No. 523] ARE FLUCTUATIONS INHERITED f 423 



comes from the shorter culms and suggests that in select- 

 ing large seed for planting we are selecting from small 

 plants. This may be true. The writer finds that the 

 average weight of seed in peas decreases with the height 

 of plant, which corresponds to the results obtained by 

 Waldron. On the other hand, when these are planted 

 the larger seed, although coining from smaller (shorter) 

 plants, produced larger plants than smaller seed which 

 came from the large plants. 



The foregoing results then indicate that there is not 

 enough evidence in favor of the inheritance of fluctua- 

 tions caused by environment to form a practical work- 

 ing basis. That is, by selecting out the individual plants 

 which give an exceptionally high yield we would not 

 obtain any higher yielding individuals than from a selec- 

 tion taken at random. 



We find then that we have given a definite answer to 

 the question which is the title of this paper. 



While such statements do not accord with those who 

 place their faith in cumulative selection, yet it is only 

 a statement of fact as shown by an analysis of such 

 data as are here brought together. 



These results accord very closely with those stated by 

 Pearl 6 and we can agree with a statement made by him 

 which is as follows: 



Altogether much evidence is accumulating from widely different 



The rapidly accumulating facts in this respect bring 

 us to face a different view of the value of selection. The 

 testing of individuals to learn their power to reproduce 

 their characters must be done just the same but a differ- 

 ent interpretation must be given the results obtained. 



Unless further studies produce different results, we 



