Genus of Fossil Hexactinellid Sponges. 21 



by polarized light. When this is done, a change in the plane 

 of polarization is distinctly produced by the fibre, since it 

 shines out with faint bluish and yellow glimmerings on the dark 

 ground produced by crossed prisms. If now some recent spi- 

 cules, or some compound Vitreohexactinellid fibre, be substi- 

 tuted for the fossil silica, no effect will be produced on the 

 light : the dark ground will remain wholly dark ; and if the 

 polarizer be turned round 90°, the light admitted will under- 

 go no change of colour in passing through the object. 



One may diverge for a moment here to speak of some addi- 

 tions to the modes of examining recent sponges which arise 

 out of these observations. First, the fact that the recent spi- 

 cule is almost invisible in glycerine jelly, while the horny fibre 

 of sponges is more than usually well defined in it, allows us 

 to optically despiculize the fibres of the Chalinida and Ecliino- 

 nemata (Carter) by immersing them in this substance, and 

 thus to observe the kerataceous material independently ; and, 

 next, the fact that the calcareous spicules of the Calcispongiffi 

 do produce a marked effect on polarized light, exhibiting bril- 

 liant colours, Avhich siliceous spicules and fibre do not, provides 

 us with a speedy method of distinguishing between these two 

 kinds of spicules, and one which may be employed in cases 

 where the use of acid is not available*. 



I cannot attempt to explain all the various mineral changes 

 and replacements which we have now described ; they are as 

 obscure as most of the pseudomorphic alterations w^iich occur 

 in fossilization ; but two most important facts stand out from 

 all the rest in my mind: — first, that siliceous fibre may be com- 

 pletely replaced by carbonate of lime without obliterating its 

 structure ; and, next, that spicular silica may with lapse of 

 time pass from the colloidal to the crystalline state. 



Alliances. — In looking for the existing relations of Stauro- 

 nenia one will not find any near ones. The absence of a 

 " lantern " about the nodes excludes the Ventriculites ; Euplec- 

 tella is characterized by ladder-formed fibre, and is in most 

 respects widely divergent. With Apkrocallistes the oscular 

 plate presents some analogy, as pointed out to me by Mr. 

 Carter, the walls of the tube-net of Apkrocallistes being per- 

 forated completely by horizontal excurrent canals quincuncially 

 arranged, just as we found in the plate-net of Stauronema ; and 

 even, as in the latter the oscules become covered up with a 

 layer of fine network, so a network, but not correspondingly 

 fine, extends itself over the oscules of Apkrocallistes^ as may 



* Mr. Carter points out to me that this latter observation has been pre- 

 viously made by 0. Schmidt. 



