426 BOTANICAL GAZETTE [DECEMBER 
families which are preserved from generation to generation by 
continuous selection. Whenever a family shows deterioration it 
is discarded, but, as new families are annually added, the number 
composing the variety is not materially altered. 
One of the most costly operations practiced in sugar-beet 
breeding is the determination of the so-called best roots, either for 
starting families or as representatives of families already established. 
This is accomplished by chemically analyzing the individual roots 
of each family and grading them upon the basis of size, shape, and 
percentage of sugar. The few best roots are saved to preserve the 
family and continue the process of improvement, while the remain- 
der are thrown into a mixed lot and used for growing stock seed. 
A fairly good conception of the number of analyses made in com- 
mercial work of this kind is afforded by the records of a single 
European beet-seed company who analyze over 300,000 individual 
beets a year. The cost of making these analyses forms a large 
part of their operating expenses. Yet, aside from the theory 
that “like begets like,’’ there seems to be little evidence that 
these highly selected roots are better for breeding purposes than 
the discards. It has apparently been assumed upon theoretical 
grounds that a high percentage of sugar tends to be transmitted, 
and therefore is the most important quality of a mother root 
without regard to the possibility of its being a fluctuating charac- 
ter and nowise indicative of the average quality of the plant’s 
progeny. 
Real differences between individuals and varieties are often 
obscured by variations caused by irregularities of the soil, which 
makes it difficult to distinguish hereditary differences from tempo- 
rary differences and thus complicates the question of methods. 
The comparative efficiency of different methods has been treated 
somewhat in detail in another paper,’ but the effect of environment 
on the behavior of consecutive check and progeny rows will be 
graphically illustrated here, as it presents a number of interesting 
phenomena which have an important bearing on the results of 
this investigation. 
* A contribution to the use of checks and repeated plantings in variety tests. 
