Sept. 1, 1925 
Net-Energy Values of Feeding Stuffs 
471 
In Table I, Part C, the metabolizable energy per pound of dry 
matter of the hay eaten is 960 Calories in period 4, and 910 Calories 
in period 3; the average remaining the same, as in Table I, Parts 
A and B, namely, 935 Calories. The apparent net-energy value 
of the hay in this case is 603 Calories per pound of dry matter, as 
compared with 502 Calories in Table I, Part A, and 401 Calories in 
Table I, Part B, although the heat increment and the average meta¬ 
bolizable-energy values are the same in all three cases. The meta¬ 
bolizable energy per pound of dry matter of the added hay is in this 
case 1,036 Calories, which is considerably higher than the value of the 
hay eaten in either of the two periods. 
The availability of the metabolizable energy of the hay, computed 
from the data which have just been given in Table I, Parts A, B, and 
C, would be, respectively, 502-^935 = 53.7 per cent; 401-^-834 = 48.1 
per cent, and 603-^1,036 = 58.2 per cent. 
It is obvious from the foregoing that the net-energy value computed 
by this method is affected not only by the difference in metabolizable- 
energy value of the feed between the two periods compared, but also 
by the direction of this difference; that is, by whether the determined 
metabolizable-energy value is greater in tne heavier or the lighter 
of the two rations. Since the net-energy value is computed from the 
difference in gain between the periods, a relatively small error in the 
determination of the metabolizable-energy value, which affects the 
gain directly, may cause a large percentage error in the net-energy 
value, especially if the difference in gain between the two periods is 
small. Obviously, errors in heat production would tend to reduce 
the effect of errors in metabolizable energy, if these errors are in the 
same direction. If, however, they are m opposite directions, their 
effect on the gains and on the net-energy value computed by difference 
would be magnified. It is possible, therefore, to obtain absurd 
results by this method of computation. 
This method was originally used by Armsby and Fries (7, 2, 8 , 4) 
in the computation of the availability of the metabolizable energy, 
which means the net-energy value expressed as percentage of me¬ 
tabolizable energy, and may account to an appreciable extent for 
the discordant results contained therein. 4 
In their later publications (5, 6 , 7, 9), Armsby and Fries expressed 
the net-energy values as per kilogram of dry matter of the feed, and 
have also recomputed the results of the early experiments on the 
same basis, using a simplified method of computation. This method 
consists of averaging the metabolizable-energy values from the several 
periods, and subtracting from this average the average heat incre¬ 
ment of the feed, usually determined by a comparison of the greatest 
and least amounts of feed eaten. The net-energy value thus obtained 
is not affected by the factors previously discussed, as in computations 
by the first method. It is, however, an average value derived in some 
instances from quite discordant results, and as such does not show the 
variations in the individual periods. 
The new modification described later makes possible the computa¬ 
tion of a net-energy value of the feed for each period separately, and 
besides it has other advantages which will be illustrated in the 
following examples. 
4 These publications cover experiments 174, 179, 186, 190, 200, and 207, the last three being included in 
Bulletin 128 of the Bureau of Animal Industry, United States Department of Agriculture. 
