246 W. J. SOLLAS ON PHARETROSPONGIA STEAHANI. 



filling them has been removed, so as to leave the skeletal fibres ex- 

 posed and bare, as in a deciduous recent specimen. 



The thickness of the fibres varies very considerably. The smallest 

 measure about 0-0025", and the largest 0-025" in diameter, the 

 average thickness being 0-006". 



The superficial aspect of the oscnlar network shows that its fibres 

 are wider than those of the interior, while these sections prove 

 that they are also thicker ; and although no definite canals can be 

 traced in the transparent sections, there is yet observable a general 

 tendency of the intermeshes to open more freely into one another, 

 and thus to form more open and less obstructed passages, in a direc- 

 tion crossing obliquely outwards from the poriferous to the oscular 

 surface ; and as these passages seem to become confluent in certain 

 of the intermeshes of the oscular network, to the exclusion of others 

 that have become nearly or quite obliterated by an enlargement of 

 the oscular fibres, we may conclude that the excurrent canals found 

 their way along these more open paths, and drifted together to open 

 into the intermeshes of the oscular network, which thus, though 

 minute, are really oscular openings. 



The Fibre. — Under a power of 60 diameters and by transmitted 

 light, the fibres present a transparent and colourless ground of crys- 

 talline calcite, which is finely striated in a direction more or less 

 parallel to the general direction of the fibre by fine black lines about 

 0*00035 of an inch apart. The apparent blackness of these lines is 

 due to their opacity, since, when viewed by reflected light, they shine 

 brightly with a pure white colour. 



Examined by a higher power (say from 150 to 500 diameters, though 

 with a very careful examination an enlargement of 60 diameters wiL 

 suffice) these striae are resolved into a perfect spicular structure ; 

 one sees clearly that the fibres of the sponge are composed of a 

 number of transparent acerate spicules which lie side by side, over- 

 lapping at their ends and separately defined by the fine granular 

 opaque white material, which coats them externally and produces 

 the black lines seen by transmitted light (PI. XI. figs. 5, 6, 7, 8). 



This is the appearance produced when the fibre happens to have 

 been sliced along its length ; when the section cuts the fibre trans- 

 versely, so that it appears as an isolated round transparent space 

 instead of a broad band, the striation vanishes and is replaced by a 

 " dotting " of minute transparent circles, defined from each other by 

 the fine opaque material before mentioned, which either thinly invests 

 each of them or may be present in sufficient quantity to produce a 

 black ground with the transparent circular areas set in it (PL XI. 

 fig. 6, *). 



Since the spicules lie with their long axes in the same direction 

 as the length of the fibre, it is evident that a slice transverse to the 

 axis of the fibre must also cut the spicules transversely ; and since 

 the spicules are more or less cylindrical in form, the circular spaces 

 observed can be nothing else than the spicules seen in transverse 

 section. 



This explanation is supported by such appearances as that repre- 



