VOLUMETRIC ESTIMATIONS 93 



sometimes cause the precipitate to subside ; if not, it will 

 be necessary to boil a fresh portion of the original solution 

 and then to add lead acetate; after filtering, the filtrate is 

 saturated with hydrogen sulphide to remove excess of lead, 

 and the titration is then carried out on the filtrate after boiling 

 off the hydrogen sulphide. 



Plant extracts may also contain tannins which must be. 

 removed before estimating the sugars. There are various 

 ways of doing this ; and in all cases it is best to try the 

 separation with a small portion of the material first, in order 

 to determine the efficiency of the method. 



1. The aqueous solution is mixed with about half its vol- 

 ume of ethyl acetate, and thoroughly shaken. The mixture is 

 then placed in a separating funnel until the two solvents have 

 separated, when the lower aqueous part, containing the sugar, is 

 drawn off, and with it the process is repeated. The aqueous 

 solution is now warmed on a water bath until all the ethyl 

 acetate is driven off, and the cooled solution tested for tannin. 



2. To every 100 c.c. of the solution add 4 grams of pure 

 lead carbonate. Shake thoroughly at intervals for about four 

 hours. Test the filtrate for tannins. 



The chief objection to this method is that lead salts soluble 

 in water may be formed. 



3. Add gelatine solution until no more precipitate is formed. 

 Filter and wash the precipitate very thoroughly. The filtrates 

 resulting from any of these operations may then be examined 

 for the sugar. 



Estimation of Pentoses. 



When pentoses alone are present, they may be estimated 

 in the same way as glucose ; if, however, they are mixed with 

 other carbohydrates, or are present in the form of pentosanes 

 other methods must be employed (see p. 101). 



Estimation *of Glucose. 



The following precautions must be taken in estimating 

 glucose by this and similar methods involving the reduction of 

 metallic salts.* 



* For the estimation of levulose in the presence of glucose see Loewe : 

 " Proc. Soc. Exp. Biol. Med.," 1916, 13, 71. 



