458 Bulletin 279. 



manner to changes in climatic conditions. This is undoubtedly an 

 important factor in bringing about place variation and is probably one 

 of the chief causes of the oscillations of the empirical modes. 



Another factor which must be taken into consideration is that of 

 the unequal action of environment on the population as the result of 

 greater susceptibility of certain individuals to environmental influences. 



One of the most noticeable features in the distribution curves for 

 height and weight for the three years is the change in skewness. An 

 inspection of Fig. 142 shows that in 1905 the curve for height was very 

 nearly normal, while in 1906 there was a marked negative skewness 

 which was still more pronounced the following year. In the case of 

 weight we find conditions reversed. In 1905 there was a strong positive 

 skewness, while in each of the succeeding years there was a change 

 toward the normal type. 



There have been several interpretations of the significance of skew- 

 ness. One of the first of these and one which has undoubtedly been 

 most widely accepted until recently, was that it indicated an evolution 

 of the species either by the cutting off of certain individuals from one 

 side of the mode by natural selection or that a shifting of part of the 

 population had resulted from the establishment of a new race within 

 the range of the old (Davenport 1901). 



Later studies have shown that these views do not explain the majority 

 of cases of skewness. Kapteyn (1903) points out the fallacy of suppos- 

 ing that the causes which produce variation act equally on all indi- 

 viduals, and that the action of these causes is not such as to induce 

 half the population to deviate to a certain degree in one direction and 

 the other half to a corresponding degree in the opposite direction. He 

 also maintains that causes dependent on size result in skewness. Tower 

 (1906) believes that in the material with which he worked, the skew- 

 ness observed was due to place variation and that this was the result 

 of " fluctuations from time to time of the individual factors in its environ- 

 mental complex." 



In timothy the increase in skewness in the case of height is evidently 

 due largely to environmental conditions. In order to study in detail 

 the changes which may have taken place in the individuals which were 

 in part responsible for the skewness, the record of the extreme variants 

 on the negative side of the curve for 1907 was looked up and the data 

 given in the following table: 



