19 
Species of Echinonematous Sponge . 
given rise to the smooth paired rays, while a third bud must 
have started from its outer or distal face, and by growing out¬ 
wards have produced the third unpaired ray. From the sides 
of the third ray additional offshoots have been developed 
forming the spines, into each of which a branch from the axial 
canal of the ray on which they occur is continued. 
As it is very probable that the spines of spinose spicules in 
many other kinds of sponges are solid, and not provided with 
lateral diverticula from the axial canal, it may become neces¬ 
sary to distinguish between two kinds of spines, those with and 
those without a central canal, just as one distinguishes between 
those spines of plants which are mere epidermic outgrowths 
and those which are true aborted branches. Indeed, since 
writing the preceding sentence I have been able to verify the 
existence of solid non-canaliculated spicular spines in several 
instances, one of which will be described in a forthcoming 
paper on a new species of Plocamia. 
The variations from the average form of ecninating spicule 
are exhibited by some 4 or 5 per cent, of these spicules, and 
extend over a very wide range. In figs. 17 and 18, PI. V., 
we have a form possessing two rays only, one of the proximal 
rays having disappeared and the remaining one grown back¬ 
wards in the same line as the distal ray. Fig. 20 is another 
case of a similar kind, but with the single proximal ray in¬ 
clined at an angle of 125° with the distal one. 
Thus figs. 17 and 18 are biradiate but uniaxial forms, and 
in this respect resemble ordinary straight acerate spicules j 
while fig. 20 is biradiate and biaxial as well, and thus re¬ 
sembles a curved acerate, which is both biradiate and biaxial 
and therefore cannot, either in a strict or even in a wide sense, 
be said to be monactinellid : the only truly monactinellid 
spicules are the pin-headed ones and some other (but not all) 
forms of acuates. The variations shown in figs. 22, 24, 27, 
and 81 exhibit changes in the relative length of the three rays, 
and in the angle which the paired rays make with each other 
and with the third ray. Thus in fig. 27 the three rays are 
inclined at nearly equal angles of 120° with each other, and 
are all of nearly the same length ; in fig. 22 one smooth ray 
continues the direction of the single ray, and the other projects 
at right angles from it, producing a form which recalls that of 
some commonly occurring hexactinellid spicules. In fig. 31 
both paired rays diverge nearly at right angles from the third; 
in fig. 24 one smooth ray points forwards and the other keeps 
its normal backward direction ; in some instances, not figured, 
both smooth rays point forwards. 
Fiffs. 23 & 25 illustrate variations in the terminations of the 
O . 
