159 
to be a negative. This depends evidently upon the 
fact that a partial inversion of the cane-sugar has taken 
place. If only cane-sugar and dextrose had been pre- 
sent the percentage of sugar should never have been 
stated as a negative one; both these sugars turn the 
plane of polarization in the same direction. 
Stephani (1911) and Jekelius (1912) state that 
the polarization decreases considerably during winter 
in Beta , on account of the inversion of the cane-sugar 
to a large extent; the case may be similar in carrots. 
However, a much larger part of the cane-sugar is in- 
verted in carrots than in Beta already at the harvesting. 
In their investigations of the inversion of the cane- 
sugar in Beta , Stephani and Jakelius have found that it 
takes place much more rapid in the average proofs of 
varieties with a low percentage than in sorts with a 
high percentage of sugar originally; the inversion is also 
said to be more complete in the former sorts. A negative 
correlation seems thus to excist between the percentage 
of cane-sugar and invert-sugar in Beta, and the writers 
cited suppose the existence of such a correlation; the 
facts presented are not quite convincing, however. Thus 
Jekelius has found a considerable variation, in this case 
when individual roots were studied. 
Any correlation between the percentage of cane-su- 
gar and invert-sugar in carrots does not seem to exist. 
They seem to vary quite independent of each other with- 
in the same sortas well as within different sorts. It is 
seen from table 2 that all the sorts with exception of St. 
Valery, crop B, have about the same percentage of total 
sugar, whereas the percentage of invert-sugar shows a 
very marked variation. The case seems to be the same 
within the different varieties (see table 4). The number 
of analyses, however, are too small — and the same holds 
true with regard to those made by Stephani and Jekelius 
— to make possible a correct answer to the question if 
