[Vol. 10 



224 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



flask, this being stoppered and shaken and permitted to stand for 

 10 minutes. Five cc. of cone. H 2 S0 4 should then be added and 

 titrated with 0.1 iV thiosulphate, using starch solution as an 

 indicator. The difference between this titration and the titration 

 of a blank on the chemicals represents citric acid (1 cc. of 0.1 N 

 I 2 theoretically represents 3.5 mgms. of citric acid). To the 

 amount of citric acid found by multiplying the number of cc. 

 by 3.5 a correction of 6 per cent should be added if the amount 

 is over 115 mgms., 8 per cent if from 90 to 100 mgms., and 10 per 

 cent if less than 90 mgms. It is desirable to run known amounts 

 of citric acid and calculate the corrections for the particular 

 apparatus and procedure used. 



Where too large amounts of interfering substances are presen t 

 the citric acid must be separated by precipitation. It is not 

 necessary, however, that this separation be complete, but only 

 that the citric acid be precipitated quantitatively and the bulk 

 of the interfering substances removed. Two methods are sug- 

 gested as convenient and practical, the one to be used depending 

 upon the conditions under which the work is carried out. 



Melhod 1 . Precipitation as barium citrate. — To a volume of solu- 

 tion equivalent to approximately 100 mgms. of citric acid a drop of 

 phenolphthalein is added and the solution neutralized with NaOH. 

 Just enough dilute CH,COOH (1 to 2 per cent) is supplied to 

 destroy the pink color and sufficient barium acetate solution to 

 completely precipitate the citric acid, then 2 volumes of 95 per 

 cent alcohol, and the mixture shaken. If a centrifuge provi 

 with large tubes is available the precipitation may be carried out 

 in one of these. The supernatant liquid should be centrifuged 

 and decanted (the precipitate comes down rapidly when centri- 

 fuged) and the residue stirred up at the bottom of the tube with 

 a stream from a wash bottle filled with 26 per cent alcohol and 

 centrifuged again; repeat if necessary. The precipitate should 

 be warmed with 5 cc. of 5 N H 2 S0 4 and a few cc. of water and 

 the whole washed into the Kjeldahl flask and made up to 125 cc. 

 The flask may be boiled vigorously before connecting with the 

 condenser to remove the alcohol, this taking the place of the 

 preliminary distillation. The precipitated BaS0 4 does not inter- 

 fere with the distillation and oxidation. 



led 



