greater live weight than actually existed. A better way, doubt 

 less, and one adopted in another experiment, would have been 

 to have added one-half oi this gain, thus figuring the ration on 

 approximately the average live weight for the week. On the 

 other hand, it may be reasonably urged that so long as the food 

 was consumed without waste it was evident that these particu- 

 lar pigs, at least, were so constituted that they could handle 

 more than Wolff's standards call for. 



HOW THE RATIONS WERE COMPUTED. 



As already mentioned, it was decided to make the grain ra- 

 tion a mixture of equal parts of corn meal and wheat middlings, 

 as this mixture would have a nutritive ratio of i : 6.5, which was 

 exactly what was desired. A sufficient amount was then taken 

 to meet the requirements for the live weight involved, estimated 

 as above explained. 



The lot not having the mixed grain were to have such a 

 mixture of skim milk and corn meal as should give exactly the 

 same amount and proportion of digestible matter. It was found 

 that one part of corn meal and two parts of skim-milk gave near- 

 ly the right proportion, or nutritive ratio, and in general this was 

 the combination fed. 



As the feeding progressed, samples of the foods were ana- 

 lysed, and when it was observed that the skim-milk ration inva- 

 riably gave the greater growth, the digestibility of each constit- 

 uent of the food was determined, as reported in Part II, where 

 will be found the " Composition of Feeding Stuffs," Table I, 

 page II ; "Composition of Dung," Table II, page 12 ; and Di- 

 gestion Coefficients, Table IV, page 13. 



The following table (I) is computed from the total composi- 

 tion and the digestion coefficients given in Part II, and is in the 

 form which I have used for the last six years in feeding tables; 

 the first column gives the digestible albuminoids, the second the 

 non-albuminoids made up of digestible nitrogen-free extract, fibre 

 and fat, the latter multiplied by 2>^ to place it on a starch basis: 



TABLE I. 



