l.")G Tlie Starch Kf/uim/nif T/iror// 



shows in his appendix talilos' that tho maintfiuanc*^ ration from wliioh 

 his <) lb. of starch ('((uivali'iit is (lorivcd must contain ()-(i to (>!S lb. of 

 digestible protein, (•■! of diu'cstiblc fat and T-") to '.••") lb. of digestible 

 carbohydrates and fibre. This anioimts to from 81 to lo-j lb. of "starch 

 equivalent" reckoned by the formula given by Wood and Yule. 



We have here to deal with a serious defect in Kellner's system. 

 He found ^ that, on the average, only some 50 to 70 per cent, of the 

 available energy of the foods used for maintenance rations is of dynamic 

 value. Consequently the starch e(juivalents of these foods are from 

 30 to 50 per cent, below the full value. The maintenance requirements 

 of the animals, when expressed in terms of .starch equivalent, must 

 therefore be reduced by a similar amount. The quantities of feeding 

 stuffs necessarv to maintain the aninials may be estimated with 

 approximate accuracy on this basis provided the foods are of the value 

 assumed in the estimate of the animal's rei|uirements. but not otherwise. 

 For example, the quantity of niai/e meal which corresponds to G lb. 

 of starch equivalent would be too small for maintenance of a 1000 lb. 

 ox because it yields little or no thermic energy in addition to that 

 represented by the starch ecpiivalent. In practice, of course, no one 

 would use maize meal for maintenance rations of oxen; but in the 

 investigation undertaken by Wood and Yule the fattening ration 

 (most of which it may be assumed was of nearly full value) was assigned 

 for maintenance functions in order that the fattening value of the roots 

 might be determined. In this case lb. of Kellner's starch equivalent, 

 derived from such foods, would be inadequate. As Wood and Yule 

 did not etnploy Kellner's formula but another which includes the whole 

 of the available energy of the food, this arg\iin('nt does not apply. It 

 remains to be seen, however, whether the r)-35 lb. of starch equivalent, 

 reckoned by that foirnula. was sufficient. 



It will be seen that Kellner's estimate of fi lb. of starch eq\iivalent 

 for maintenance is a purely artificial number arbitrarily adapted to 

 the needs of his convention. That being so. it makes little diiiierence 

 whether the ])rotein be niulti])lie(l by 1-25 or by 0-94. The net result 

 of using the higher factor is that the estimates of the requirements 

 for maintenance must be correspondingly increased. Kellner appears 

 to have deliberately decided to ignore the differetice between the heat 

 producing and fat producing values of protein in order to attain that 

 uniformity which his system demands, and because he saw that it 



' The Srienlijic Fmliiig nf Aiiiiimli. p. 3112. - Ibid., \\ .■JGO el scq. 



