FIELDS AND GRADIENTS IN NORMAL ONTOGENY 317 



the centre of the gradient, and constituting a subordinate region, 

 must in the vegetative half have turned into a dominant region and 

 come to control the new complete gradient-field of the fragment. 

 The curious and apparently anomalous production of notochord 

 and mesoderm by various isolated regions of the blastula of the 

 newt, although these regions may possess the most diverse pro- 

 spective fates (p. 139, footnote), may perhaps be explained on 

 these fines when it is remembered that isolation of a piece of tissue 

 removes it from the control of the dominant region to which it has 

 been subjected. As noted on p. 285, experiments on Planarians and 

 Hydroids have shown that the tendency in such cases is for a small 

 isolated piece to develop by self- differentiation into an isolated 

 dominant region. The dominant and only self-differentiating region 

 in the late blastula of the newt is the organiser, and the tissue into 

 which it differentiates is notochord and mesoderm : other regions 

 develop in subordination to it. On this assumption, therefore, a 

 piece from any other region, when isolated, should, if environmental 

 circumstances permit, come to be the site of a new dominant 

 region, and differentiate accordingly.^ However, the occasional 

 differentiation of such pieces into tissues which represent neither 

 the presumptive fate of the piece nor that of the dominant region 

 (organiser) presents a difficulty. We should however recall that 

 whereas in the Invertebrates only a simple field is involved, in 

 Amphibia there are two interacting gradient-fields (pp. 310, 318). 



(iv) The modifying influence exerted by the dominant 

 region on other parts 



This is obvious in the example just given of the formation of 

 miniature wholes from animal and vegetative portions of Coelenter- 

 ate eggs. Regions originally containing but half the length of the 

 main gradient become reorganised to contain whole gradients. 



In most of the well-analysed types of ontogeny, however, con- 

 ditions are more complex than in the regeneration of Hydroids or 

 Planarians, for the main organising activity proceeds from the high 



^ In connexion with the environmental circumstances, it is a curious fact that 

 pieces of presumptive neural tube tissue (which has been the tissue most fre- 

 quently used in these experiments) show a much greater tendency to differentiate 

 into notochord when interplanted into the coelomic cavity of an older larva than 

 when explanted in an inorganic medium (Holtfreter, 1931 a). See also Huxley, 1930. 



