Dec. 2, 1908.] Agricultural Gazette of N.S.W. 1017 



^iniount of each constituent thcit should be present. This method is much 

 better than another which is sometimes used — viz., to assume that the ash is 

 unchanged in amount, and to use it as the basis of calcuhition ; in the first 

 instance, as we have just seen, the soluble ash is liable to wash downwards, 

 and in the second place the percentage of ash is so small that a trifling error 

 in sampling or determination very considerably affects the result. Of cour.se 

 there is in any case the dithculty of getting the average composition of the 

 green material and the silage, which is only got over by taking a number of 

 samples. In our experiments more than one-third of the total dry matter 

 was lost, including more than half the nitrogen-free extract and the protein ; 

 in the latter case some of the decomposition products, the amino-acids, etc.-, 

 remained in the silo, so that the actual loss of nitrogen was only 2G per cent. 

 The fibre appeared to undergo no change. 



Experimental details. 



Analytical Methods. — The determinations of fat, total nitrogen, fibre, etc., 

 were all made in the ordinary manner. N^on-protein nitrogen was estimated 

 by Stutzer's method, the substance is boiled with a mixture of copper 

 hydrate and glycerine, which dissolves the non-protein but leaves the protein 

 in the insoluble residue. The method is no doubt open to objection, because 

 some proteins may dissolve and some insoluble non-protein bodies, e.g., the 

 purin bases, may remain insoluble. But when used for purposes of com- 

 parison, and not with an idea of getting absolute results, we consider that it 

 gives quite valuable information. 



To determine the quantities of nitrogen present as annnonia, amide, and 

 amino-acid respectively, we adopted the very elegant methods used by Drs. 

 Horace Brown and Millar, described in the Transactions of the Guinness 

 Research Laboratory, Vol. I, Part 1, 1903. We have fouml them work very 

 satisfactorily with our substances. 



The f urfurol obtained on distillation with hydrochloric acid was converted 

 into its hydrazone and weighed. Krug's method of working was adopted, 

 but we prefer not to attempt expressing our results in terms of any particular 

 pentosan. It is well known that several groups of bodies give f urfurol on 

 distillation, including pentosans, celluloses (the so-called " oxy-celluloses "), 

 glycuronic acid, etc., but that certain members of the groups yield instead 

 non-volatile Iwdroxy-furfurols, and hence cannot be estimated in this way. 

 In view of the fact that the furfurol yielding bodies of green maize ha\e not 

 been carefully examined, we prefer to give the experimental figure only, and 

 in the Tables have given the actual weight of furfurol obtained from lUO 

 parts of substance. No doubt the less resistant or "oxy-celluloses" are 

 responsible for much of the furfurol. 



Method of Procedure.— The maize was cut up, roughly dried at about SO' 

 C, ground in a coffee mill, and reduced to a fine powder in the Maercker 

 mill. Drying could now be completed, and samples taken tor analysis. 



A different scheme had to be adopted for silage. As the bags were 

 recovered from the silo they were weighed, Ijrought to the laboratory, and 



