286 Mr. W. J. Sollas on the Action of Caustic 



in 1868, stated it as his opinion that the network of the sili- 

 ceous-netted sponges was produced by an anastomosis, fusion, 

 or coalescence of sexradiate spicules with one another. 



This opinion was at once endorsed and supported with 

 various new observations by Oscar Schmidt. The objection 

 urged by Dr. Bowerbank to Gray and Thomson's view was, 

 that true spicules never evince any tendency to fuse together : 

 thus in 1869 he states that he has " never yet seen a case of 

 the anastomosis of spicula. The normal condition of these 

 organs is never to anastomose, however closely they may be 

 packed together." 



So the question stood till 1873, when it was reserved for 

 Dr. Carter to bring these opposing views into harmony, and 

 so to explain in the most beautiful manner the real structure 

 of the netted Hexactinellidse. In a paper on the Hexacti- 

 nellidae and Lithistidte he proves, from observations of very 

 young specimens, that at first the netted sponge Aj)hrocaUistes 

 possesses no other skeletal elements but separate free sex- 

 radiate spicules, and that the formation of a network is a sub- 

 sequent process brought about as growth advances by the 

 envelopment of the free spicules in a coating of siliceous 

 material, which, running over each and from one to another of 

 them, at length involves them all in continuous siliceous fibre. 

 The network of Aphrocallistes and of other netted sexradiate 

 sponges is accordingly a composite structure, not simply 

 fibrous in the sense that the horny skeleton of a washing- 

 sponge is so, nor, on the other hand, simply spiculose, as if it 

 consisted of ankylosed spicules merely, but spiculo-fibrous 

 like the network of the Chalinids ; only in this case the simple 

 acerate spicules of the Chalinidse are represented by sexradiate 

 ones, and the horny substance of their fibre is replaced by 

 siliceous material. 



This interpretation was supported by the fact that the 

 original spicules on which ^the fibre of the network is built 

 become absorbed some time after the death of the sponge, and 

 so reappear in the interior of the fibre as hollow sexradiate 

 casts, which in time, however, may, with continued internal 

 absorption of the fibre, so increase in size as to become con- 

 tinuous axial canals, in which state they seem to have been 

 observed by Dr. Bowerbank, and so came to be regarded by 

 him as a proof of the existence of normally fistulose fibre 

 amongst the siliceous sponges. It is important to add that, 

 besides casts of ordinary sexradiates. Carter also observed 

 those of other forms, notably of the curious besom-shaped 

 spicule which is so characteristic of Aphrocallistes. 



As regards the cause of the absorption by which the im- 



