4 BULLETIN 1144, TJ. S. DEPARTMENT OF AGRICULTURE. 
Obviously all these different materials must be reduced to fairly 
definite relations to some standard if feeding animals is to be any- 
thing short of pure guesswork. Profitable feeding is a fine art, but 
an art developed on a basis of countless experiments to determine, by 
chemical analyses and by physical measurements, the relative values 
of different feeds for production and the feed requirements of cows 
of different capacities. The standards thus established serve as a 
means of reducing the different feeding practices to a common basis 
for comparison, and as a rough guide to the prices which may be 
offered or demanded for feeds of various kinds. 
Armsby's feeding standards, as given in Henry and Morrison's 
" Feeds and Feeding," 1917 edition, have been used in the calculations 
of feed values and feed requirements for this study, the principal 
unit of measurement being the therm of net energy. Digestible pro- 
tein must not be overlooked, however, in compounding rations or in 
comparing the values of concentrates, especially the cereal and oil- 
mill by-products which are used to balance rations. 
The quantities of grain, hay, silage, and fodder fed to herds in the 
five groups are shown in Table 2. The term " grain " includes, be- 
sides the home-grown corn, oats, and barley, the various purchased 
mill feeds, brewers' grains reduced to a dry basis, and beans. The 
hay fed was mostly clover, with some alfalfa. Silage includes, in 
addition to corn silage, pea silage, with small amounts of soiling 
crops on a dry-matter basis. "Fodder" was mostly cornstalks, in- 
cluding little or no grain. Besides the feeds thus supplied, the cows 
had pasture in varying amounts and of varying qualities. The term 
" pasture " covers all the feed the cows gather for themselves during 
the summer, from wild or rough land unsuitable for crop production, 
from meadows or fields definitely set aside for the purpose, and from 
the aftermath of the other fields. Thus, pasture is an exceedingly 
variable quantity, both in the amount of feed provided and in the 
number of cows it will support and the time it will provide feed. It 
is unfortunate that pasture can not be more clearly denned and di- 
rectly measured, as it plays a very important part in the management 
of dairy herds. 
In order to approximate the amount of feed supplied by pasture 
and to explain the relations between feed consumption and milk 
production, the feeding standard and analysis of feeds mentioned 
above were used in calculating the feed requirements of the cows 
and for comparing the amounts of feed supplied in the several areas. 
The pasture season may begin in April and may continue into De- 
cember. For present purposes the year was divided into two periods 
of six months — the winter season November 1 to May 1, and the " sum- 
mer," or pasture season, from May 1 to November 1. Many farmers 
do not turn the cows out to pasture until the latter part of May and 
are obliged to resume feeding by the middle of July. The practice 
varies, of course, with the amount of feed supplied by pasture and 
the number of animals to be fed. The price of feeds and of milk 
affect the use of pasture: in 1920. with feed high in price and falling 
milk prices, many farmers did, not resume feeding as early as they 
would under normal conditions. 
The estimated annual feed requirements of the cows included 
in the studv, due allowance being made for differences in size of the 
