1 2 HEREDITY. 



parents only, not in some vague way from more remote ances- 

 tors. We believe the facts all tend to make the view more and 

 more untenable, that there is a distinction between germplasm 

 and soma. Everything points to it, that in essentials all the 

 cells of an organism, up to the moment of a formation of game- 

 tes are identical, that they gave at least an identical set of 

 genes, namely the set present in the zygote from which they 

 all descend. It is only a matter of technique which prevents us 

 to show that every body-cell, every somatic cell is a potential 

 producer of germ-cells. It is possible to show that under certain 

 circumstances one single epidermis-cell of Begonia or of Carda- 

 mine produces a complete plant, which in its turn is capable of 

 sexual reproduction, and the only possible explanation of the 

 formation of adventitious buds is, that in these instances soma- 

 tic cells, which normally would not have germ-cells in their des- 

 cendant, can be induced to produce branches capable of flow- 

 ering and seeding. Whereas, it has not been shown to be 

 possible to trace the origin of germ-cells in animals to cells, 

 which normally would not have germ-cells in their descendants, 

 the work of Carrell and others on cultivation of somatic animal 

 tissue cells in vitro shows show these difficulties are mainly 

 a matter of technique. 



An organism inherits, whatever it does inherit, exclusively 

 from its parents as part of its zygote. Such things as Hackel's 

 recapitulation theory, which states that the ontogenetic deve- 

 lopment is a recapitulation of the phylogenetic development of 

 the species, have now only historical interest. Phylogenetically 

 new characters, qualities which the members of a group possess 

 since a short number of generations, are not shown at some late 

 stage of development, but at the exact stage in development in 

 which the peculiarity in genotype on which the new character 

 is partly dependant, is exerting its influence. If, by cross-breed- 

 ing, we produce a rumpless fowl or a waltzing rat, we do not 

 observe that in our new strain the chicks after, having devel- 

 oped normally, lose their tails, or that the rats begin to waltz 

 after they have developed normally at first, but the character 



