304 PLANT-BREEDING 



life-conditions cannot, of course, be set from these observa- 

 tions. But we do not need them for our present discussion. 

 The main fact is that all organs and qualities must go through 

 such a responsive period, and that this period coincides, 

 at least partly, with their extreme youth. Hence we may 

 derive a general rule for the correlations depending on 

 fluctuating variabinty. For it is evident that, with the 

 continuous change of the weather, only such organs are 

 really exposed to the same life-conditions, as are sensible 

 to them at the same time. As soon as the period of sensi- 

 bihty has passed, the weather can no longer have an influence, 

 but new parts are produced which will be exposed to the 

 influences prevailing at that time. In this way we may 

 conclude that one of the great factors of correlation is 

 equahty of age, because it exposes the organs, during the 

 period of their sensibiUty, to the same conditions. 



Another cause of correlation may be looked for in the 

 mutual dependency of different organs, the one affording 

 or controlhng the nourishment for the other. This case, 

 however, though of quite common occurrence, cannot always 

 be easily separated from the first, both causes ordinarily 

 combining their efforts in the same direction. 



It may be studied by the statistical method. A hundred 

 or more individuals are measured, and the result is represented 

 by a single Hne. In order to do this, the individuals are 

 arranged according to the degree of development of the 

 measured quahty. Suppose we compare the amount of sugar 

 in the sugar-beets of a field. This amount can be measured 

 for a part of the tissue, without sacrificing the beet, allowing 

 thereby the selection and the ulterior cultivation of the best 

 samples. As is generally known the sap is pressed from 

 the tissue and after clearing it, the percentage of sugar is 

 determined by the method of polarization. In the year 

 1896, I had an opportunity of studying these percentage 



