XV111 INTRODUCTION. 



stance (flocculite) , whereas the dendritic bodies referred to are 

 unmistakably the indirect products of crystalline development. 



The point now introduced consists of the flocculent variety of 

 the fc canal system/' which, composed of a homogeneous or 

 structureless substance, is unlike in its arborescent forms ordinary 

 crystalline products ; while its branchlets resemble casts of tubes 

 and other cavities conceivably produced by tunnel-enclosed 

 vessels of an organism, as predicated of them by eozoonists. 



We must except in this connexion a variety of the " canal 

 system" consisting of malacolite, as it is palpable beyond dispute 

 that its branchlets have resulted from the decretioii of clusters 

 of irregularly arranged crystalloids of a mineral readily affected 

 by solvent action along its eminent cleavage-divisions, and thus 

 etched into shapes equally simulative of arborescent growths. 



But in the case of the flocculent variety of the " canal system" 

 we are dealing with bodies possessing forms which, although 

 we maintain that they have been equally produced by decretion, 

 cannot be positively said to have been predetermined by cleavage- 

 structure *. We do not propose to enter on an explanation of 

 this point ; all we have to do is to prove the existence of examples 

 simulative of organic features, in the elaboration of which crystal- 

 line forces alone have been concerned, and which are apparently 

 as unlike crystalline products as the typical examples of flocculent 

 " canal system." We refer to the foliaceous expansions charac- 

 teristic of the mineral silicate metaxite, also the coralloids and 

 other configurations of the Sunderland magnesian limestone. 



First, the configurations in metaxite have a strikingly organic 

 aspect ; and it is undeniable that they are purely of mineral 

 origin. It must also be remarked that this mineral, like floccu- 

 lite, is without internal structure ; properly speaking, it has a 

 pasty consistency. Moreover it is closely related to, if not an 



* Still we have already adduced the most conclusive evidence that rod-like 

 cylindrical processes in serpentine (undergoing change into flocculite) have 

 had their shape determined by rectangular prismatic cleavage. See Proc. 

 Roy. Irish Acad. vol. x. pi. xliii. fig. 8. 



