74 W. WOODLAND. 



different species of animals in all cases appreciably differ in 

 chemical constitution from each other/ and even supposing 

 that the substance of all calcareous spicules^ e. g. is constant 

 in composition, the different colloidal natures of the various 

 organisms in which spicules are found are in all probability 

 amply sufficient, in conjunction with factors (a) and {h), and 

 the different crystalloid media present to account for the 

 various spicule forms encountered. ^ 



As showiug in some degree the complexity of this subject 

 of the causes determining the forms of spicules I may instance 

 the extraordinary nature of even the common process of 

 simple crystallisation, as ascertained by Frankenheim, 

 Vogelsang, and several other observers. Vogelsang observed 

 that the first visible stage in the formation of certain crystals 

 (sulphur crystallising from carbon bisulphide solution, e. g.) 

 was the appearance of liquid globules, which subsequently 

 aggregate to form small solid isotropic spheres which he 

 termed glob u lit es, these again arranging themselves in de- 

 finite patterns successively coalesce to form still higher aggre- 

 gates (crystallites, margarites, etc.) until finally the 

 crystalloids (the "integrant molecules " or crystal particles) 

 are produced.^ It would be interesting to know whether the 



1 Also the well-known fact that different tissues and regions of the 

 substance of the same organism differ widely in chemical constitution must 

 be remembered, this possibly largely accounting for the different forms of 

 spicules found in different parts of the same animal (Chiridota and other 

 Holothurians, Hexactinellids, etc.) ; moreover, in some cases, as Maas (16) 

 has shown in Tetliya, e.g. the scleroblasts differ among tiiemselves as regards 

 total size, size of nucleus, character of cytoplasm, etc., and this consequently 

 introduces another cause for the diversity of form of spicules contained in the 

 same organism. 



2 Arguments for the inheritance of spicule forms, based upon the numerous 

 transitional forms connecting together spicules of different types, are of little 

 value, since all crystallomorphs exhibit exactly the same phenomenon on a 

 very large scale, and, as is well known, all transitions exist even between true 

 crystals of very different shapes. Moreover, in the case of spicules, there 

 exist more causes to produce these transitional forms than in the case of 

 crystals or crystallomorphs. 



' Dr. Rosenheim [22j has brought to my notice certain incidental 



