R. W. H. ROW—REPORT ON THE SPONGES : CALCAREA. 195 
entirely obviated if the further growth of the paired rays continued along the 
original lines occupied by the spicule, 7. e. in the primitive facial plane, each 
portion of calcite secreted being thus a little more out of line with the basal 
ray than the last. This would fully explain the peculiar curvature of the 
paired rays of the prochiact, and since the dermal and gastral cortical 
triradiates in Grantilla are very nearly regular, the oral angle between the 
rays of the original cortical triradiate is very similar to the oral angle of a 
prochiact. 
It is, however, more difficult to understand how the growth of the 
flagellated chambers can cause the change of position in the case of the sub- 
dermal triradiates and prochiacts, though the evidence in favour of a similar 
origin to that of the subgastral spicules is very strong. Like the similar 
subgastral spicules, they lie in the region of the oscular rim in Grantilla 
hastifera in intermediate positions between tangential and radial, while, as 
quoted above, Jenkin (9) states that in Tenthrenodes such a “tipping” is 
actually caused by the growth of the chambers. 
These arguments imply that the prochiact is more primitive than the sub- 
dermal sagittal triradiate, which will then be derived from it by the widening 
of the oral angle by the turning of the paired rays until they lie in one plane 
with the axis of the basal ray (fig. 1, p. 188); and the immediate derivation of 
one from the other is rendered certain by the absolutely complete series 
of intermediate forms which occur in Grantilla. This series is as complete 
on the subgastral surface as on the subdermal, and leads to the conclusion 
that the subgastral sagittal triradiates of Grantilla are derived, like the sub- 
dermal, vié prochiacts from the cortical triradiates. It is, of course, not 
suggested that spicules once formed change from prochiacts to sagittal tri- 
radiates, but that this change has occurred during the evolution of the 
group. 
It also leads to the conclusion that the subgastral sagittal triradiate of such 
a form as Grantilla, and probably also of the Heteropide, is entirely different 
from the sagittal triradiate of the first joint of the ordinary articulate tubar 
skeleton, being secondary in origin, and therefore that the inarticulate tubar 
skeleton is an entirely new structure, supplanting and replacing the articulate, 
instead of being derived from it. 
In dealing with the question of these secondarily formed sagittal triradiates,, 
the researches of von Ebner (19) into the physical characters of spicules are 
of very great importance. 
Von Ebner (19) studied this question in a large number (14) of species, 
including members both of the Homocoela and the Heteroccela, and deduced, 
from the behaviour of their spicules when examined under polarized light, the 
occurrence of “secondary” sagittal triradiates differing markedly in their 
optical orientation from the primitive sagittal triradiates. 
In examining a typical regular triradiate by polarized light, it is found that 
