288 Mr. W. J. Sollas 07i the Action of Caustic 



specimens of Farrea occa and F. jacunda exist, and which 

 occurs sometimes in Aphrocallistes Bocagei and, according to 

 Bowerbank, in Pm-isiphonia Clarkei^ while the second stage is 

 exhibited in skeletons of Ajjhrocallistes Bocagei^ Dactylocalyx 

 jpumiceus^ and D. crisjyus (O. S.), which, however, like the 

 genera Iphiteon (Bk.) and Myliusia^ are usually found in the 

 fresh state and consequently with solid fibres. 



As regards the characters of the canals themselves, they 

 are in my specimens exclusively sexradiates ; but they differ 

 in a marked manner from the loose separate sexradiates which 

 do not become involved in the siliceous fibre ; for while the 

 latter are always rectangularly triaxial, the involved spicules 

 are neither rectangular nor triaxial at all, except in the rarest 

 cases. Each arm of an imbedded spicule may make any angle 

 with its fellows ; there is no constancy in this respect, nor any 

 approach to it. This will be seen by reference to fig. 1, PI. 

 IX., and is even more strikingly exhibited in some specimens 

 which are not here figured. 



It would seem as though the free sexradiate spicules of the 

 sponge lost their usual rigid regularity and became instead 

 perfectly pliable to all the changes in direction of the siliceous 

 fibre as soon as they became involved in it, the fibre governing 

 the direction of their rays to the complete subordination of 

 their intrinsic tendency of growth. 



On the other hand the centre of each sexradiate spicule 

 appears in most cases, but not in all, to determine the position 

 of a node of the fibrous skeleton, each node corresponding in 

 position to the centre of an imbedded spicule. The exceptions 

 are where three arms of a sexradiate are 

 bent in one direction approximately Fig. 1. 



parallel to each other, and the other 

 three in the opposite direction, so that the 

 whole spicule comes to be imbedded in 

 one and the same straight fibre (fig, 1). 



Here, then, a sexradiate exists without a corresponding 

 skeletal node ; and more frequently still it happens that a node 

 may exist independently of the presence of a sexradiate centre, 

 ex. gr. where three or more arms of adjacent sexradiates join 

 together. 



It need not be added that though the nodes of the skeleton 

 are generally determined by the ])resence of sexradiate spi- 

 cules, they are not on that account necessarily sexradiate 

 themselves, since by the coincidence in direction of two or 

 more arms of one spicule, or by the addition of rays from an 

 adjacent spicule, a node may become four-, five-, six-, seven-, 

 or even eio'ht-rotulate. 



