408 GERMINAL ORGANIZATION INDUCTION PHENOMENA 4 



mesenchyme, which would explain the constriction zones between the five basic 

 brain parts. This hypothesis does not need further development. Let us only add 

 that indeed a brain is not as specifically distinct from a spinal cord as we have 

 the tendency to assume ; the prosencephalon is only remarkable in the size it can 

 acquire and in the refinement of its differentiations. The characteristics of brain 

 morphogenesis can be reduced to : an initially larger amount of neuro-epithelium; 

 an exceptional capacity of growth in certain territories; a gradual thinning of 

 definite areas of the wall, and a correlated expansion of the ventricles. All these 

 features could be explained by an interaction between myogenic and non- 

 myogenic mesenchyme, with a possible collaboration of the ectomesenchyme. 



In view of this hypothesis, we feel entitled to raise the question : is any part of the 

 neurectoblast endowed with a capacity of self-organization (Holtfreter, 1951, 

 p. 140; Nieuwkoop et al., 1952, p. 99; 1955, p- 272) or is the evolving of structures 

 from a neural anlage always dependent on a combination of inducing factors? 

 Could the hypothesis outlined above be applied to all cases in which an organized 

 prosencephalon appears? A continuation of this analysis will show that certainly 

 the true mesenchyme cannot be considered as the agent always present in such 

 cases. Thus, if we feel reluctant to admit self-organization, some other approach 

 has to be found to solve the riddle (see p. 444). 



Before leaving the consideration of purely microsurgical experiments, we should 

 mention the remarkable results first obtained by Niu and Twitty (1953) and 

 further studied by Niu (1956) under conditions which, although concerning in 

 vitro cultures, seem the most direct approach for the analysis of the agents respon- 

 sible for normal induction. In this new technique, the reactor is still the young 

 ectoblast, but it is divided into pieces of about 20 or 30 cells; the inductor is not 

 strictly an adjacent piece of tissue, but a modified Holtfreter solution especially 

 "conditioned" by having contained, for several days, an explant of some actively 

 inducing material. The advantage of using tiny explants of ectoblast is that they 

 possess only a limited amount oi coat, just enough to cover their previously outer 

 surface; when they round up, more than half of the sphere surface remains naked, 

 thus permitting any agent in the medium to act directly upon the exposed cells. 

 It is obvious that in case of a positive result obtained with a fluid inductor, the 

 way would be opened for a biochemical analysis. Naturally, with such small 

 reacting pieces, no proper morphochoresis can be achieved, but only cell differ- 

 entiation. With proper controls available, if the ectoblast differentiates into 

 neuroblasts or pigmentocytes or myoblasts, an inducing effect of the medium 

 can be demonstrated. 



This in fact did happen. An explant of the dorso-marginal zone of a newt early gastrula 

 was cultivated for 7-10 days in a drop of the medium, thereby conditioning it; the tiny 

 reactor was then introduced at a fair distance from the inductor, to be reared there some 

 10-20 days more and a positive induction effect was observed. The posterior (myogenic) 

 part of the neural plate (Fig. 70) was also used as an active explant for such trials. The 

 presence of the inductor in the culture is not necessary. The fluid in which it has been 

 cultivated may act in the same way, and homoeogenic induction in vitro may be demon- 

 istrated. Substances having diffused from the posterior neural plate cause the differentiation 

 of myoblasts (up to 73% of the cells), or of neuroblasts and pigment cells if the duration 

 of previous cultivation has been longer. In all these results, comparison with the controls 



