344 PROCEEDINGS OF SECTION B. 



The Determination of Moisture. — We have considered it sufficient 

 for practical purposes to dry in the steam oven at 98°-99° C. 



The Determination of Protein. — The total nitrogen is determined 

 by the Kjeldahl method, and the " true proteid nitrogen " by Stutzer's 

 cupric-hydrate method. The water soluble nitrogen is also determined, 

 and the soluble and insoluble protein calculated, by the use of the 

 factor 6-25. "We consider, on account of their wide variation in amount 

 at different stages of plant growth, and their inferior food value, the 

 determination of amides is of considerable importance. 



The Determination of Crude Fibre. — The method of Konig (Zeit. 

 fiir Untersuch der Nahrungs-und Genussmilttel, i., 3-16, Abstr. Analyst, 

 1898, XXIII., 47) for the determination of crude fibre has found wide 

 acceptance among European chemists, and has been largely used by us. 

 According to Konig, a body consisting of cellulose and lignin, and 

 practically free from pentosans, is obtained, though it is slightly richer in 

 nitrogen than the fibre of the Henneberg (acid-alkali) process. This is, 

 however, of little consequence in materials poor in nitrogen, and we 

 have found the amount of contained pentosans to be negligible. 



The method has been unfavorably criticised by American chemists 

 (Nineteenth Annual Convention, A.O.A.C., 1903, Referee's Report on 

 Cattle Feeds) on account of its slow filtering and variation in duplicates. 

 We prefer to filter through a layer of glass-wool with suction, and, pro- 

 vided the material be kept off the filter till well washed, no special 

 difficulty has been experienced. Duplicates agree within 1 per cent. 

 The method is recommended as giving an estimate, practically free of 

 pentosan, of the least valuable portion of feeding materials. Since 

 pentosans are separately determined, this is to be commended. In 

 treatment of materials rich in protein we would suggest the further 

 digestion of the fibre with dilute alkali for the removal of residual 

 protein which Konig has shown may amount to between 1 per cent, 

 and 2 per cent. Or an estimation of the contained nitrogen may be 

 made and correction of the percentage of crude fibre. In our work 

 this refinement has not been introduced. Determinations of crude 

 fibre in a number of fodder materials by acid-alkali and Konig methods 

 follow in table II. 



The Determination of Pentosans. — We employ the method of Tollens, 

 as modified by Krober and Rimbach (Abstr. Analyst, 1902, p. 279). 

 This method has been found so satisfactory that no trial has been made 

 of the recent method of A. Jolles (Abstr. Analyst, May, 1906, p. 163), 

 in which the furfurol is volumetrically estimated by titration with 

 sodium bi-sulphite solution. 



The Determination of Starch. — It had been our custom to determine 

 the starch in fodder materials by the method of Maercker and Morgan, 

 as for grain (Konig: Untersuchung landwertshaft-licker Stoffe, p. 221), 

 the determination generally being carried out on the material without 

 previous extraction of the existing sugars, for which allowance was 

 made in the starch figure obtained. In spite of the contention of 

 Sherman (School of Mines Quarterly, 1896, xvii., 356-365, Abstr. 



