FOEMS OF SPICULES. 61 



these scleroblasts are waudering-cells. Now tlie theory of 

 the inheritance of spicule forms asserts at the very least that 

 this scleroblast or scleroblast-fusion inherits the property of 

 becoming enormously distended in size, of so distending- as 

 to assume externally the form of a spicule adapted in its main 

 features to its immediate envii'onment, and of correspondingly 

 forming' an internal mould necessarily identical in shape in 

 every detail with the spicule to be deposited, which shape, as 

 just implied, varies according to the position in the organism 

 which the wandering-scleroblast happens to occupy at the 

 time of deposition. 1 And these wandering-cells termed 

 scleroblasts are supposed to inherit the property of undergoing 

 all these changes (in the main d'fefinitely related to the archi- 

 tecture of the rest of the organism and complex in the 

 extreme) without having any connection with the other com- 

 ponent cells of the body — an assumption which, if my above 

 remarks are true, is quite unwarranted, to say the least. This 

 hypothesis of the inheritance of the forms of spicules, in fact, 

 either contradicts the ascertained truths of experimental 

 embryology or implies that scleroblasts are not wandering- 

 cells. 



It is true that in the development of Balanoglossus, 

 annelids and other animals, masses of cells are stated to be 

 often completely detached from the rest of the embryo, and 

 that these sometimes undergo changes of form related to the 

 surrounding tissues whilst so detached, but it will invariably 

 be found that either these changes of shape are simple 

 mechanical adaptations to the architecture of the organism 

 (formation of ccelomic sacs, etc.), or are due to the fact that 

 these cells have quite recently been connected with the rest 

 of the organism, the changes of shape being adapted to the 

 general scheme of development because the mass of cells has 



dem Dampfer ' Valdivia,'" Bd. iv, 1904) that these spicules are completely 

 invested by scleroplasm. 



' A scleroblast can be supposed, on the hypothesis, to inherit the property 

 of giving rise to spicules of different shapes, according to its localisation in 

 the organism, on the principle of the " equipoteutiality " of blastomeres. 



