400 
Journal oj Agricultural Research 
Vol. XXX, No. 5 
Recurring now to the computation of 
the heat production in the steer 
experiments, the feces and urine were 
collected, dried, and analyzed sepa¬ 
rately. The carbon in the urine was 
determined in the fresh sample, and the 
correction for energy lost on drying 
was computed from the nitrogen lost. 
The carbon in the feces was, however, 
determined in the air-dried sample, 
ignoring the possibility of loss of carbon 
and energy during the drying. There 
is no data covering the losses on drying 
steer feces. There is, however, data 
which show a considerable los^ of 
carbon in drying cow feces (not a 
mixture of urine and feces), which also 
implies a loss of energy. The results 
of the analysis of several samples of 
cow feces, which were collected by an 
attendant on a single day of the 
digestion experiments, are given in 
Table V. 
POSSIBLE ERROR IN THE DETERMINA¬ 
TION OF DRY MATTER OF FECES-AND- 
TJRINE MIXTURE 
The determination of the dry matter 
of the feces-and-urine mixture is 
another source of error to which 
attention is called. 
For determinations of energy, nitro¬ 
gen, carbon, etc., in the combined 
feces and urine, the composite sample 
representing several days’ collection 
was used. In all cow experiments 
txiis composite was made up of 10 
daily aliquots. The dry matter, how¬ 
ever, was determined in this composite 
and also in each of the daily samples. 
The same routine was followed in the 
drying of the daily samples as in the 
case of the composite, except that the 
former were generally left in the dry¬ 
ing closet for a longer period of time, 
since, as a rule, they were placed in the 
Table V.— Losses of nitrogen and carbon occurring during the drying of cow feces 
Experiment and period 
Dry matter 
of feces 
Loss of nitrogen 
Loss of carbon 
Loss of 
carbon 
per gram 
nitrogen 
lost 
Grams per 
Grams per 
Grams per 
221 f: 
day 
Per cent a 
day 
Per cent « 
day 
Grams 
874 I_ 
2,013. 7 
0.092 
1.86 
1.045 
21.0 
11.29 
874 II_ _ . 
1,409. 7 
1 . 208 
17.0 
887 I_ 
1, 560. 2 
.367 
5. 73 
3. 379 
| 52.7 
9.20 
887 II_ 
1,086. 3 
.240 
2 . 61 
1.891 
20.6 
7.89 
886 I_ 
2 , 762. 2 
.236 
6 . 53 
1.872 
51.7 
7.92 
886 II_ 
2,796. 0 
.208 
11.41 
3.114 
87.1 
7.63 
221 g: 
887 I_ 
2,755.4 
.165 
4. 54 
1. 672 
46.1 
10.15 
887 II_ 
1,147. 2 
.124 
1.43 
1.000 
11.5 
8.04 
887 III___ 
1, 583.4 
.160 
2. 53 
1. 384 
21.9 
8.66 
887 IV___ 
2, 665.1 
.096 
2. 56 
3. 228 
86.3 
33.71 
a Computed to dry-matter basis. 
These results show that while the 
loss of nitrogen is slight, the loss of 
carbon is several times as great. 
It would, of course, be unsafe to 
assume that identical losses occurred in 
the drying of steer feces since the 
composition of the rations, and, there¬ 
fore, presumably, of the feces differed 
from that applying to the experiments 
with cows. Furthermore, the losses on 
drying are shown to be too variable to 
be regarded as accurately applying 
to other conditions. However, the 
data suggest the probability of signifi¬ 
cant loss of carbon during the drying of 
steer feces also. 
The bearing of the facts as to losses 
on drying, as disclosed by the above 
data, is also obviously important in 
connection witli digestion experiments. 
closet as soon as obtained, and were 
allowed to remain until the composite 
sample also was thoroughly dry. 
In all cases but one the percentage 
of dry matter as determined on the 
composite was found to be higher than 
the true average of determinations on 
the daily samples. Such a uniform 
difference must have an explanation 
other than errors of sampling or weigh¬ 
ing. Aside from these errors, such 
differences could be due either to an 
unweighed evaporation of moisture 
from the composite during preparation, 
or to a greater loss of dry matter during 
the drying of the daily samples. Con¬ 
sidering the precautions taken to pre¬ 
vent evaporation the loss by this means 
could have been but slight. The loss 
of dry matter due to fermentation 
