160 
THE VOYAGE OF H.M.S. CHALLENGEE. 
has increased to 0*008 mm. in diameter, two of the cladi are 0*0236 mm. in length, 
and the third, 0*004 mm. long, appears merely as a rounded tubercle. In succeeding 
stages the tylus and cladi increase in size ; but the rhabdome grows so much more rapidly 
than the tylus, that eventually it becomes of the same thickness, and the distinction 
between the two, though it persists for some time, is finally obliterated. In a spicule 
with the rhabdome 0*0118 mm., the tylus is 0*0158 mm. in diameter, and the axial 
fibre of the third and shortest cladus is traceable into it for a distance of 0*007 5 mm. The 
axial fibre of this third cladus, which in the adult spicule is usually entirely suppressed, 
can be traced for a long distance through the spicular series, indications of it persisting 
in some cases in spicules with a rhabdome 0*039 mm. in diameter, and therefore presum- 
ably almost adult ; in this case it is represented by its axial fibre, which is about 0*0035 
mm. in length. The two persistent cladi of the young spicules do not lie nearly in the 
same plane with the rhabdome, i.e., at about 180° with each other, but diverge at an 
angle which is more nearly 120° than that. 
Having shown that the adult disene originates in a trisene form, and that in all proba- 
bility in an oxytylote, which is descended in turn from an oxea, we may next turn to enquire 
into the cause of the suppression of the third cladus of the trisene. This appears to be 
traceable to pressure. The suppressed or arrested cladus, in the early stages of develop- 
ment of the spicule, is pressed against the rhabdome of the spicule next behind it in the 
series (PI. XVII. fig. 12), and, considering the sensitiveness of more fully grown spicules 
to the slightest action of pressure, it is no wonder that a young cladus so unfavourably 
situated should suffer an early arrest of development. To a similar cause in aU pro- 
bability the suppression of one or more cladi of the anatrisenes of the sponge-body is 
due. 
The scleroblasts of the megascleres are well displayed in several preparations of this 
sponge. The characteristically large nucleus of the spicule-cell is situated at about the 
middle of the oxeate spicules, just over what we may presume is the position of the 
actinal origin. In the trisenes it always occurs nearer the cladal than the oxeate end. 
In a young orthotrisene of the sponge-body the distance of the nucleus from the cladal 
origin measures 0*146 mm., and from the oxeate termination 0*288 mm. ; in a more fully 
grown form these distances become 0*592 and 1*203 mm.; in an anatrisene they are 
0*217 and 0*414 mm. In these three cases the distance of the nucleus from the cladal 
origin is about one-third the total length of the spicule, and if we suppose the actinal 
centre corresponds, as appears very probable, with the position of the scleroblastic nucleus, 
we may express this fact in the statement that the length of the cladal actine is to that 
of the acladal actine as 1 ; 2. 
Thus a kind of balance appears to exist between the two actines, the oxeate actine 
compensating by its greater length for the greater thickness and the branches of 
the cladal actine. As a notion seems to exist that this may be the case, Carter often 
