396 GERMINAL ORGANIZATION INDUCTION PHENOMENA 4 



the archenteron roof, in which the relative proportion between the inductive and the 

 transformative factor is changing steadily in favour of the latter. So, the differentiation 

 tendency of the most posterior part of the presumptive neural plate changes during the 

 period of gastrulation under the influence of the invaginating archenteron roof from 

 prosencephalic through rhombencephalic to spinal cord, and finally probably into tail 

 mesoderm. More anterior parts of the presumptive neural plate are not transformed as 

 far as that; the process of transformation stops with the differentiation tendency to spinal 

 cord. Still more anterior parts, also first activated to prosencephalic differentiation tend- 

 encies, are only transformed as far as the rhombencephalic level, while the most anterior 

 part of the presumptive neural plate continues its further differentiation in the original 

 prosencephalic direction, since no transformative influences are exerted upon it" (p. 235). 



Such is the new proposal. After gathering the fuH amount of available information, 

 •we shall have to make our final conclusion on this theory (p. 473). Let us state here 

 that the activation of the ectoblast by the archenteron roof which passes beneath it 

 seems to be an established fact. In the implanted folds used to reveal this activation, 

 it expresses its influence by producing diverse prosencephalic structures, perhaps 

 due to some factor inherent in this experimental technique. I also agree that the 

 inductive action of the chordomesoblast differs from that of the prechordal 

 mesoblast, that the structures of the hind-brain have partly a transitional character, 

 and that the chordomesoblastic type becomes purer in more caudal parts. How- 

 ever, whether or not the succession of activation and transformation is real and 

 has a physiological meaning remains, I think, open to discussion. Let us look at 

 other techniques which are already helping us to solve this complicated problem^ 



[b) Explantations of unmodified territories 



A quarter of century ago, there was a hope that the in vitro cultivation of cell 

 groups would immediately reveal the true potencies of the chosen area. This 

 expectation has not been realized either in amphibians or in birds. 



In urodeles and anuran eggs, all parts of the very early gastrula have been 

 systematically and most skilfidly isolated aiid aseptically ctiltivated by Holtfreter 

 (1938b) in a saline solution (which is since that time recognized as the most 

 appropriate). Hardly any region shows typical morphogenesis. If the cells disperse 

 on the bottom of the dish, as often happens, they generally remain "indifferent". 

 If the explanted fragment preserves its cohesion and form, it generally yields more 

 than would be expected from its presumptive value. The fate of the marginal 

 zone is the most interesting. Most frequently, parts of the cells flatten and stretch 

 over the surface, covering the fragment with a kind of epiblastic layer. Inside, 

 tissues or groups of cells differentiate remarkably. The degree of paragenesis is 

 moderate for pieces explanted from the medio-ventral region of the marginal 

 zone, more accentuated for the lateral region, and still more for the dorsal one. 

 Here, the notochordal and somitic areas behave similarly. Notochord and muscu- 



' This account has been entirely devoted to amphibian embryos, on which most of the 

 recent experiments have been performed. Chick blastoderms have however been used also. 

 Abercrombie and Bellairs (1954) have exchanged the anterior node of primitive streak 

 stages with the posterior part of this streak. The results are essentially variable, but in 

 some cases at least the unity of the embryo is restored. Grafts labelled with ^^p were used 

 to show that the notochord, often V-shaped, was formed from host material. 



