PROCEEDINGS OF SECTIOX B. 345 



Analyst, 1897, p. 19), that the method of Maercker gives results com- 

 parable with those of the diastase method, we have come to consider 

 it quite unsuitable for the estimation of " true " starch in grasses and 

 like materials. The susceptibility of pentosans and hemi-celluloses to 

 hydrolvsis with the formation of reducing bodies has been noticed by 

 Krug and Wiley (J. Amer. C.S., 1898, xx., 266-268). 0. Lietz (Abstr. 

 Analyst, 1903, p. 150) states of a variety of methods, including digestion 

 under pressure, that none give accurate results, owing to hydrolysis of 

 pentosans. The influence of the presence of pentosans upon the 

 estimation of starch by Maercker and Morgan's method is undoubted. 

 That these bodies are rendered soluble is shown by the fact that the 

 residues from the autoclave of two grasses (Astrebla pectinata and 

 Panicmn jubiflorum), themselves containing respectively 20-52 per cent, 

 and 20*32 per cent, pentosans, yielded only 7*96 per cent, and 7*98 per 

 cent. The method described by Lietz {loc. cit.) has not invariable given 

 lower starch percentages. If the pentosans are unaffected by the solutions 

 employed it is evident that substances of cellulose nature are brought 

 into solution, and are reckoned as " true " starch. No test of the 

 author's contention as to the non-interference of pentosans has been 

 made by us. A short experiment was carried out upon a sample of 

 corn-cob husk, previously washed to remove water-soluble material. 

 A qualitative examination led us to believe that the amount of starch 

 present was very small. Determinations of starch were made in this 

 substance by autoclave digestion at 5-6 kgm. per sq. cm. ; 2-3 kgm. 

 per sq. cm. by Lietz's method, and by the method of Wilforth, as given 

 in " Researches on the Assimilation of the Elements of Nutrition of 

 Plants during different Periods of their Growth " (translated by Emslie : 

 Vinton & Co., London, 1905). We consider this last identical with the 

 diastase method referred to by other chemists. The starch percentages 

 by the different methods, together with the percentage of pentosan 

 found in the autoclave or diastase extracts, and which would be esti- 

 mated as starch, are herewith given. 



The sample of husk itself was found to contain 37*83 per cent, 

 pentosan and 29*50 per cent, crude fibre (Konig). 



We conclude that the diastase method gives the most correct 

 estimate of the " true " starch in plant substances of the class dealt 

 with. Li view, however, of the possible close approximation in nutritive 

 value to starch of the substances soluble under the conditions of the 

 Maercker-Morgan method, we would tentatively recommend its pro- 

 visional adoption, and the estimation of all substances behaving similarly 

 to starch under well-defined conditions as such. 



