Ill PRIMARY INDUCTION EXPERIMENTAL DATA 379 



blastoporal lip were tested by an other method. The pieces were grafted to a neurula and 

 covered with ectoblast of a late gastrula. They formed mostly notochord, and induced 

 hind brain and sometimes an ear vesicle (Okada and Hama, 1948). 



Thus, the situation which this investigation reveals is not without variations accord- 

 ing to the species considered, the iTioment where the change takes place, the factors 

 able to cause it, etc. These nuances explain, for example, the apparent discrepancy 

 between the above results and some of Gallera. The species were not the same and 

 the technique was not identical. In Gallera's cases, where the ability to induce a 

 complete head was recognized in the chordomesoblast, it probably did not exist there 

 beforehand, but was acquired, as suggested, under the experimental conditions. 



At this stage of our analysis, we may keep in mind that two distinct inductors 

 are operating to form a normal neural organ, but that, at least in Urodeles, both 

 are not effectively present before gastrulation. This ripening or emergence of the 

 acrencephalic inductor seems to be more precocious in Anurans, and this may be, 

 after all, the essential difference between their eggs and those of the Urodeles. 



Epigenesis of the acrencephalic indvictor may be quite a useful concept in 

 reconsidering the results I obtained, from 1933 to 1947, on anuran eggs, first Rana 

 temporaria and then mainly Discoglossiis pictus. These experiments were begun at a 

 time when the distinction between head and trunk organizer (Spemann, 1931) 

 had barely been realized; they were planned as a further step in the exploration 

 of the dorso-marginal zone, the recognized site of the organizer. 



The first operations performed were of limited extent. However, since they 

 were of a new type and were mostly concerned with the blastula, they afforded 

 some valuable information. 



By a temporary opening through the roof of the blastocoele, the inner material of the 

 dorso-marginal zone was detached and transported to the ventral part of the cavity. As a 

 result, the primary embryo remained normal, but secondary structures were obtained 

 and induced in its ventral side, including parts of definite acrencephalic character (Fig. 48). 

 It thus seems that this kind of implant not only contained presumptive chorda, somites 

 and pronephros, but also prechordal plate, or else was able to produce the acrencephalic 

 inductor itself. Yet, the dorsal part of the egg was capable of compensating for the loss 

 inflicted upon it. This regulative capacity was also observed after an excision of a rectangle 

 including the endoblast above the blastopore, the prechordal plate and some mesoblast. 

 This remarkable ability had been previously observed by several workers in Spemann's 

 group {cf. Schenk, 195 1), but it is worth stating that this is also true for the Discoglossits egg. 

 Another limited but interesting operation consisted of subjecting this same central region 

 to a 180 rotation (Fig. 49). The expected result would have been the formation of a 

 blastoporal lip at the upper border of the rotated fragment, and a downward invagination. 

 However, instead of this, the upper part progressively sank and migrated toward the 

 animal pole in such a position that the pharyngeal endoblast and the prechordal region led 

 the way as they would have done normally. This change may be described as a true 

 kinematic regulation, a process also observed more recently in newt gastrulae' (p. 386). 

 The space corresponding to the graft was covered by the expanding ectoblast, which was 

 submitted to induction. Acrencephalic structures did appear, although they were not 

 always normal (Fig. 50). No doubt their inductor was present in the rotated material. 

 This result agrees with the contention of Okada and Takaya that in Anurans the prechordal 

 properties exist before gastrulation. 



^ Waddington and Yao (1950) have also performed the same operation on the newt. Regu- 

 lation occurred in the majority of cases, while other embryos were affected with spina bifida. 



I.ilrrature p. 483 



