514 EXPEEIMENT STATION RECORD. 



the maintenance values of feeding stuffs, in which a beginning has 

 been made by the Pennsylvania Station. The residts thus far reached 

 hardly do more than show the need for further investigation. The 

 experiments should be rejoeated witli additional animals and extended 

 to cover at least typical members of the dilferent classes of feeding 

 stuffs. Several years' w^ork might be expended profitably on this 

 single branch of the subject. Were this done we might fairly hope, 

 I think, to secure approximate factors which could be applied to 

 those feeds whose maintenance values had not been directly deter- 

 mined. If successful we should secure an indispensable factor for 

 the discussion of the results of productive feeding, since it is evident 

 that we must know liow much of our ration has served simply for 

 maintenance before we can rationally consider its productive value. 



As regards the production values of feeding stuft's, we are better off, 

 in that we have the results of Kellner's elaborate and exceedingly 

 valuable determinations. For the feeding of mature fattening cattle 

 it seems probable that the " production values "■ thus worked out may 

 at least be regarded as a very close approximation to the truth. Here 

 again, however, it is to be feared that we are in danger of repeating 

 the mistakes of HennebergV; successors. Kellner has published a very 

 complete table of actual or computed production values. It would not 

 be sufprising if there should develop a tendency to accept these values 

 as measuring the worth of feeding stuffs for all productive purposes. 

 In other words, there is danger that we may rest in an assumption 

 instead of actually determining the facts for ourselves. The results 

 of our Pennsylvania experiments certainly indicate that the values for 

 maintenance are higher than those for production. It would not be 

 at all surprising if the jDroduction values for different purposes or 

 for <lifferent classes of animals should prove to be materially different. 

 At any rate we have no warrant in advance of investigation for 

 assuming that thej^ are identical. We need similar determinations 

 by equally rigorous methods of the productive values for other pur- 

 poses, which are by no means necessarily the same as for fattening. 

 Here, again, Ave have a vast field open to the qualified and patient 

 investigator, equipped with the necessary appliances. 



It may be objected that investigations such as those just outlined are 

 to a degree empirical. They treat a single feeding stuff as an entity, 

 while as a matter of fact it is not. We can determine the net available 

 energy of a given sample of timothy hay, but we can hardly be said 

 to determine the available energy of timothy hay. The justification 

 for making such determinations lies in the fact that they promise to 

 yield approximately correct and useful results within a reasonable 

 time. Too great emphasis can hardly be laid upon the importance of 

 studies of the physiological values of individual chemical ingredients. 

 Such results, when once established, are definite, because thev refer to 



