No. 2.] CLEAVAGE OF ARENICOLA CRISTATA. 83 



* 



not Y-Z or y^ larvae, but have become more or less complete 

 wholes. Wilson appears to have overlooked this point entirely, 

 in quoting this work in support of his views. He asserts that 

 "apart from a few unimportant details, blastomeres of the 

 two-cell or four-cell stage, whether isolated singly or in 

 groups, segment precisely as if the missing portions of the egg 

 were present, and the resulting larval fragment is completely 

 devoid, not only of the power of a regulatory rearrangement of 

 its material, but also of regenerative or post-generative power. 

 The cleavage is thus demonstrated to be in the gasteropod 

 precisely what I have asserted that of Nereis to be, viz., "a 

 visible mosaic-work." 



Even here, then, in this highly specialized type of cleavage, 

 the fate of the cells has been altered by a change of conditions. 

 If we accept Roux's " Reserveidioplasson," it is of course pos- 

 sible to escape from this difficulty. 



The oblique cleavage is not, then, a direct division of the egg 

 into so many sets of organs which exist as " Anlagen " in the 

 cells, but the organs are differentiated as the result of proc- 

 esses going on in the egg as a whole, though, later in devel- 

 opment, self-differentiation may occur to some extent. The 

 cell is after all no true morphological unit in cleavage, as, 

 indeed, an increasing number of facts is showing us, e.g., the 

 papers of Hammar ('97b) and Andrews ('97a), and it is difficult 

 to believe in the face of these and other facts that it is 

 physiologically isolated. 



I believe that any theory of development, proceeding on a 

 strictly cellular basis, must fail in its attempt to explain ontog- 

 eny. The organism from the unsegmented egg to the adult 

 is a whole, and at every stage of its existence acts as such 

 ('93b). There is often visible organization in the unsegmented 

 egg, but this, by no means, corresponds to the adult organiza- 

 tion. That organization is the final result of the processes 

 constantly going on in the developing egg, and changes occur- 

 ring in one part are the result of the preceding changes in the 

 other parts and in their turn become causes of others. 



Cleavage, however, if not necessarily cell-differentiation, is 

 not merely a mechanical splitting up of the egg, nor is its form 



