THE FOOD AS A SOURCE OF ENERGY. 325 
diminish it slightly. On the other hand, the formation of hydro- 
carbons was markedly increased except in two cases. It has al- 
ready been shown that the proteids of the food do not take part in 
the production of these gases, and that the same is probably true 
of the fat under normal conditions. Neglecting the small effects 
upon the urine, therefore, we may compare directly the increase in 
the digested carbohydrates with the increase in the gaseous hydro- 
carbons, using for this purpose the differences between the two 
rations uncorrected for the slight variations in the consumption of 
dry matter. 
Taking first the last five of Kellner’s experiments,* which seem 
to represent the most normal conditions, we have the following: 
Difference in Carbohydrates 
Digested. Difference in 
Energy of 
Nitrogen-free ar a 
Crude Fiber, , zs 
| “eta” | Bataet 
Ox D, Period 2............. —64 +1388 +424.4 
pages th Ne eBay dea ee sack elsy —64 +1609 +822.0 
“G OO And ec cahelirescat —50 +1598 +645.8 
te EE) AS Ber sae are scauesee —26 +1861 +604.5 
ohare} SS Ou sees seek e es —9 +1501 +769.9 
LOtals) 0 sie ee Aaeoatwia See doe —213 +7957 3266.6 
Assuming that the same proportion of hydrocarbons is pro- 
duced in the fermentation of crude fiber as in that of starch, we 
may compare the algebraic sum of the two with the energy of the 
methane as follows: 
3266.6 Cals. + (7957 — 213) =0.422 Cals. per gram. 
Subtracting the latter result from the gross energy of the digested 
nitrogen-free extract of starch, we have for the metabolizable 
energy of the latter 
4.185 Cals. — 0.422 Cals. =3.763 Cals. per gram. 
In the experiments on Oxen B and C the basal ration was a 
heavy one, with a rather wide nutritive ratio, and already con- 
tained large amounts of digestible carbohydrates. Under these cir- 
cumstances the added starch was very imperfectly digested, while 
* Loc. cit., 68, 422. 
