Stromatoporoids and of Eozoon. 457 



like Eozoon should have a calcareous skeleton and why other 

 lihizopoda should have a siliceous one. 



I think we must go back to the cooling minni-al mao-ma 

 which wmild varv greatly in character over diffjretit areas 

 and which would form shall )wer or deeper trouo-Iis, Probably- 

 at first, before diffusion produced more or less uniformity, 

 seas filling those trouglis w )uld vary in composition. Rocks 

 are classilied as acid or basic according to the percenta -e ot" 

 silicic acid present in them, and similarly a classification 

 based on mineral constituents might have been extentled to 

 primitive seas (just as doctors classify mineral waters). 

 Sarcode living in seas with a high percentage of silica and a 

 sm dl one of calcium salts would become saturated with the 

 former mineral. Apparently silica forms a more intimate 

 union with protoplasm than carbonate of lime does and 

 though isotropic when thus u lited, seems to impress some of 

 its mineral characters on that protoplasm, as witness the 

 beautiful symmetry of Radiola'ia and of Hexactinellid 

 spicules*. Acanthin of R idiolaria seems to be halfway 

 bcitween silica-saturated protoplasm and silica of the skeleton. 

 Organisms with skeletons of silica, probably on a:,'ouiit of 

 their want of flexibility, have oidy travelled a short (lisbaiice 

 along the path of evolution. The calcareous animal oriTauisnn 

 on the other hand, have deposited their skeletal material in 

 masses (acicula, dorsal pillar, limb-girdles, &c.) udiich have 

 served us points iVappid tor contractile protoplasm, and thus 

 they have been enabled to come itito relation with a varied 

 en\iionment. Consequently the path of evolution from the 

 Dawn Animal to human civilization has been alou"- t!ie 

 calciuvous way. 



* For some time past I have tboug-ht that possibl}' sponj>'es are tri- 

 phyletic and that there were at least three ancestral olynthuses (and not 

 one olyathiiri), viz. a shallow-water calcareous, a shallow-water pi e-Demo- 

 spiuge, and a deeper water Hexactinellid olyuthiis, all three orii^iuatin(»- 

 Iroiu colonial VoLoox-Yikii Flairellates or Choanotlageilates. In all three 

 a free-swimming phase would be followed by a fixed one brought about 

 by the increasing disproportion between weight and carrvino- capacity. 

 On coming to rest the little organism would form a disk with an inferior 

 layer of granular cell^ and a superior layer of tlagellate cells. The 

 increasing growth of the lower nutritive layer would soon cause it to 

 enco.iipass the upper motile layer, which would sink into it and an 

 olyntluis would be formed. 



The Cho.molLigdllate pre- Hexactinellid was highly vacuolated and its 

 delicate reticulate strands would be best sustained by a typo of spicule 

 having three axes crossing a common centre at right angles, like building- 

 scaffolding, and the loose-textured protoplasmic network would permit of 

 the formation of large ti igellated ch imbers. (I would add that I hold 

 this homoiousion heresy very liglitly.) 



Ann. & Mag. N. Hist. Ser. b. Vol. x. 31 



