DIFFERENTIATION 675 



ant ; for example, the larvae of many marine bottom inverte- 

 brates only metamorphose when they receive the mechanical 

 stimulus of a suitable substrate for the adult. 



At least up to a late stage in development the division of the 

 nucleus at each cell division is mitotic (p. 688) and equal, so that 

 the differentiation of cellular type is not directly due to differ- 

 entiation of the chromosomes, and it is very unlikely that any 

 process akin to mutation (p. 720), which is rare and irregular, 

 could take place with the frequency and regularity with which 

 embryonic cells change their form. Yet cells, as we have seen in 

 Chapter 26, acquire irreversible characteristics, so that it seems 

 that there must be important changes in the cytoplasm. The 

 only other possibility is that the chromosomes change under the 

 influence of their environment ; this cannot be disproved, but 

 where, as in many of the lower animals and in almost all plants, 

 regeneration of a whole organism can take place from a piece of 

 specialised tissue, it is unlikely, and any suggestion that the 

 nucleus can be altered in any regular way by outside influences 

 is rejected by modern geneticists. We must therefore seek factors 

 which can influence the cytoplasm, and some of these are known. 



The earliest specialisations of the embryo are its main axes, 

 and these are determined by accidents of position. In the frog, 

 for instance, the fore-and-aft axis is determined by the position 

 of the developing egg in the ovary ; the part of the egg which is 

 deeper in the ovary is that in which yolk is laid down, and this 

 becomes, as we have seen, the posterior part of the embryo. 

 In most instances known, the dorsal-ventral axis is determined 

 by the presumably accidental point of entry of the sperm ; 

 in the frog this side becomes ventral. 



The effects of spatial distribution of material within the egg 

 go deeper than the mere mechanical action of the yolk in delaying 

 division. In the molluscs, as described on p. 672, the cytoplasm 

 which is to form mesoderm is already recognisable in the egg, 

 and the same is true of many other animals, especially the 

 tunicates, where five distinct zones can be recognised (Fig. 529). 

 In animals like the frog, where this is not so, a separated blasto- 

 mere of a two-celled, or even later, stage may develop into a com- 

 plete individual, but sooner or later specialisation sets in. This 

 happens chiefly under the influence of a group of cells, itself 

 determined by the accidents of position, which acts as an 

 organiser for the rest of the embryo. In vertebrates (and especially 



