Quantitative Estimation of the Carbohydrates. 109 



which removed about 97% to 98% of the sugars. 



Davis, Daish and Sawyer (1916) consider this method of treat- 

 ment is unsatisfactory in the case of moderately thick leaves such 

 as that of the mangold, where heating up may be slow and a certain 

 amount of enzyme action is possible before the enzymes are 

 destroyed. They therefore adopt the followinjf method in their 

 work. About 1 kilo, of freshly picked leaf material is dropped in 

 small quantities at a time into 2 litres of boiling 95% alcohol con- 

 tained in a large zinc beaker to which 20 c.c. of ammonia of S.G. 

 0'880 is added in order to neutralise the acids present in the leaf. 

 After boiling the alcohol for half an hour the fui'ther extraction of 

 the alcohol-soluble contents of the leaf is carried out in an extraction 

 apparatus on the principle of the Soxhlet extractor. The extraction 

 is complete after 12 to 18 hours. The final separation of the 

 extract from the residue is effected in a Buchner press. The 

 residue is dried on paper trays in a steam oven for 18 hours and 

 from it the total insoluble matter, the starch and pentosans are 

 estimated. The alcoholic extract is analysed for total soluble 

 matter, sucrose, maltose, glucose, fructose, and pentoses. It can 

 be kept in a waxed-corked bottle for 3 to 6 months without any 

 change occurring in the sugars if about 10 c.c. to 20 c.c. of toluene 

 are added. 



2. Estimation of Starch. 



Brown and Morris estimated the starch in their dry leaf powder 

 by O'Sullivan's method (1884) which consists in converting the 

 starch into a mixture of dextrin and maltose by means of diastase. 

 The leaf material usually contains tannins, amino-acids, etc. which 

 influence the optical activity and reducing power of a solution and 

 these accompanying substances have therefore to be removed by 

 precipitation by means of basic lead acetate. Davis and Daish 

 (1914) find that this method does not give correct results 

 because some of the dextrin is carried down with the precipitate 

 and so is lost to the analysis. It is estimated that as a result 

 of this the starch estimations in leaf material made by O'Sullivan's 

 method may be 15% to 20% below the actual starch content. 



Davis and Daish therefore treat the plant material with taka- 

 diastase which converts starch wholly fnto maltose and dextrose 

 which are then estimated by measurement of the cupric reducing 

 power and the optical activity. The leaf material is first treated 

 for 24 hours at 38"C with about 20 times its weight of water con- 



