76 DIFFERENTIATION III 



the cytoplasm, though the organs themselves of course are not. 

 But though preformed they are not prelocalized since the 

 apical sense-organ in Cerebratulus and Dentalium they need 

 not be initially present in their ultimate position. 



On the other hand it is possible to divide the ovum into 

 totipotent parts, which may be either fragments of unseg- 

 mented eggs or isolated blastomeres. This possibility obviously 

 depends on the arrangement in the cytoplasm of the various 

 necessary materials, which must be such that a fragment or 

 blastomere can receive a portion of each. It decreases with 

 time in both cases, though for different reasons. The totipo- 

 tence of egg-fragments diminishes apparently because there 

 is a redistribution of the materials during maturation and 

 fertilization, when for instance a bilateral replaces a previously 

 radial symmetry, but isolated blastomeres cease to be able to 

 give rise to whole embryos or larvae so soon as the cell- 

 divisions fall in such a way as to divide the cytoplasm into 

 unlike portions. Where the cell-plasma is highly differentiated 

 this occurs at once, as in the bilateral eggs of Cynthia and the 

 Ctenophores and usually in Molluscs : where the structure is 

 polar, but not markedly bilateral, cells separated by meridional 

 divisions are totipotent, those produced by equatorial or lati- 

 tudinal divisions less so or not at all (Sea-urchins, Nemertines, 

 Amphioxus). The case of A scar is is very instructive : nor- 

 mally the first division is equatorial, but in the centrifuged 

 egg meridional : in the former the half -blastomeres develop 

 partially, in the latter totally. The egg of Cerebratulus pro- 

 vides a similar example ; for the second furrow, which is nor- 

 mally meridional, may be made equatorial by pressure at right 

 angles to the egg-axis (Yatsu). The quarter-blastomeres 

 which in the first case are totipotent, are now no longer so, 

 since two are animal and two vegetative. Lastly, where the 

 egg-structure is radially symmetrical about the centre (as in 

 Hydromedusae, with the exception of Carmarina), the cells 

 remain totipotent as long as the divisions are perpendicular to 

 the surface, but when tangential divisions occur by which the 

 endoderm is delaminated from the ectoderm, this capacity is 

 lost ; at least it is known that these two germ-layers cannot 



