24 HEREDITY. 



or labile conditions, and substitute the hypothesis that genes 

 are so many relatively simple chemical substances, we cannot 

 conceive of spontaneous variation, variation that is not given 

 by the potential variabihty, genotypic diversity of the group 

 which presents it, as a function of the genes, as a process taking 

 place from within, and we must have recourse to some explan- 

 ation of spontaneous variation which sees the cause in the rela- 

 tion of the organism to its surroundings. De Vries assumes 

 certain labile states of his vital pangenes, and ultimately he 

 looks to the ambient circumstances to bring about these states, 

 in which the pangenes occur. It is not difficult to see, how even 

 if genes and the action of genes are the same in all organisms, 

 spontaneous geno-variation may in its causation, be entirely 

 different in unicellular from what it is in multi-cellular orga- 

 nisms. It is certainly a significant fact, that no authentic cases 

 of a positive mutation in the higher plants or animals, are on 

 record. Therefore, we may very well leave a discussion about 

 the ways in which plants and animals can acquire new genes 

 outside crossing, till the moment on which such an occurrence 

 will be observed. On the other hand, in uni-cellular organisms, 

 the taking up of a gene, new for the group, may not be an 

 impossible rare process. It is for this reason, that we would 

 like to warn against accepting instances of the effect of selec- 

 tion in apparently pure clones, of uni-cellular organisms as 

 invaUdating Johannsen's law. 



If we accept the h37pothesis that genes are relatively Simple 

 chemical things, and we search for an explanation of such cases 

 of mutation as have been observed, that is to say of cases of the 

 apparent loss of a gene, we must remember, that a distribution 

 of a quantity of the gene in question over all the cells of the 

 complete organism, is possible only, if the materials the "ingre- 

 dients" of the substance, are available. To take an example. 

 Let us suppose oxide-of-iron, which is a substance with auto- 

 katalitical properties, to be a gene, and a developmental factor 

 in a family of plants. Only so long can the distribution of the 

 substance be carried on, as iron and oxygen and oxide-of-iron 



