 
current during development is an obvious one; we shall 
return to this point later. 
