MATERIALS FOR A MONOGRAPH OF THE ASCONS. 339 
for the present to confine the discussion to the spicules of 
Ascons. 
In my former memoir (1898) I rejected the notion that 
physical factors, such as crystallisation, had anything to do 
with the.forms of the triradiates, and considered their 
symimetry to be entirely the result of adaptation. I have 
been forced to reconsider my position in this matter. The 
regular and unfailing symmetry of the triradiates in the 
species of Clathrina is a very striking thing, and becomes 
more so when we take other facts into consideration. Von 
Ebner has shown that in Clathrinid triradiates the facial 
plane, i.e. the plane containing the apices of the three rays, 
is at right angles to the crystalline optic axis, so that 
the morphological axes of the three rays lie entirely in three 
planes of crystalline symmetry which contain the optic axis, 
and intersect at angles of 120°; and, further, that the straight 
gastral rays grow up in the optic axis, with which their 
morphological axis coincides. As Bidder (1898) has pointed 
out, the minute spines on the gastral rays of Clathrina 
cerebrum also form angles of 120° with one another, and 
repeat the symmetry of the basal rays. In the abnormal 
spicule figured by me (fig. 101) the additional rays come off 
from the main shaft at the same unfailing angle. If this 
angele is the result of adaptation, as I formerly argued,! it 
must be supposed that, by the operation of this principle over 
avast time, the cells have acquired a hereditary tendency 
always to attach one spicule-ray to another at a fixed, invari- 
able angle. This becomes difficult to imagine, when we find, 
as is frequently the case, that the three rays of a triradiate 
_ may be very irregular in the earliest stages, when they are 
still more or less separate from one another, and that they 
only become completely symmetrical when fused together ; 
an observation directly opposed to the theory ef hereditary 
tendency, which, as it seems to me, cannot go further than 
1 Maas, from his embryological observations, has pointed out that the form 
of the triradiates is in no way correlated with the arrangement of the pores, 
at least in ontogeny (1900 [2], p. 227). 
