436 Prof. W. J. Sollas on the 



which corresponds really to the side of other specimens. 

 These last two specimens may be instances of a local variety, 

 which, however, I shall not dignify by a name, as Schmidt 

 has his T. fenestrata. The size of the sponge averages about 

 1-5 inch ill diameter by 1'2 inch in height; the rooting-fibres 

 extend downwards for 0-8 inch usually before fraying out. 

 Mr. Norman's largest specimen measures 3 inches by 2'5 

 inches in width and breadth, by 1"5 inch in height. 



External Surface. — The outer surface of the sponge is 

 felted and. thatched by obliquely-projecting, long, slender, 

 acerate, and grapnel-shaped spicules. Round the middle of 

 the upper half, midway between the oscule and the lower 

 edge (tegminal edge we may call it, since it covers or roofs 

 over the equatorial recess) , the spicules, lying prostrate almost 

 parallel with the surface, point this way and that, and by 

 their intercrossing form a loosely-felted thicket above the 

 skin — the home of all kinds of animals, Foraminifera (some 

 form of which covers the surface with long strings of sand- 

 grains) , Ascidians, worms, and Crustacea. Above this zone the 

 spicules, still projecting obliquely from the skin, point directly 

 towards the apex, so that within a radius of half an inch from 

 it they form a close, regular, but inverted thatch, the free 

 ends of the spicules projecting upwards, and those immediately 

 around the oscule fencing it in with a forest of bristling 

 points. Below the middle zone the spicules proceeding ob- 

 liquely from the skin point directly downwards towards the 

 tegminal edge, beyond which they project in a fringe of long 

 line lashes ; the thatch is here in the right direction ; and the 

 fringe reminds one of the uncut straw hanging over the eaves 

 of a cottage. The lower half of the sponge is covered by 

 obliquely-projecting spicules, showing no regularity in direc- 

 tion, except opposite the tegminal edge ; here they point 

 upwards and intercross with the spicules descending from the 

 fringe, forming with them a defensive sieve of great efficiency. 



Great variation exists in the distribution and disposition of 

 the spicules as just described ; sometimes projecting acerates 

 seem contined to the margin of the oscule and the tegminal 

 edge, or even to the oscular margin alone. Probably in some 

 of these cases the spicules have been lost since the specimen 

 was obtained ; in others, on the contrary, they seem never to 

 have been present. Owing to one or other of these causes, 

 i. e. abrasion or non-development, or to both, projecting spi- 

 cules are usually absent over a large part of the skin, the 

 outer surface of which is then clearly exposed to view ; it has 

 a greyish tint in spirit-specimens, is often nearly pure white 

 in dried ones. Examinuig it with a lens, we perceive the 



