SoLLAs — On Sponge Spicules. 385 



about 60° to 75°, and sometimes of about 96° (PI. XV., figs. 5 and 7) . 

 In some cases near the sharp ends of tbe spicules, which have been 

 eroded to a greater depth than the thicker central part, we find 

 projecting angles of evident crystals. These angles, which are of 

 faces and not quoins, measure usually from 64° to 77° as projected 

 by camera lucida on a plane surface (pi. xv., figs. 1 and 2). 



The sagittal spicules display a curious difference in the distri- 

 bution of erosion, interesting as showing the influence of crystalline 

 structure on solution ; their basal, or paired rays, are always well 

 etched with striee, and sometimes show as well a fringe of projecting 

 angles on the distal or both margins (pi. xv., fig. 3). The un- 

 paired ray, on the contrary, is never visibly eroded ; it appears to 

 enjoy complete immunity from the solvent power of balsam. 



Wishing to develop etch-figures at will, I experimented first 

 with Grantia, but found its spicules were too minute to give satis- 

 factory results. I then managed to extract from under the cover- 

 glass of my mounted specimen of Leuconia glomerosa (?) a bundle 

 of colossal acerates ; after washing the balsam away with ether 

 they were placed in dilute acetic acid, and the process of solution 

 watched under a Zeiss D. Strise were soon developed similar to those 

 produced by the balsam ; the groups of striee being margined by one 

 of themselves, strongly developed, and truncated by a ledge which 

 sometimes made an angle of about 77° with them, the acute angle 

 of a face of a rhombohedron ; in some cases three ridges were 

 produced, giving a triradiate figure with angles of about 108°, 

 125°, and 127° : such a figure would be produced by the edges of 

 a rhombohedron placed similar to that shown in the projection, fig, 7, 

 p. 388, where the corresponding angles are 104°, 116°, and 140°. 

 The large difference in these angles and those just noted is partly 

 to be accounted for by a slight difference in the position of the 

 two rhombohedrons, and partly by the difficulty of obtaining exact 

 measurements of angles determined by such extremely short lines 

 as those of the etch-figures. Partly, also, the plane on which the 

 figures were projected by the camera lucida would not be exactly 

 at right angles to the reflected rays of light, and some distortion is 

 inevitable. It is clear, however, that the etch strise are the edges 

 of rhombohedrons, of each of which four faces, lying in a zone, are 

 elongated to such an extent as to convert the rhombohedron into 

 a prismatic fibre. On one side of an edge, exposed as an etch 



