422 PRINCIPLES OF ANIMAL NUTRITION. 
The great variations in the results, as well as the large propor- 
tion of cases in which the availability appears to exceed 100 per 
cent., show that little value attaches to them as quantitative deter- 
minations, although they undoubtedly show that rhamnose pos- 
sesses a comparatively high nutritive value. 
Crude Fiber—The experiments of v. Knieriem have already 
been cited in Chapter V in their general bearings upon the metabo- 
lism of matter. As was there noted, certain corrections were neces- 
sary on account of the residue of undigested cellulose remaining 
in the digestive canal at the close of the experiment. The results 
given below are based on those computed by the author, as sum- 
marized on p. 161, on the assumption that the resorption of the 
remaining digestible crude fiber was complete after two days. 
B Metab lie 
A olizable ee 
os] ws Food per Day. Energy Over ability 
- ys a Total, eee Per Cent. 
&| ez H Cals. é a 
D9: | Millkies cova es caer se ox ue ns 341.7 |—46.8 saahte pais 
II} 10 | “ + 22grms. crude fiber ' ied 39.9 | 121.3* 
for eight days..... -874.6 |— 6.95) 31°65 -| 126.54 
TED fe MAME os aiscse:o ecpcscaie 400 iets 8 350.1 |—37.9 Siete eaten 
* Compared with Period I. {+ Compared with Period III. 
It is evident from the above figures that while the experiments 
show qualitatively a nutritive value for the cellulose, they are in- 
sufficient for a quantitative determination of its amount. 
In striking contrast with these results are the conclusions drawn 
by Zuntz & Hagemann from their experiments upon the horse 
which have already been considered in the previous chapter (pp. 
389-391). As was there explained in detail, these investigators 
have estimated the expenditure of energy in the digestion of crude 
fiber from a comparison of the computed heat production in two 
sets of experiments in which the proportion of coarse fodder eaten 
differed considerably, it being assumed that 9 per cent. of the metab- 
olizable energy of the nutrients other than crude fiber was consumed 
in their digestion. On this basis the digestive work caused by the 
crude fiber is computed at 2.086 Cals. per gram, or rather more than 
