1916] PRITCHARD—SUGAR-BEET BREEDING 461 
or less were to lie in one group and all the remaining roots in the 
other, the chances of finding the mutant by taking an equal num- 
ber of roots from the respective groups would be as 1:6. 
The ratios 1:4.5 and 1:6 are small and hardly warrant the 
expenditure of large sums of money for chemically selecting 
y Swe 
221g 
23 
ta. i 5 » OS 2S“ 3 sa 8S 
SAELCSGESERC Resse ss sgz 
enters of percentage classes of $¢o of sper cent range 
Fic. 51.—Most probable location of mutants 
individual roots of high sugar content to increase the probability of 
finding mutants. Tests of 6-10 times the usual number of roots 
taken at random would be more economical and fully as effective. 
As continuous selection has resulted in failure and only the 
mutation theory remains as a promising means of sugar-beet 
improvement, some very decided changes should be made in 
methods. First of all, more efficient methods of making tests 
should be devised in order to reduce the effects of soil irregularities 
