INTERRELATIONSHIP OF EG(, RECORDS 
99 
TABLE H 
Comparison of correlation between monthly egg production and annual 
egg production, and monthly egg production and the record of the remain¬ 
ing eleven months of the second year. 
Month 
Correlation 
with total 
production 
r/E r 
Correlation with 
other eleven 
months 
production 
r|E r 
Difference 
1 Vi 
centaKe 
reduc¬ 
tion 
November 
•3467 
.02S2 
12.29 
.21 13 
.0306 
6.91 
— 1354 
39 1 
December 
• 3 2 72 
1 .0285 
11.44 
.22 19 
! -0305 
7.28 
—•1053 
32.2 
January 
.4188 
.0264 
15.86 
.300S 
.0291 
10.34 
— . 11So 
2S.2 
February 
•4643 
. 0 25I 
18.50 
• 33 88 
.0284 
U -93 
—•1255 
27.0 
March 
•5549 
.0222 
25.00 
.4682 
.0250 
i 8-73 
— .0867 
15.6 
April 
•4795 
.0247 
19.41 
• 3805 
.0274 
13-89 
- .O99O 
20.6 
May 
•5324 
.O23O 
23 -'5 
.4196 
± .0264 
15-89 
—. 1128 
21.2 
June 
•5947 
: .0207 
2S.73 
.4801 
.0247 
19.44 
—. 1146 
■ 9-3 
July 
.6647 
.0179 
37-13 
■ 5362 
.0228 
23 52 
—.1285 
' 9-3 
August 
.7206 
.0154 
46.79 
• 5 8 43 
± .0211 
27.69 
—•1363 
18.9 
September 
• 7257 
± .0152 
47-74 
.6054 
.0203 
29.82 
—.1203 
16.6 
October 
• 5204 
± .0234 
1 22.24 
.410S 
.0266 
15-41 
—. 1096 
21.1 
III. Presentation and Analysis of Data 
i . The correlation between the records of the individual months 
and the annual egg records. 
The coefficients showing the relationship between the egg pro¬ 
duction of the individual months and of the year as a whole 
appear in table i for the first and in table 2 for the second laying 
year. The coefficients vary considerably among themselves in 
each year. They are represented graphically by the series of 
solid dots bordering the shaded areas in the two figures of dia¬ 
gram 1. These show that for the first year the correlations for 
November and October are somewhat lower than those for 
December and January, and August and September, but higher 
than those for the period from February to July. In the second 
year the correlations increase with fair regularity from November 
to September, but the October correlation is materially lower 
than the September constant. It does not seem desirable to dis- 
