178 PHYSIOLOGICAL GENETICS 



.irtcr; and the Crataegus epidermis shows a clear influence of the 

 underlying Mespilus tissue. The structure is actually more 

 intermediate than in real hybrids. If from such a chimera pure 

 Mespilus shoots arc formed, containing only Mespilus cells and 

 chromosomes, an influence of the Crataegus genotype is visible 

 and persists. No doubt in this case morphogenetic stuffs, 

 controlled by the genes of one species, pass from cell to cell and 

 act upon the distant cells. 



None of these cases is yet quite clear, and at present it may 

 suffice to register the facts. 



There can be no doubt that some gene-controlled reactions take 

 place within the same cells in which the respective genes are 

 situated; furthermore, that this action may take effect even if it 

 is started rather late in development. It seems, moreover, that 

 this type of genie action is confined to such reactions as occur 

 either at the very beginning or at the end of development and 

 that in the latter case they lead to no further pattern formation. 

 In some cases, however, the products of these reactions may 

 diffuse into the surrounding cells. 



In his w r ork on sex, from which Goldschmidt originally had 

 derived his view r on genie action, he thought that he was dealing 

 with a typical case of purely intracellular action of the genes in 

 question. Embryological and morphological facts which were 

 found later convinced him, however, in agreement with the 

 experimental work of Seidel (see Seidel, 1936), that in insects, 

 also, a part of the determinative reactions are not of the intra- 

 cellular, strictly self-determining type. We shall mention these 

 facts later. The foregoing discussion shows that such types of 

 action are to be expected wherever pattern formation is 

 involved. 



It has to be kept in mind also that in experimental embryology, 

 the limits between self-differentiation and induction have turned 

 out to be rather vague', if not actually nonexisting, and that the 

 actual difference is a difference in time of irreversible determina- 

 tion. But there are also some cases existent in which the deter- 

 minative reaction, controlled by genes, actually takes place 

 within highly differentiated cells without induction from cells in 

 the neighborhood. Such cases may be analyzed in the mosaics 

 that the experimenter produces by transplantation as well as 

 in the genetic mosaics reported above. 



