1090 
Journal of Agricultural Research 
Vol. XXXI, No. 11 
Table III. — Increments of heat production per kilogram of dry matter of feed — 
Continued 
Ani¬ 
mal 
Feeding stuff 
Periods compared 
H 
J 
J 
K 
K 
Alfalfa hay__ _ 
1 and 3... 
Do ___ . _ _ 
3 and 5. _ 
Do . .. 
1 and 5._... 
Average.._ 
Alfalfa meal___ _ 
2 and 4.. 
Do ... _ _ _ . 
4 and 6... _ _ 
Do_ 
2 and 6_ 
Average.._ 
Alfalfa hay___ 
5 and 6.. 
Do__ 
6 and 7_ 
Do.... 
5 and 7.. 
Average.. _ 
Alfalfa hay and starch__ 
1 and 2.. 
Do_ 
1 and 3__ 
Do.. 
1 and 4_ 
Do_ 
3 and 2._.. 
Do 
2 and 4_ 
Do__ 
3 and 4_ . 
Average..___ 
Alfalfa hay and grain mixture No. 3 _ 
1 and 2_ 
Do _ _ r__ 
3 and 4.. 
Red clover hay __ 
1 and 2_ 
Red elovor hay and mm mp.al 
3 and 4_ 
Do . 
4 and 5_ 
Do .. _ 
3 and 5._.. 
Average. ___ 
Experiment No. 
Heat in¬ 
crement 
per kilo¬ 
gram 
of dry- 
matter 
212 . 
216. 
217. 
220 . 
Cals. 
980 
1,080 
1,043 
1,034 
934 
1,287 
1,160 
1,127 
1,066 
1,006 
1,025 
1,029 
1,188 
1,334 
1,198 
855 
1,239 
1,028 
1,140 
1,206 
1,360 
921 
1,132 
980 
(658) 
1,056 
* Fattened. 
VARIABILITY OF HEAT-INCREMENT VALUES 
The variations in the heat increments per kilogram of dry matter 
exhibited in Table III call for a consideration of their true signifi¬ 
cance and of several factors possibly contributing to apparent 
discrepancies. 
These variations are not to be considered in the same lig;ht as vari¬ 
ations in other determinations, for example, of digestibility or of 
metabolizable energy. We may have, for instance, in two periods, 
independent determinations of metabolizable energy per kilogram of 
feed, but, in connection with the same, only one heat-increment 
value. On the other hand, three periods on the same feed make 
possible three comparisons, and, therefore, as many determinations of 
heat increment; and four periods make possible six comparisons, and, 
therefore, six determinations of heat increment, while no one of these 
heat-increment values represents the result of a single period. Vari¬ 
ations in heat increments, therefore, can not be referred to the heat 
production of individual periods. The extent of these variations, 
and their effects on the average heat increment per kilogram of feed, 
depend not only upon the magnitude of the errors in the values rep- 
