Ill PRIMARY INDUCTION EXPERIMENTAL DATA 37I 



and grafted to the ventral region of a host. Gallera (1947) observed that the struc- 

 tures obtained from prechordal brain fragments of Pleurodeles IV. and of axolotls 

 were generally less complex than they would have been in situ. The degree of re- 

 gression was less accentuated for large pieces than for small ones; when small ad- 

 jacent fragments were compared, the median and anterior ones preserved their 

 individuality better than lateral or posterior fragments. Presumptive ocular ma- 

 terial covdd form a piece of brain with paraphysis or epiphysis. Regularly, a large 

 part of the graft transformed into neural crest. However, the transformation of a 

 graft into a spinal cord was not observed. Working on the spinal cord level with 

 young neurulae of Axolotl and Xenopus, Damas (1947) also described the partial 

 transformation of the graft into neural crest and its derivates'. The residue 

 remained of spinal type, but its differentiation was distinctly hindered. 



In a later investigation, Gallera (1949) reexamined in newt embryos the 

 inducing ability of the prechordal chordomesoblast. He took the grafts from 

 three stages: young neurulae, gastrulae with circular blastopore, and young 

 gastrulae. Care was taken to insert the fragments in the most ventral part of the 

 blastocoele with their superficial side adhering to the presumptive ectoblast of 

 the host. No vital staining was necessary to recognize the fate of the grafts and 

 thus these data are all the more reliable. The results of this extensive experiment 

 (157 grafts) were far from schematic but afford valuable information. 



The prechordal territory of the neurula (Fig. 42, a) was used either entire, halved, or 

 divided into one median and two lateral fragments. The latter had a strong tendency to 

 disperse into the mesenchyme (35 out of 40 cases), and in the few cases where an induction 

 was exerted, it always produced balancers, but only once induced a brain (anterior) ; 

 a modelling of the endoblast into a pharynx (twice including teeth) was observed. Thus, 

 the graft behaved like cephalic mesoblast and pharyngeal endoblast, and not properly like 

 prechordal material. With the whole fragment or its halves, individuality was better 

 preserved (44 out of 54 operations) and manifested itself in producing muscles and indefinite 

 cartilages and by inducing, in order of decreasing frequency, pharynx (39 cases), stomo- 

 doeum (36 cases, eventually with teeth in 27 cases), balancers (28 cases), anterior brain 

 (rather poor, 26 cases), nasal pits (simple or double, 19 cases), and optic vesicles (unique 

 or double, 18 cases). These sense organs were always accompanied by brain. The size of 

 the neural structures was compared to the normal dimensions of these organs and was 

 found to be more reduced for the brain parts (maxiinum io°o of the normal) than for the 

 eyes or nasal pits (nearly half of the normal). Thus, the neural inductions were definitely 

 of an acromeritic character, but the optic vesicles were exaggerated at the expense of the 

 brain, (Fig. 42, b, c) and the induction of the olfactory pits from the epiblast stronger than 

 expected. 



When the preneurula stage was used (small yolk plug. Fig. 43, a), an anterior and a 

 posterior fragment were dissected and inserted in the blastocoele. All the anterior ones 

 (20) were assimilated and could be considered as purely endoblastic. Out of 18 posterior 

 fragments, 3 only (i entire, i median part, i lateral half) caused inductions. In all of these 

 cases the inductions were strictly prosencephalic: brain, eyes, nose, balancers. The lateral 

 part also induced a small otocyst, probably because the dissection had encroached on 

 parachordal mesoblast (Fig. 43, b, c). The frequency of assimilations is probably due to 

 the loose structure of this still migrating material. This difficulty was considerably reduced 



^ The experiment of Damas has been repeated, with the same results, by Denis (1957) 

 and by Gallera (ig58b). It remains however possible that the implantation of the graft 

 on the abdomen may cause in it some stretching which could partially inhibit neuralisation 

 {cf. Denis, 1958). 



Litiraiure p. 483 



