THE NITROGEN-FREE EXTRACT OF FEEDING STUFFS. 651 



A fuitlier cause of difference in results lies in the uuderstandiuo- of 

 wluit is meant l>y 1.25 per cent sulplnuic acid, and 1.25 per cent potas- 

 sium lij^droxid, and this is seldom defined in the description of methods. 

 Do the percentages refer to SO3 or H2SO,, to K2O or KOH ? Henne- 

 berg does not specify in his original description of the method, but in 

 another paper' he states explicitly that he uses "50 cc. of 5 per cent 

 sulphuric acid (50 gm. suljihuric hydrate per liter) and 150 cc. of water," 

 aiul also " 50 cc. of 5 per cent potassium liydroxid (50 gm. of fused 

 caustic potash per liter)." 



This is further confirmed by Prof. F. Lehmann, Professor Henneberg's 

 successor at the station, who states that in making the 5 per cent stock 

 solutions 255 gm. of concentrated i)ure sulpliuric acid (98 per cent 

 H^SOj) is mixed with 4,74:5 gm. of water, and 300 gm. of potassium 

 hydrate is dissolved in 6 liters of water. 



E. Wolff,"' in his book on agricultural analysis, directs that in making 

 the 5 per cent potash solution 50 gm. of fused caustic potash be dis- 

 solved in 1 liter of water. 



From the evidence adduced, the writer believes the correct solutions 

 to be 1.25 per cent ILSO, and 1.25 per cent KOH. In a method 

 whose regularity depends upon using solutions of constant strength, it 

 is very necessary to follow exactly the original directions of the author, 

 or at least the method by which the majority of analyses are made. 



Experiments made in the United States have shown that sulphuric 

 acid and potash solutions of different strengths leave different amounts 

 of crude fiber undissolved. For instance, Huston and McBride^ found 

 that filter paper treated with 2,5 per cent sulphuric acid and 2.5 per 

 cent potassium liydroxid lost 17 per cent, and even more by other 

 treatment. 



In other instances linseed meal^ gave with 1.25 per cent solutions 

 8.13 per cent of crude fiber, and with 2.5 per cent solutions 7.35 per 

 cent of crude fiber, with differences of about 1 i^er cent between the 

 results of different analysts. ^ 



Cellulose (dicai/s atiach'd in the Weende method. — As cellulose is a 

 principal constituent of crude fiber, the question as to the nature and 

 behavior of crude fiber can be answered at least in part by determining 

 whether in the Weende method the cellulose of feeding stuff's remains 

 intact or whether it is partly dissolved and so goes to swell the 

 nitrogen-free extract. 



For this ])urpose investigations on pure cellulose treated according 

 to the Weende method will be helpful. 



Perhaps the first to make investigations along this line was H. 



'Laudw. A^ers. Stiit., 6 (1864), p. 4i>7. 



- Aiileituiijf zur cheiiiisclicu Uutersnchniig- hmdwirtlischaftlicli wichtiger .Stolie, 

 Berlin, 1875, Sd ed., p. 17.",. 

 ■^Associatiou Otilicial Agricultural Chemists at Chicago, 1893, p. 140. 

 'Ibid., p. 131. 



