528 Notes 011 the Siigar Beet and its Juice [July 



seems to us, the strongest evidence for a negative relationship 

 between the weight of the beet and its sugar content in commercial 

 cultures. Novotny selected from his laboratory record books 

 analyses of samples of beets which had been divided into two sub- 

 samples, one of large and one of small beets. Altogether there are 

 2y of these pairs of samples, taken from 1892 to 1910. With only 

 two exceptions (in which the sugar percentage was identical in the 

 two cases) the sugar content of the lighter was higher than that of 

 the heavier fraction.^ 



3. Summary and discussion. There is a more or less wide- 

 spread idea that the percentage sugar-content of large beets is lower 

 than that of small roots. This belief, which has often been opposed, 

 was placed on a scientific basis for American commercial cultures 

 by an earlier study, in which we showed that there is a significantly 

 negative (and sometimes numerically very substantial) correlation 

 between the weight of the root and its total solids, its percentage 

 sugar-content and its coefficient of purity. The strongest evidence 

 in Support of our conclusions (so far as sugar content is concerned) 

 is furnished by the data of Novotny. Recently, however, wide cur- 

 rency has been given to the conclusions that there is no necessary 

 negative relationship between weight of root and composition of 

 Juice, and that when analyses are made of a series of beets derived 

 from the same mother plant no such correlation is demonstrated. 



The data upon which this Statement has been made seems to be 

 six series by Andrlik, Bartos and Urban. Of these the figures for 

 one series only are given in a form really suitable for Statistical 

 analysis. This gives a low positive correlation. The tables for the 

 other five series contain so many obvious inconsistencies that they 

 cannot be used. But three other series by Andrlik and Urban, also 

 from beets derived from an individual mother but analyzed for a 

 different purpose, all give negative correlations between weight and 

 sugar content. One of these is numerically insignificant ; the other 



ö Notwithstanding the fact that his data bear evidence without a single ex- 

 ception against the Statement that there is no negative relationship between 

 weight of root and composition of juice, Novotny attempts, by the use of a 

 formula which seems to us to have no theoretical justification, to show that the 

 difference in sugar content between large and small beets is becoming smaller 

 and that his results are consequently in accord with those of Andrlik, Bartos and 

 Urban. 



