60 W. WOODLAND. 



as it is legitimate to assert such a thing it is impossible that 

 it should ; au ovum is a mechanism which evolves into an 

 organism — " crystallises out " so to speak — and any portion 

 which separates off necessarily remains uninfluenced by the 

 "^ crystallising out " ^ process. The whole fabric of 

 Weismannism is based upon this assumption, and whether 

 the superstructure be sound or not no one doubts but that the 

 assumption is valid. Moreover, all histological reseai-ches of 

 recent years confirm the supposition that, apart from the 

 various classes oE wandering-cell s, tlie whole organism is a 

 syncytium. Now the theory of the inheritance of spicule 

 forms asserts that most scleroblasts (not all, since many 

 spicules are not in any way adapted to the architecture of 

 the organisin) are unique in this respect, being able, though 

 having lost connection with the rest of the organism, to give 

 rise to a complicated structure (an organ — the spicule-mould) 

 adapted in shape and function by inheritance to 

 that part of the organism in which it happens to 

 be situated. Take, as a concrete example, one of the com- 

 plicated siliceous hexactines of hexactinellid sponges so 

 marvellously adapted to fit in, so to speak, with the architec- 

 ture of the sponge-wall which it inhabits. The hexactine, 

 like other spicules, has undoubtedly originated as a simple 

 deposit within one or more scleroblasts (see Ijima" [11]), and 



separately that part of the organism which they would have done if unde- 

 tached. But this capacity would be of no use in the case of a wandering 

 cell which, as the term implies, is constantly changing its position, since tlio 

 organ produced would not be in relation with the rest of the organism — in 

 Ctenophora it would be futile e.g. for a detached blastomere to produce say 

 an eighth of au embryo in one of the tentacles (see Wilson's "Tlie Cell," 

 chap. ix). 



' This likening of the developmental process to crystallisation is a truer 

 analogy tiian might be supposed. In the formation of complex aggregate 

 crystals (the familiar fronds on the window-pane, e. g.) the co-ordination of llie 

 multitudinous simple crystals to form one pattern is only possible by mutual 

 contact between the constituent crystals ; the least separation leads to tlie 

 production of another centre of crystallisation. 



2 I am at present working on hexactinellid spicules, and can amply con- 

 firm the statements of Ijima and Schulze ('Hexactinellida,' "Expedition auf 



