240 Phictuatton [ch. 



round the most frequent or modal weight characteristic of 

 each individual. There will then be seen that some rouo-h 



o 



correspondence exists 'between the modes for the individual 

 plants and the weight of the individual seeds from which 

 they sprang. The heavier families on the whole come from 

 the heavier seeds, and the lighter from the lighter seeds. 

 Such correspondence is nevertheless very imperfect, and the 

 weight of the seed chosen from the original mass gives only 

 a vague indication of what the modal weight of that plant 

 will be. 



But when any one family raised from a single seed was 

 taken, and the heavier and the lighter of its self-fertilised 

 seeds chosen and separately sown, Johannsen found that 

 the modal weights were approximately the same for the 

 produce of both the heavier and the lighter seeds. Selection 

 inside the family 7'aised fro7n a single seed did not alter the 

 7nodal weight, which ive7tt back or regressed to that of the 

 individual common pareiit. Such a family thus raised from 

 the single seed constitutes what Johannsen has called a 

 pzcre line. Within the genetically pure line there are 

 fluctuations in weight, but these fluctuations are due to 

 interference which Is external, or environmental in the 

 wide sense, and selection of those extremes which are due 

 to such interference produces no effect on the result, for the 

 differences in the weights of the seeds of a pure plant do 

 not indicate differences between the germ-cells which pro- 

 duced them (164). 



It is to variation of this type that the name fluctuation 

 is given. 



Johannsen has made a similar series of experiments In 

 regard to the proportion of defective grains present in 

 certain barleys and has reached a similar conclusion. 



Whether the distinctions characteristic of the pure lines 

 segregate or not cannot yet be said. By the nature of such 

 cases the distinctions themselves are fine, and even if 

 segregation occurred the proof of its existence would involve 

 many and serious practical difficulties. 



When the complex diversity of Intermediates is thus 

 appreciated it will be understood that extreme caution is 

 needed before the statement is made that in any specific 



