No. 535] 



NOTES AND LITERATURE 



419 



hand, of germ-plasm (i. e., germinal determinant factors of 

 whatever sort), which presumably was identical for each plant 

 during the three years, and environmental factors, on the other 

 hand, in determining observed degrees and kinds of variation in 

 the adult organism. The results taken as a whole show that what 

 might be called the general variation fades of a population of 

 Phleum must depend to a very high degree upon "nurture" 

 rather than "nature." The degree of variation, the degree of 

 skewness of the variation curves, the closeness of correlation 

 between different characters of the plant — all these are changed 

 by general environmental conditions to a marked extent. Thus 

 to take an example: the coefficient of correlation between weight 

 and height of plant is given as .274 ± .011 in 1905 and as .718 

 in 1907. This is a relative change of nearly 200 per cent. In 

 another case a significant positive correlation one year becomes 

 significantly negative two years later. In general the heights (or 

 weights) of timothy plants in any one year are correlated with 

 the heights (or weights) of the same identical plants in another 

 year only to about the degree indicated by a coefficient of around 

 .5, which is but 50 per cent, of perfect correlation. 1 In other 

 words, it appears on the basis of this result that in determining 

 what a given timothy plant shall be like next year in respect to 

 such characters as height and weight the innate constitutional, 

 hereditary factors within the plant are on the whole of neither 

 greater nor less importance than external environmental cir- 

 eumstances. In this case, and in respect to the characters dealt 

 with, "nature" and "nurture"* are about evenly balanced, with 

 what advantage there is on the side of "nurture." The author 

 ciiipha.siz.-s the practical significance of a result of this kind to 

 the man who is carrying on selective breeding, and who obviously 

 must make his selections at the outstart on the basis of the 

 visible somatic characters as they are developed at the particular 

 place and time at which he is doing his selecting. The paper is 

 unfortunately marred by arithmetic errors. 



It is a well-known fact that European workers (other than 

 English), generally speakimr. have very little acquaintance with 



a paper of Grabner on the problem of correlated variation in 

 barley. The investigator doir.-d t-. h-arn what relation existed 



'Cf. Clark's Table VIII. 



