Further Examination of the Cell-Theory 207 



in a general way the course pursued by Roux relative to the 

 observations by other biologists and by himself which are 

 clearly hostile to the mosaic theory. To meet the fact that 

 under some conditions, even in the frog, not a half-embryo 

 but a smaller whole embryo develops from the isolated blas- 

 tomeres, he advanced the notion of "postgeneration." By 

 this he means the '''supplementary restoration, completely 

 or incompletely, of a half- or quarter-embryo, or other 

 'Partial-product' formed in consequence of the destruction 

 of a part of the cgg.^^ ^^ 



By coupling this idea with an earlier speculation of his 

 about the nature of the cell nucleus, he tried to make the 

 nucleus, instead of the cell, the really responsible element 

 in the mosaic, at least in those cases in which the whole cell 

 clearly does not conform to the mosaic conception. 



Provision for retreat to the nucleus upon occasion, is 

 made in his definition of mosaic theory already quoted.^ It 

 will be recalled that the theory rests on the "assumption 

 of different qualities in the individual cells (cell body or 

 cell nucleus, etc.) capable of self differentiation." 



Alongside this modification of the original mosaic theory 

 made by transferring the role of building stones from the 

 cells to the nuclei, it is instructive to place what is in effect 

 another modification of an opposite nature invented by G. 

 Born, in connection with his extensive experiments on graft- 

 ing together the larvae of various amphibians. Born found 

 that larvae, not only of different species, but even of diifer- 

 ent genera and families, could be made to unite to some 

 extent, the union between the more closely related species 

 being in general the easiest to accomplish and the most per- 

 manent, but that in any case each component of the grafted 

 specimen maintained its specific attributes uninfluenced by 

 the individual with which it was united. In other words, 

 animal grafts, so far as these experiments went, follow the 

 well known rules of plant grafts. The maintenance of iden- 



