866 
Journal of Agricultural Research 
Vol. XXX, No. 9 
ment will invariably be erroneous. And if the neck 
is turned to the right or left, the measurement of 
length of body will also be erroneous. Then fol¬ 
lows the ascertaining of the weight, which must 
invariably be obtained by reliable scales; the date 
of service is then ascertained from the owner and 
the condition described. The date of the examina¬ 
tion should also be recorded in some convenient 
place in the memorandum book. 
The necessary analyses of these data 
show the averages and coefficients of 
variation seen in Table I. 
The averages for age, butterfat per¬ 
centage, and milk yield show that this 
group of cows is younger, has a higher 
butter-fat test, and is less productive 
of milk than are the cows of the whole 
Advanced Registry. 
The series of means of the different 
parts of the body gives an idea of what 
might be called the “ average’' Hol- 
stein-Friesian cow. This average cow 
The variation of the different parts 
of the cow is of considerable signifi¬ 
cance. The characters used in score- 
card judging are essentially divisible 
into two classes, physiological and 
morphological. The writer has studied 
the records of over 1,600 score cards 
taken by more than 200 judges on 
Jersey Registry of Merit cattle (1, 2). 
The judges who made these scores 
have undoubtedly taken cognizance 
of the increases in size of the parts 
scored as age advances, and compen¬ 
sated for such increase in recording 
their judgment on the cows. The 
amount of variation would conse¬ 
quently be expected to be less than it 
would have been had age been left 
out of consideration. The data on 
the Holstein-Friesian cattle are actual 
Table I.— Means , standard deviations , and coefficients of variation for the physical 
measurements of Holstein-Friesian cows 
Characters 
Mean 
Standard 
deviation 
Coefficient 
of variation 
i 
Age-- 
Milk yield_ 
Shoulder height_ 
Hip height_ 
Body length_ 
Rump length_ 
Body width_ 
Thurl width_ 
Body girth_ 
Weight_ 
i 
3. 90±0. 08 
... 338 ±2 
52.8 =t . 1 
...! 53.8±.l 
62. 0 ± . 1 
20.4 zh.l 
21.3 ±.l 
19.0 ± . 1 
73. 2 ±.2 
... 1088 ±6 
i 
1 
2. 22±0. 05 
72. 6 ±1.8 
1. 94db . 05 
1.91± .05 
3. 73± . 09 
1.44± . 04 
1. 72± . 04 
2 . 63± . 06 
4. 59± . 11 
164. 0 ±4. 2 
56. 9±1. 8 
21. 5± . 6 
3. 7± . 1 
3. 6± .1 
6 . 0± .2 
7. 0± .2 
8 . 0± .2 
13. 8± .3 
6 . 4± . 2 
15.1± . 4 
may or may not be a desirable speci¬ 
men. It would be interesting to see a 
reconstruction based on these measure¬ 
ments. It might be found that the 
average is disproportionate for some of 
the parts, as was the case for the typical 
American soldier. 
The weights of the cows are either 
estimated or actual. There are 161 
records of actual weight and 178 of 
estimated weight. The actual weights 
have a mean of 1,078, a standard de¬ 
viation of 162, and a coefficient of 
variation of 15.0. The distributions 
are consequently closely similar to 
each other. 
The measurements of this average 
cow are, on the whole, quite a little 
less than those of the present Holstein- 
Friesian ideal. Thus in shoulder 
height the ideal is nearly 2 inches taller, 
in body length nearly 10 inches longer, 
in rump length 2^ inches longer, 
in body girth about 15 inches greater, 
in hip width 2 inches greater, in thurl 
width 2 J /2 inches larger, and in weight 
about 200 pounds greater. Only in 
hip height do the average and ideal 
correspond. 
measurements. In general, the com¬ 
parison of these actual measurements 
with the scores shows that the scores 
of the judges on conformation are ac¬ 
tually more variable than are the 
parts of the cows from which their 
scores are taken. In other words, the 
judgment by eye is, on the whole, less 
reliable than estimates based on the 
yardstick. 
The comparable results found on 
other forms in conjunction with those 
found for these Holstein-Friesian cattle 
are worthy of record. The weight of 
these Holstein-Friesian cows has prac¬ 
tically the same variation as the weight 
of men or of the domestic fowl. 
Shoulder or hip height in cattle is 
slightly less variable than length of 
the forearm or femur in man. Rump 
length, body girth, and body length 
are slightly more variable than stature 
in man. 
Table II reveals several facts of no 
little importance in the selection of 
dairy cattle. Every one of the varia¬ 
bles has a rather large positive relation 
to the quantity of milk secreted by the 
mammary gland. As shown elsewhere 
