STUDIES IN SPICULE FORMATION. 549 
only of the scleroblast, and, similarly correlated with the 
quality of the cell substance, the scleroblast adheres to the 
grown spicule as a spherical mass with a very small basis of 
attachment. Now in regard to the character of the cell sub- 
stance, the Cucumarian scleroblast agrees with that of the 
echinopluteus, and in consequence the young sclerite origin- 
ates in the ectoplasm or extreme peripheral layer of the cell 
and the scleroblasts of grown spicules adhere as spherical 
masses, as shown by the figures. If we imagine say two 
indiarubber balls covered with a layer of some still softer 
material, which, we will also suppose, is secretory in function, 
to be situated between two boards (to represent the position 
of the scleroblasts between the external wall and internal flat 
layers of fibres), and whilst in this position to come into close 
contact, we can realise that this apposition, under such con- 
ditions, would produce the particular stresses in the soft layer 
lying between the two balls necessary to result in the forma- 
tion of a transversely-disposed rod of matter secreted by the 
soft layer. ‘he secreted matter cannot elongate in the direc- 
tion of the line joining the two balls, since the indiarubber is 
not secretory. On the other hand, the soft layer, which is 
secretory, is already elongated at right angles to this line by 
the mutual apposition of the two balls, and the secretion will 
occur in the plane parallel to the boards owing to the com- 
pression exerted on either side by these. Substitute endo- 
plasm for indiarubber, ectoplasm for the soft layer, and lime 
for the “ secreted matter,” and my meaning in reference to 
the possible reason for the deposition of the Cucumarian needle 
at right angles to the line joining the two primary scleroblasts 
will be evident. ‘The same argument applies to the bi-division 
of a scleroblast since here also the plane of separation of the 
two cells is occupied by a layer of ectoplasm necessarily situ- 
ated at right angles to the line joining the two cells. In 
sponges, on the other hand, the young sclerite naturally 
elongates in the direction of the line joing the two primary 
scleroblasts, since the elongated portion of endoplasm joining 
the two cells is the plasm that secretes the lime. Further, in 
vou. 49, PART 4,—NEW SERIES. 4] 
