J. Rakshit modifies Messinger's method by substituting lime water or baryta water for caustic potash solution, with the view of diminishing the error due to formation of iodoform from the ethyl alcohol.1

The sample to be examined, containing about 0 05 gram of acetone, is placed in a 750 c.c. flask, and 300 c.c. of freshly-prepared lime-water added; the flask is loosely closed with a rubber cork, and heated to about 35°. Drop by drop 5 c.c. of N/5-iodine solution are added, and the mixture shaken for five minutes, then another 5 c.c. of iodine are similarly added with shaking, and so on until 40 c.c. of iodine have been run in. If during the addition the colour of the iodine persists after thorough shaking, more lime-water should be added. Ten minutes after the final addition of iodine, a few drops of starch solution are added, the contents of the flask shaken and cooled, 15 c.c. of N/1 -sulphuric acid run in, and the excess of iodine titrated with iV/10-thiosulphate. The number of c.c. of N/5-iodine used up, multiplied by 000193, gives the quantity of acetone present.

Each c.c. of ethyl alcohol present was found to use up 0.8 c.c. of iV/5-iodine, and this correction should be applied. Accurate results are given with 1 part of acetone to 10 of alcohol; with 1 to 100 parts of alcohol the results are less satisfactory, as the correction is then relatively great.