FORMS OF SPICULES. 59 



this supposition that the forms o£ spicules are inherited is 

 still a desideratum. 



Although, therefore, we cannot as yet state decisively that 

 the forms of spicules are not inherited, yet there exist certain 

 facts which appear to justify us in provisionally making this 

 statement.^ In the first place the results of recent inquiries 

 in experimental embryology prove that for any given cell 

 (blastomere) of a growing organism to be able to produce a 

 living part of the adult organism adapted in form to the other 

 parts (an organ), this must be connected with the 

 other cells. The part which a given blastomere plays in 

 the formation of the adult organism is a function of its con- 

 stant position relative to the other blastomeres, i. e. a function 

 of its localised connection with the other blastomeres. 

 Detach a blastomere from its fellows, and it either gives rise 

 to a complete though diminutive organism or becomes a 

 wandering-cell or a germ-cell ; in no case does it give rise to 

 a part of the adult organism adapted to the general economy 

 in structure and relative position, i. e. an organ.'^ And so far 



(Auricularia wheels) and Theel (spicules of Elasipoda) were, without doubt, 

 merely cases of decalcification, as I have elsewhere pointed out (Study IV). 



' lu what follows the general tone of the argument would imply that I 

 allogether reject as unwarranted the hypothesis of the inheritance of spicular 

 forms. This is not the case. I have taken up, somewhat vigorously perhaps, 

 the opposite position to that held by most zoologists who have worked at 

 spicules solely in order to see what can be made of that position, and, rather 

 to my surprise, I find that there is a good deal to be made of it. For 

 example, in connection with my main argument, based upon the results of 

 experimental embryology, if it should turn out that the forms of spicules are 

 inheritable, it seems to me that, in view of these additional facts, the problem 

 of heredity will be rendered more complex than it already is. I do not reject 

 as unwarranted the inheritance hypothesis in the case of spicules, for the 

 obvious reason that at present we are so ignorant as to what heredity can and 

 cannot do. Nor, without definite proof to the contrary, is it justifiable 

 to summarily reject the convictions of those who have worked for years on 

 siliceous sponges e.g. I think it right, however, that the objections to the 

 inheritance hypothesis here elaborated should be published (see Note at end). 



- Occasionally, as in Cteuophora, detached blastomeres, in virtue of once 

 having been joined to the other blastomeres in a definite position, produce 



