The actual process is as follows:--10 grms. of the sample are weighed out and run into a litre flask, water added, and the mixture shaken. The flask being finally filled up to the neck, some of the solution is now filtered through a dry filter, and 10 c.c. of this filtrate is placed in a 6 or 8-ounce stoppered bottle, and 30 c.c. of the alum solution added. In a similar bottle 10 c.c. of the standard solution of carbolic acid are placed, and a similar quantity of alum solution is added, as in the first bottle. The bromine-water is now run into the bottle containing the standard solution of carbolic acid from a burette until there is no further precipitate; the bottle is stoppered and shaken after every addition. Towards the end of the reaction the precipitate forms but slowly, and when the carbolic acid is saturated, the slight excess of bromine-water gives the solution a pale yellow tint. The solution from the sample is treated in the same way, and from the amount of bromine-water used, the percentage of the sample is obtained by making the usual calculations. Thus, supposing that 5 c.c. of the standard required 15 c.c. of the bromine-water for precipitation, and 10 c.c. of the solution of the sample required 17 c.c., the calculation would be 15 × 2 : 17 = 100 : _x_ per cent. With most samples of crude carbolic acid, the precipitate does not readily separate. It is then best to add a little of the precipitate already obtained by testing the standard solution, which rapidly clears the liquid. =Koppeschaar’s volumetric method= is more exact, but also more elaborate, than the one just described. Caustic normal soda is treated with bromine until permanently yellow, and the excess of bromine is then driven off by boiling. The liquid now contains 5NaBr + NaBrO₃, and on adding this to a solution containing carbolic acid, and a sufficient quantity of hydrochloric acid to combine with the sodium, the following reactions occur:-- (1.) 5NaBr + NaBrO₃ + 6HCl = 6NaCl + 6Br + 3H₂O; and (2.) C₆H₆O + 6Br = C₆H₃Br₃O + 3HBr.