386 Scientific Proceedings, Royal Dublin Society. 



stria, a transverse ridge corresponding to another edge of the 

 rhombohedron would make an apparent angle of about 96°, and, 

 on the other, another transverse ridge of 60° to 75°, as observed. 



In the projection shown in fig. 7, p. 388, the edges correspond- 

 ing to the truncating striae make apparent angles of 94° and 40° 

 with the edge which furnishes most of the striae. 



As the solution of the spicule under observation continued, a 

 beautifully defined hollow rhombohedron made its appearance 

 (pi. XV., fig. 11), and traces of others were seen. Finally, the 

 spicule broke up into segments ; but I was unable to determine 

 whether these were bounded by cleavage planes. 



Cleavage. — The spicules of calcisponges usually break with a 

 conchoidal fracture ; but it seemed possible that traces of cleavage 

 might be obtained by heating the spicules to short of redness, and 

 then plunging into water. This was tried with the coronal acerates 

 of Orantia ciliata. 



After igniting, the spicules are perfectly white and opaque, 

 as seen by reflected light; mounted in glycerine, they become 

 transparent, are brown in colour, and finely granular in texture. 

 The edges of the spicules are broken (pi. xv., fig. 8) by straight 

 lines, the fractures being apparently those of cleavage ; cleavage 

 planes frequently traverse the spicule, dividing it into rhomboidal 

 fragments, and often it becomes entirely resolved in places into a 

 collection of navicula-like bodies (pi. xv., figs 9 and 10), which 

 appear to be rhombs with rounded edges. 



Form of Transverse Section. — This experiment enabled me to 

 definitely ascertain a fact I had long suspected, but had up to this 

 been unable to prove : the acerates of Grantia, as they roll over, 

 when seen under the microscope, appear to be greatly flattened 

 along two opposite sides, one diameter of a transverse section being 

 much longer than the other. So long as the spicules were trans- 

 parent I was not sure that this might not be an optical delusion ; 

 now that they were opaque, when seen by reflected light, the matter 

 could be put to a certain test, as it was, with the result of proving 

 that these spicules are in fact, as in appearance, far from cylin- 

 drical, being greatly compressed from side to side. The colossal 

 acerates are also of oval section, and possibly rhomboidal, the angles 

 of the rhomb being rounded off. At all events, one which had 

 been etched (pi. xv., fig. 7) showed, after it had broken into 



