THE UTILIZATION OF ENERGY. 475 
In Kellner’s two experiments on heavy rations the enormous losses 
in the feces cut down the percentage utilization to a very low 
figure and thus render difficult a direct comparison with the 
other averages. 
While the above form of stating the results appears the simplest 
and most direct, it is of interest also to eliminate the influence of 
varying digestibility by computing the percentage distribution 
of the gross energy of the apparently digested portion of the starch. 
This is particularly the case since Kellner’s computations of his 
experiments are made in a somewhat similar way. Combining 
the data given on p. 461, regarding the percentages of metaboliz- 
able energy utilized, with those on p. 301 for the energy of the 
apparently digested matter, we have the following: 
DISTRIBUTION OF ENERGY OF APPARENTLY DIGESTED STARCH. 
Work of 
Digestion, 
In Urine. | In Methane. |Assimilation,}| In Gain. 
Per Cent. Per Cent. and Tissue Per Cent. 
Building. 
Per Cent. 
Kithn’s experiments ........ —1.19 13.42 43 .89 43 .88 
‘ Kellner’s experiments: 
Moderate rations......... —0.92 11.12 37 .36 52.44 
Heavy rations ........... —4.95 —6.15 42.77 68.33 
Kellner’s computations are made in a different manner.* Omit- 
ting in the computation of metabolizable energy the correction for 
nitrogen gained or lost, he compares the period in which starch was 
fed with that on the basal ration substantially as has been done 
above. He then, however, introduces a correction for the influence 
of the starch upon the digestibility of the basal ration. For ex- 
ample, comparing Periods 3 and 4 on Ox H, he finds in the manner 
shown on p. 307, Chapter X, that the equivalent of 820 Cals. less 
of the basal ration was digested in the period in which starch was 
added to it, while there is a further correction of 112 Cals. to be 
made for the less amount of organic matter of the basal ration con- 
sumed in Period 3, making a total difference of 932 Cals. Of the 
gross energy of the basal ration, 79.9 per cent. was found to be met- 
* Compare Landw. Vers. Stat., 53, 450. 
Digitized by Microsoft® 
