402 
Journal of Agricultural Research 
Vol. XXX, No. 5 
Table VII.— Effect of possible error in the dry matter of feces-and-urine mixture 
on the energy, nitrogen, carbon, and computed heat production 
Experiment 
and period 
Possible 
error 
in daily 
dry 
matter 
(from 
Table 
VI) 
Energy 
per 
gram 
dry 
matter 
Carbon 
in dry 
matter 
Nitro¬ 
gen in 
dry 
matter 
Possible 
error 
in 
energy 
Possible 
error 
in 
carbon 
Possible 
error 
in ni¬ 
trogen 
Energy 
equiv¬ 
alent 
to the 
carbon 
CX9.4 
Energy 
equiv¬ 
alent 
to the 
nitro¬ 
gen 
NX5.45 
Energy 
equiv¬ 
alent 
to both 
N and 
C 
Possible 
error 
in com¬ 
puted 
heat 
pro¬ 
duction 
221 f: 
Grams 
Calories 
Per cent 
Per cent 
Calories 
Grams 
Grams 
Calories 
Calories 
Calories 
Calories 
8741.... 
±64.7 
4.425 
43.801 
2. 845 
±286.3 
±28. 34 
±1.84 
±266.4 
±10.0 
±276.4 
±9.9 
874 II... 
±52.1 
4. 381 
43. 648 
2.975 
±228. 3 
±22. 74 
±1. 55 
±213. 8 
±8.4 
±222. 2 
±6.1 
887 I.... 
±62.2 
4. 451 
44. 207 
2. 397 
±276. 9 
±27. 50 
±1. 49 
±258. 5 
±8.1 
±266. 6 
±10.3 
887 II... 
±39.3 
4. 394 
43.890 
2.590 
±172. 7 
±17. 25 
±1. 02 
±162. 2 
±5.6 
±167. 8 
±4.9 
8861.... 
±99.1 
4. 560 
45. 368 
2.485 
±451.9 
±44.96 
±2.46 
±422. 6 
±13.4 
±436. 0 
±15.9 
886 II... 
±29.9 
4. 528 
45.101 
1.989 
±135. 4 
±13.49 
±0.59 
±127.3 
±3.2 
±130. 5 
±4.9 
221 g: 
887 I.... 
±79.4 
4.555 
45.469 
2. 557 
±361.7 
±36.10 
±2.04 
±339. 3 
±11.1 
±350.4 
±11.3 
887 II... 
±43.7 
4.547 
45. 303 
2.464 
±198. 7 
±19. 80 
±1.08 
±186.1 
±5.9 
±192.0 
±6.7 
887 III.. 
±49.8 
4.493 
45.123 
2.360 
±223.8 
±22. 47 
±1.18 
±211. 2 
±6.4 
±217. 6 
±6.2 
887 IV.. 
±61.2 
4. 545 
45.323 
2.258 
±278.2 
±27.74 
±1.38 
±260.8 
±7.5 
±268. 3 
±9.9 
It will be noted that the possible 
errors in energy of the feces-and-urine 
mixture, as a result of the possible 
errors in dry matter, appear to be 
appreciable, and they would remain so 
if the nitrogen and, especially, the car¬ 
bon as determined on the dried sub¬ 
stance were not used as a basis for 
computing the loss of nitrogen and 
carbon on drying and of an energy 
correction corresponding to these losses. 
The differences between the nitrogen 
and carbon as determined on the dry 
substance and those determined on the 
fresh represent the loss on drying. A 
positive error in the carbon and nitro¬ 
gen determined on the dried substance 
would, therefore, mean a negative error 
in the computed loss on drying and, 
consequently, a negative error in the 
computed energy correction when such 
a correction is applied, and vice versa. 
As a result of this, the initial error in 
energy is reduced by the energy equiva¬ 
lent of the error in the computed loss of 
nitrogen and carbon during drying. 
These energy equivalents and the net 
effect on the computed heat production 
are shown in Table VII. The result¬ 
ant error is negligible according to this 
method of computation. For example, 
if the error in dry matter in experiment 
No. 221f, 8741 were +64.7 gm. (see 
Table VII), the energy of the daily 
feces-and-urine mixture, as computed 
by multiplying grams dry matter by 
the number of Calories per gram, would 
be too great by the amounts indicated 
in column 5, namely 286.3 Calories 
(64.7 times 4.425). Similarly, the 
figures for carbon and nitrogen as 
determined on the dry substance would 
be too high by the amounts indicated 
in columns 6 and 7 respectively (64.7 
times ^q q-= 28.34 gm. carbon and 
2 845 
64.7 times -jqq-= 1.84 gm. nitrogen). 
On the other hand, by subtracting the 
too large amounts of carbon and nitro¬ 
gen in the dry material from those 
determined on the fresh the results for 
loss of carbon and nitrogen on drying, 
thus obtained, would be too small by 
the same amounts given in columns 6 
and 7. The energy corrections based 
on these losses would then be too small 
by the amounts indicated in columns 
8 and 9, totaling in the case cited 276.4 
Calories (column 10), and making the 
final error in the energy of these excreta 
equal 286.3—276.4=9.9 Calories, or in 
the computed heat production, —9.9 
Calories. 
DETERMINATION OF DRY MATTER IN 
FEED 
It is obvious that an error in the 
determination of the dry matter of the 
feed would directly affect its energy 
equivalent, and also the carbon and 
nitrogen balances. 
Dry matter is determined by first 
subjecting the sample to a preliminary 
determination of the air-dry substance; 
the sample is then ground, and the 
remaining hygroscopic moisture is de¬ 
termined in the ground sample. If 
there is any loss of moisture during the 
grinding, this is not accounted for in 
the determination, and would increase 
the figure for dry matter in the feed. 
The effect of grinding on the dry 
matter determination is a very variable 
factor, the exact magnitude of which is 
unknown, but a preliminary investiga¬ 
tion of this problem indicates that the 
