868 
Journal of Agricultural Research 
Vol. XXX, No. 9 
Table II .—Correlation of age and body 
measurements with the milk yield of the 
cow 
Characters 
Correlation 
coefficient 
Age- 
Shoulder height . 
Hip height. 
Body length. 
Hump length_ 
Body width. 
Thurl width. 
Body girth. 
Weight. 
• 344± .030 
. 590± . 022 
,391± .029 
. 520± . 025 
,214± .033 
. 476=t .027 
. 652±°. 021 
® The correlation for the 161 actual weights for 
milk yield with body weight is 0.623. 
The graph showing the relation of milk 
yield and body length has as its first 
observation a cow of short body length 
but of relatively high milk yield. As 
this average is based on only one in¬ 
dividual it can scarcely be significant 
except in showing the amount of varia¬ 
tion which can occasionally be found 
when considering individual cases. 
The graph for girth shows similar ir¬ 
regularity. The drop from the high 
point of milk yield with a girth of 82 
inches to the lowest point in milk yield 
with a girth of 90 inches is based on 
one individual. It seems highly prob¬ 
able that there was an errdr in this 
measurement. However, as this is 
not known, the data are used as given. 
The irregularity for the first part of 
this curve likewise is due to very 
limited data. 
The prediction equations derived 
from the correlation coefficients and 
the means and standard deviations of 
the data are shown by the straight 
lines. It is evident that these straight 
lines follow the major portion of the 
data very well for all cases except those 
for girth and age. The equation for 
girth is, as noted, very materially in¬ 
fluenced by the irregularities of the first 
three observations and also the last. 
If these observations should be dis¬ 
carded it would he found that the re¬ 
sulting equation fitted the data for 66 
to 82 inches of girth almost exactly. 
In the case of age it has not seemed de¬ 
sirable to calculate the linear line, as 
it is well known that this line does not 
fit the relation between age and milk 
yield. It has therefore been omitted 
from the original chart. It might be 
said, however, that in attempting to 
obtain the relation between the body 
measurements for a constant age this 
linear line might be used, as it would 
correct for most of the variability 
caused by age. This method of cor¬ 
rection is given elsewhere. 
Judging from these graphs it seems 
entirely reasonable to use straight lines 
to represent the increase of milk yield 
for the increase in body size. The 
linear equations for these data as noted 
above have therefore been calculated. 
These equations are given as follows: 
(t) 7-day milk yield=14.2 shoulder height—412.9 
(2) 7-day milk yield=13.1 hip height—366 5. 
(3) 7-day milk yield=11.5 body length-374 1. 
(4) 7-day milk yield=19.7 rump length—'63.5. 
(5) 7-day milk yield=21.9 body width-128 3. 
(6) 7-day milk yield=225.8+ 5.9 thurl width. 
(7) 7-day milk yield=7.4 girth-201.5. 
(8) 7-day milk yield=23.8+0.289 weight. 
These equations are used for the 
straight lines of Figure 1. We may 
note certain of the significant points 
connected with them. Thus for each 
inch increase in height at shoulders 
there is an increase in 7-day milk 
yield of more than 14 pounds. For 
each inch of increased height at hips 
there results an increase of milk yield 
of 13 pounds For an increase of 
body length of 1 inch there results 
an increase of milk yield of 11.5 pounds 
on the average. An increase of 1 inch 
of rump length represents an increase 
of milk yield of 19.7 pounds and an 
increase of body width at hips of 1 
inch increases the average milk yield 
of the cow 21.9 pounds. An increase 
in width at the thurl represents a lower 
increase in milk yield, for 1 inch in¬ 
crease in thurl width is only equivalent 
to 5.9 pounds increase in milk yield. 
An increase of 1 inch in girth represents 
an increase of 7.4 pounds of milk on 
the average for such cows. An in¬ 
crease of 1 pound of weight represents 
an increase of milk yield of nearly 
0.3 pound of milk. 
These results can not, of course, be 
compared directly, as the range of va¬ 
riation for the various body measure¬ 
ments is markedly different. If what 
might be reasonably considered the 
end points of .these different measure¬ 
ments as indicated by significant data 
are taken, it is found that the body 
length, body width, girth, and weight 
are the items which caused the most 
marked increases in the 7-day milk 
yields of the cows, as the measure¬ 
ments are varied from the lowest to 
the highest in the breed. Closely fol¬ 
lowing these are rump length, height 
at shoulders and hips. The thurl 
width is the least important measure¬ 
ment. 
It is of interest to compare the 
amount of variation in the milk yields 
controlled by these different body 
measurements. It will be remembered 
that the reduction in standard "devia¬ 
tion due to the making constant of 
