Lichinoderm-Fauna of Ceylon. 223 
being the only possibility when centralization has extended so 
far. In the next place, it is to be noted that when the divided 
disk heals and buds afresh it may give signs of the loss of 
the quinary principle, and more than two arms may take the 
place of the two that are gone. The tendency to fission, the: 
child of external irritation, became the parent of a habit of 
fission or simple reproduction. Carried on under certain con- 
ditions this habit led to yet another change; the permanently 
or characteristically sexradiate form, both in the north and in 
the south (Ophiacantha anomala and O. vivipara) ceased to 
have free-swimming embryos, and became viviparous. Here 
is cenogeny indeed! not only no trace of the bilateral sym- 
mee of the embryo, but loss of quinary proportion in the © 
adult ! 
Results so remarkable as these must not be dismissed as 
“ freiwillig ;” it seems that the tendency of an Echinoderm 
to break up under external irritation must be taken to be a 
proved fact. Passing from it forwards we recognize as an 
expression of this tendency, now crystallized into a habit, the 
fission of the Ophiurid disk; in some of the Asteroidea the 
tendency has become economized or concentrated, and only a 
single arm separates from the disk. 
If, however, instead of passing forwards and coming within 
the range of heteractinic phenomena, we pass backwards into 
an earlier condition, and try to work out the cause of this ten- 
dency to division, we find ourselves brought face to face with 
polyactinism, a stage which must certainly be regarded as 
earlier than a fixed pentactinic condition. Here we find 
(Brisinga, Labidiaster) that the separate arms appear to break 
off for the purpose of setting free the genital products—a con- 
dition not inexactly paralleled by that zoological paradox 
Palolo viridis, and bearing a significant analogy to what ob- 
tains in some Discophora, where the researches of Prof. Hiackel 
and of Dr. Romanes (whose work would not seem to have 
been consulted by Prof. Hickel) have shown that in some 
cases, at any rate (Aurelia aurita), the phenomenon of strobila- 
tion is associated with very great variability ; or, to put the 
matter more generally, reproductive fission in low forms (or 
early stages ?) is more or less indefinite in direction. 
The origin, then, of the habit of self-mutilation in the 
Echinodermata is to be sought for in the imperative necessity 
of reproducing the species ; as concentration and consolidation 
went on this habit disappeared, to be again roused into acti- 
vity by the attacks of enemies. ‘Thus roused, it has, in some 
groups, become definite in direction, and has again become a 
factor in reproduction. But the difference is a real one: at 
