A PRELIMINARY EXPERIMENTAL ANALYSIS 39 



It appears that once a differential is established, the plane of 

 symmetry will thereby be determined, and that it will be deter- 

 mined just as efficiently by a feeble differential as by a strong one. 

 The possibilities of realising normal bilaterality are thus inherent 

 in the egg ; but the factors which determine the fact of its realisation 

 and decide its localisation are external^. 



§3 



The next step in development is cleavage, the splitting up of the 

 egg by cell-division into a number of smaller cells, the blastomeres. 

 Here, one of the effects of the axes already determined (the antero- 

 posterior, and the dorso-ventral) manifests itself in a differential 

 rate of activity and cell-division, and therefore a gradient in cell 

 size, from the animal pole with its small, actively dividing cells, to 

 the vegetative pole with its more sluggish yolk-containing cells; 

 and, at any given circle of latitude, the cells on the dorsal side 

 divide faster and are therefore smaller than those on the ventral 

 side, at any given time. As will be pointed out in Chap, ix, the 

 main organisation of the developing egg at this stage consists of 

 these quantitative gradients, or, as we shall call them, gyadient-fields. 



The rate of cleavage and subsequent differentiation can be 

 locally altered by subjecting the egg to differential temperature- 

 exposure: one pole or side hot, the other cold.^ 



The amount of yolk present in the vegetative hemisphere of the 

 amphibian egg, while responsible for the larger size of the vegetative 

 blastomeres, is not too great to prevent holoblastic cleavage of the 

 egg. It is possible, however, to make the cleavage of the frog's egg 

 conform to the meroblastic type characteristic of Selachians and 



^ It might be supposed that the bilateral symmetry of the egg, once established, 

 is necessarily identical with that of the resultant embryo. However, Jenkinson 

 (1907, 1909 a) by means of an elaborate biometrical study has shown that the 

 correlation between the two, though high, is not perfect: in other words, the grey 

 crescent does not always lie exactly in the future mid-dorsal line. Thus both the 

 determination of the grey crescent in the meridian of sperm-entry, and that of the 

 axis of bilateral symmetry of the embryo in the meridian of the grey crescent are 

 imperfect. In spite, however, of the slight elasticity of the determination at these 

 two links in the causal chain, it is clear that in normal development the symmetry 

 of the embryo is mainly determined by the point of sperm-entry. See also Tung, 

 1933- 



^ Huxley, 1927; Gilchrist, 1928, 1929; Vogt, 1928 b. 



