THE LAW OF JOHANNSEN. 225 



concept is correct only on the average, and not holding for 

 individual cases. 



We have tried to show, that there is a third course, that 

 quite apart from any existence of pure lines in one species 

 or the other, the work with pure lines made us discover Jo- 

 hannsen in law, which has as little to do with pure lines as 

 Newton's law with empty spaces, and which simply states 

 that the genes are qualitatively stable. And we saw, that until 

 now no facts have been brought to light, which are not in 

 accordance with this law. We have seen how and when select- 

 ion may even become effective in pure lines, but why such 

 an occurrence has not yet been noted, and what is the funda- 

 mental difference between pure lines and clones, and on the 

 other hand between metazoa and protozoa. We certainly can- 

 not compare results in pure lines of metazoa with results in 

 clones of protozoa, or state that the results obtained in one 

 group cannot be valid, because they are not supported by 

 results obtained in the other group. 



We may not compare the inheritance of characters in higher 

 plants and animals, where such "inherited" characters are pro- 

 duced anew by the children, to inheritance of characters like 

 the red colour of a dye or any character of uni-cellular organ- 

 isms, where the children have the character of the parent 

 because a moment before we observed them as separate indi- 

 viduals, they constituted this parent. 



Selection-experiments with pure clones would be very in- 

 teresting and we would suggest to anyone, who entertains a 

 doubt about the law of Johannsen, to start series of selection- 

 experiments with pure lines and pure clones. We would think 

 that only the very best material would be good enough for 

 such experiments. The pureness of "pure lines" is of double 

 origin, made up of homozygosis for the genes present, and of 

 qualitative purity of the genes. Therefore, absolute proof of 

 homozygosis of the material is necessary, to decide for or 

 against the law of Johannsen. Loss-mutations should be watch- 

 ed lor, but homozygosis is more important than anything 



15 



