REPORT ON THE TETRACTINELLIDA. 
Ixxix 
they are converted into centrotylote microxeas. While a rhabdal form is thus produced 
in both sterraster and euaster, it is of interest to note in passing that the modification 
is accomplished in two different ways, — in the euaster by the reduction in the number 
of the actines, in the sterraster by a shortening in the length of those which in the 
euaster would be suppressed. 
The chief forms of microscleres have now been discussed, and the megascleres may 
next engage our attention ; these are clearly derivable from the microscleres by increased 
growth, and that such has been their origin there can in my mind be no doubt. 
We have already seen that as the microscleres increase in size they become increasingly 
subject to the influence of strains in the organism, while the intracellular tensions become 
less and less effectual, on passing to the still larger megascleres we should naturally expect 
to find this tendency still more marked, as indeed we do ; the tensions of the organism are 
in this case paramount, and the intracellular tensions may for the future be disregarded. 
The tensions existing in the sponge as a whole offer such a complicated problem for 
study that we can only attempt to treat it in a most general way ; it will, however, be 
clear that the growth of the sponge leading to an increase in surface and thickness may 
be regarded as taking place along radial lines and a superficial plane, the radial lines 
will furnish one direction of least resistance, the superficial plane others, the precise 
distribution of which will be considered when we come to treat of the trisenes ; for the 
present we may simply regard them as transverse to the radial lines of growth. The 
mode of growth of the choanosomal folds will be influenced by the relations between the 
increase of surface compared with the increase in thickness of the sponge, and thus some 
of the folds will increase radially and others transversely, concrescence between the 
longitudinal folds will give rise to radial tracts of mesoderm, between the transverse 
folds to transverse tracts, and along these tracts lines of least resistance will exist, 
probably lines of tension ; their existence is indicated by the fusiform elongation of the 
collencytes when the tissue of the tracts, as is usually the case, consists of collenchyma. 
The direction of tension along any tract will probably undergo more or less change as it 
is traced from point to point, but the spicules are as a rule so short compared with the 
diameter of the sponge that in most cases, but not all, this may be disregarded, and the 
tension may be assumed for all practical purposes to act along a straight line. If now 
we take the case of the Tetillidse in which the megascleres are supposed to have been 
derived from sigmaspires, and consider one of these spicules to undergo a vast increase 
of growth within the radial tracts of collenchyma, it will evidently be exposed to the 
action of a couple acting at its ends, and we should expect it to be elongated into a 
rhabdal form lying with its axis in the line of tension, ^.e,, radially, and this is the 
position which, as a matter of fact, the spicules of these tracts invariably assume. The 
transverse tension has next to be satisfied, and in the lower forms of the Tetillidae we 
find oxeas scattered transversely to the radial spicular fibres, which the radially directed 
