452 BOTANICAL GAZETTE [DECEMBER 
equivalent to only about 0.03 per cent sugar, while the average 
difference obtained was 0.5. Hence the commercial seed not only 
failed to show deterioration, but actually appeared to improve. 
Similar results were observed between the variety and its more 
. 
g a Row numbers 
«<§ #8, 
. to Ss & © 6 6 S&S 6 °& & 6 6 8 8 5 858 6 
 ~e™>S”?® CNM YE H DW KR DH GF YT nm = % 
oe 2s a ie i ie i 
BE, Sn: 
sooo | 
4000 15 Nog , 
% 
3000 /4 Alt We. os / ; 
ors a tee 
2000 13 UO oe a a el ee ed a 
ee e = bee of 
{000 /2 
Fic. 37.—Effect of soil irregularities on relationship between percentage of sugar 
and yield of sugar in consecutive check rows, Madison, 1914; solid line indicates 
percentage of sugar; broken line, yield of sugar. 
highly selected families in 1912 which led to the planning of this 
experiment. 
RELATIONSHIPS BETWEEN PERCENTAGE OF SUGAR, YIELD OF 
SUGAR, AND AVERAGE ROOT WEIGHT OF SUGAR-BEET ROWS.—AS 
as 
aszs, Row numbers 
‘S553 
SSSSESSSSSSESSESSSSERSSSLSERSSES 
3000 | 9 
2000 / ths NT aa” ener eal : 
1000 
Fic. 38.—Effect of soil irregularities on relationship between percentage of sugar 
and yield of sugar in consecutive check rows, Madison, 1913; solid line indicates 
percentage of sugar; broken line, yield of sugar (corrected yield). 
beets are grown under approximately uniform spacing, the same 
correlations which obtain between weight, percentage of sugar, 
and yield of sugar in individual roots would be expected to occur 
in sugar-beet rows. An ideal stand is never obtained under field 
