660 EXPERIMENT STATION RECORD. 



It is seen that the sum of the criule fiber and nitrogen-free extract 

 digested bj- animals in many cases agrees ahnost exactly with the total 

 amount of nitrogen-free extract lound by tlie Weende method. In 

 some instances, as in Nos. 129 and 225, the agreement is complete, 

 while in most cases the difference amounts to onh'- a few per cent, being 

 large in only a few cases {'Nos. 24 and 178). 



As a rule, then, the ]iercentage of nitrogen-free extract found by the 

 Weende method is approxinuitely equivalent to the sum of the digesti- 

 ble crude fiber and tlie digestible nitrogen-free extract. 



The difference hefn-ecn the nitror/en-free extract anil the aqueous extract 

 gives the digestible ceUnJose. — We can go one step farther and combine 

 the figures for nitrogen-free extract and aqueous extract. As stated 

 above, the aqueous extract agrees approxinuitely with the digestible 

 nitrogen -free extract, and it follows that deducting the percentage of 

 aqueous extract from the percentage of nitrogen-free extract will give 

 as the remainder the percentage of uudigestible nitrogen-free extract. 

 But, as stated above, the undigested portion of the nitrogen-free 

 extract is approximately eijual to the digested portion of the crude 

 fiber; and hence subtracting the percentage of aqueous extract from 

 the percentage of nitrogen-free extract gives the percentage of digesti- 

 ble cellulose. 



Although these rules are only approximately correct, comparing in 

 some cases very well and occasionally not so well with the results of 

 digestion experiments, the figures obtained by means of them are at 

 least good indications of the value of feeding stuffs, and they can be 

 widely employed since they are derived with much less time and hibor 

 than the more exact figures obtained in digestion experiments. 



It is recommended to continue the use of Ilenneberg's original 

 Weende method of analysis until a better method is elaborated, and to 

 supplement this by tlie determination of aqueous extract, which here- 

 tofore has not been general. 



If more explicit information is desired as to the various constituents 

 of the nitrogen-free extract, the starch, sugar, organic acids, etc., they 

 may be determined by the methods referred to above. 



The pentosans should be more frequently determined than at present, 

 because it is probable, though not entirely certain, that the pentosans 

 have the same nutritive value as the hexose carb(diydrates.' 



CONCLUSION. 



From the above discussion of the subject it follows that the Weende 

 method for the analysis of feeding stuffs is not to be regarded as an 

 exact method, since the constituents determined by it are not chemically 



'See investigations of Stone and Jones, Ber. deut. cliem. Ges., 25 (1892), p. 563; of 

 Lindseyand Holland, Massachusetts State Sta. Rpt. 1894, p. 186; of Weiske, Ztsehr. 

 pliysiol. Chem., 20 (1895), p. 494; of Pfeiffer, Landw. Vers. Stat., 47 (1896), p. 59; of 

 Tolleus, Jour. Landw., 44 (1896), p. 171. 



