310 E. A. MINCHIN. 
thickener rounds off the proximal end to some extent and 
completes it. Full-grown spicules can always be distinguished 
from those not quite completed by the condition of the proxi- 
mal end (compare figs. 87, 88, and 89). With a little prac- 
tice the eye picks out the unfinished spicules rapidly in the 
preparation, and more careful inspection reveals at once the 
two formative cells on the shaft. ‘The abandonment of the 
spicule by the formative cells is practically the only point in 
which the development of the monaxon spicules differs from 
the rays of the triradiates, on which one cell, at least, namely 
the thickener, is retained. he difference is perhaps related 
to the fact that the triradiate systems are entirely imbedded? 
in the body, while one end of the monaxons always projects 
freely. The monaxons are probably being continually shed, 
or, it may be, drawn out of the sponge, like the hairs of many 
caterpillars. Hence the cells leave them, and new monaxons 
are continually being formed. ‘The triradiate systems, on 
the other hand, are more permanent skeletal elements, their 
chief function being support, and not defence or protection, 
and in consequence they retain their formative cells. 
I pass on now to describe briefly some of the peculiarities 
of the formation of monaxon spicules in particular cases. 
(1) Leucosolenia complicata. In this sponge there 
are two sharply distinct types of monaxon spicules (a) small, 
straight, slender forms, and (b) large, curved, stout forms. No 
transitions are found between these two types (Minchin, 1905 
[2]). 
(a) The slender monaxons are very abundant, as a rule, 
and their early stages are easily found, especially in the 
oscular rim. Sections show that in this region they may be 
1 Compare p. 319 below. 
* Woodland (1905, p. 237) states that in many Ascons the triradiates pro- 
trude through the thin body-wall, “so resembling monaxons.” I can only say 
that no such Ascons are known to me. I may point our further that the 
development shows clearly that the apices of the rays of the triradiates are 
comparable to the proximal ends of the monaxons, not to the distal pro- 
ecting ends, 
