METHODS OF ESTIMATING CARBOHYDRATES 49 



under conditions such that all the maltose (and about 

 98 per cent, of the glucose) is destroyed, it is found that 

 about 3 per cent, of fructose has also disappeared. Accord- 

 ingly it is necessary to proceed as follows: Quantities of 

 the three reducing sugars, glucose, fructose, and maltose, 

 are weighed out and treated separately with bromine, 

 maintained as a saturated solution, for some convenient 

 period about sixty-six hours at 16. By this procedure 

 the major part of the glucose and maltose is destroyed, 

 whilst the fructose is only slightly diminished in quantity. 



The following three equations may then be obtained, so 

 it is possible to solve for the three unknowns : 



The original rotation of the solution 



(1) 100 a = a [a] 2 D o + 6[a]/ + c [a^m, 



where a, 6 and c are the original amounts of the sugars. 

 The initial reducing-power, in terms of cupric oxide 



(2) w=ak' g + bk'f+ck' m , 



in which k' g) k' f and k' m represent the weights of oxide 

 obtained when definite quantities, say 0-1 gramme, of the 

 various sugars are used. 



After treatment with bromine, only quantities ak g , bk f , 

 and ck m , are left. The final reducing-power will then be 



(3) w' = ak g V +bk f k' f + ck m k' . 



It is more convenient to determine the final reducing- 

 power than the final rotation, since the oxidation products 

 are optically active and corrections have to be made for 

 the rotation due to them. 



If in other experiments the time or temperature is varied 

 slightly, it is possible to substitute new values of k g , k f , and 

 k m , or to correct k f and k m by a fresh determination of k,. 

 This method is being carefully tested. [See note, p. 52.] 



4 



