THE LAW OF JOHANNSEN. 207 



if the circumstances favour such a predominance. Such quan- 

 titative changes in the proportion of the substances which 

 make up protoplasm, are transmitted from cell to cell. And it 

 becomes evident that, in those cases where the whole organ- 

 ism is but one cell, this transmittance of definite proportions 

 between the substances which make up protoplasm is in a great 

 measure "inheritance". 



The proportion in which the constituents occur, is greatly a 

 matter of circumstances, and changes in this proportion are 

 mostly reversible, though not necessarily so. 



This potential reveisibility of changes of proportion be- 

 tween cell-constituents is in a very sharp contrast to the irre- 

 versability of another process, namely the loss and the (at 

 least hypothetically thinkable) acquisition of a gene. 



It appears at first sight as if the processes, which change the 

 equilibrium in the proportion between the autokatalytical 

 constituents of protoplasm, could easily bring about the com- 

 plete exclusion of one or several of these constituents, genes. 



Why is loss-mutation such an exceedingly rare phenomenon ? 

 Thousands of genes must be present in the higher organisms, 

 to judge from the number which we can study in some, whereas 

 we know that such a study becomes possible only through 

 comparison of individuals with and without, and is therefore 

 limited to a great extent to relatively vitally unimportant 

 ones. But loss of a gene as distinguished from absence from a 

 new combination, caused by redistribution, which is commonly 

 confused with loss-mutation, is a phenomenon so rare as to 

 be certainly negligable as a factor in evolution. 



As we will see later, the facts point to it that within the nu- 

 cleus the original set of genes, such as it is inherited by the 

 original zygote, is conserved intact. The great diversity of 

 function and of gross chemical constitution of the cell-com- 

 plexes within one organism, the facts of functional adaptation 

 admit of a simple explanation on the hypothesis that the genes 

 have autokatalytical properties. It is evident however, that 

 the relative quantitative preponderance of genes which we 



