REPOET OR" THE TETRACTINELLIDA. 
Ixxvii 
Tlie growth of a scleroblast, either as a prolate or oblate ellipsoid, is determined by 
tensions existing in the organism, thus the oblate spheroids in the Tetillidse are situated 
with the equator near a free surface and with their equatorial plane parallel to it ; here the 
same surface tension which has given a pinacoid form to the epithelial cells has deter- 
mined the oblateness of the spheroid ; on the other hand, the toxaspires, which suggest 
a prolate form for the ellipsoid, occur only in the cortex in Chrotella, the spires of the 
choanosome in this sponge retaining the more usual sigmoid form. The cortex in this 
case exhibits the usual signs (tangentially directed fusiform cells) of tangential growth, 
which has in all probability led to an elongation of the scleroblast in its own direction. 
The spiraster, which must next engage our attention, offers much that is suggestive 
and much that is difficult of explanation. The spines proceed only from the outer side 
of the spire, and the inner side which partly surrounds the nucleus is smooth or devoid of 
spines ; this is in correspondence with the general behaviour of the nucleus, and suggests 
that the region of deposition in the scleroblast is as we have supposed situated near the 
surface ; the production of spines, which may be regarded as due to radial tensions, at 
first suggests a difficulty, which can only be met by hypothesis ; we have to suppose that 
the scleroblast, at first growing in such a manner as to produce an oblate spheroid, as we , 
should expect from its position near the epithelial surface, subsequently altered its rate 
of growth in different directions so as to enlarge in every direction at an approximately 
uniform rate ; the sigmaspire, thus becoming immersed in the scleroblast, would grow 
outwards in a radial direction, and if the tensions were uniform a radial lamella might 
be expected to result; that the tensions should be uniform, however, is exceedingly 
unlikely, considering the general want of uniformity in nature, and thus instead of a 
continuous lamella a discontinuous one, or in other words a series of spines, results. We 
have here supposed the scleroblast to continue to grow with a more or less spheroidal 
surface as it certainly does in the sterraster, but it is possible that under the influence of 
tensions in the organism it may itself assume a radiate form ; at any rate the spines once 
developed very soon betray the influence of such tensions, which lead to the trans- 
formation of this spicule first into the metaster, then into the plesiaster, and finally into 
the aster. Now it is a most important point, and one which we have already indicated, 
that in the Theneidse, in which this transformation can be traced through each successive 
stage, the passage from the spiral to the astral form is accompanied by an increase in size. 
This association I believe to be in the nature of cause and effect ; for as the increase in 
size takes place the spicule is increasingly brought under the action of the tensions 
existing in the organism as distinct from those of the cell itself, now these tensions in 
a mass of collenchyma must always be symmetrically distributed about any given point, 
and thus when the actines are definite in number we shall find them symmetrically 
disposed also ; if two actines occur they should be directed along the same axis, and thus 
the preponderance of monaxons among the diactinate forms of aster results ; if three are 
