158 Prof. W. J. Sollas on the 



trunk, which traverses the cortex in a manner not yet ob- 

 served, and then continues over its surface beneatli a tunnel- 

 like extension of the skin in the manner previously described, 

 finally terminating in the oscuhim. 



The Sl'eleton. — The main spicules, which arc collected into 

 fibrcKS, are developed in granular spicule-cells, as will be more 

 fully described in treating of the embryo. The cortical ace- 

 rates and the fibrous layer are clearly homologous with the 

 globate and fibrous layer of Geodina ; and the inference is 

 also deducible that the cortical acerates are likewise homo- 

 logous with the geodine globates ; surj^rising as this infer- 

 ence is at first sight, it is ])artly su|)ported by the fact that both 

 are developed in remarkably similar nucleated cells. Again, 

 as the trichite sheaf is homologous with the globate, so it is 

 also homologous with the TettUa acerate ; and here we are 

 brought to see the essential difference between the sheaf and 

 the acerate, the former being a fibrillated rod and the latter a 

 concentrically-layered one. 



The hamate spicules are found embracing a small round 

 nucleus with a little granular sarcode ; but no cell-wall is ever 

 seen (PI. VII. figs. 4 and 10) ; so that one is led to conjecture 

 that the spicule may be the cell-wall, especially as it closely 

 resembles in size and appearance the annular cell- wall, to which 

 we have made frequent reference in describing the dermis. 

 The nucleus of the hamate spicules is entirely different in 

 size and character from that of the large spicule-cells and of the 

 geodine globate; it is much smaller, no larger than the 

 nucleolus of the latter, and shows no distinct nucleolus. It 

 is of importance to notice that the hamates are not developed 

 several in one cell, as Carter has asserted of the common 

 tricurvate spicules, as Schmidt has shown for the trichites of 

 Es2)eria, and I, subsequently, for the trichites of SteUetta 

 Normani ; each hamate has sole possession of its own nucleus ; 

 in other words, each hamate cell produces but one hamate 

 spicule. Carter mentions that he has also found two examples 

 of a bihamate occurring singly in its mother cell (A. & M. N. 

 H. 1874, vol. xiv. p. 104, pi. x. fig. 11). 



The Ernhryo. — The segmentation of the ovum has not been 

 observed in any of its stages ; but sections of three embryos 

 are shown in very thin slices ; they each lie in a brood- 

 chamber lined by a distinct membrane (endothelial) and a thin 

 layer of fibrous tissue. Two, nearly spherical and 0-033 inch 

 in diameter, are still solid throughout ; the third (PI. VII. 

 fig. 1), oval, measuring 0-04o inch along its major and 0*033 

 inch along its minor axis, is also solid, except for the presence 

 of the subdermal cavities, wliicli are well developed over one 



