This method is not used in this country, but is official in France, particularly as a sorting.out procedure preliminary to the application of the longer and more exacting Bardy method, described later on. It depends upon the increase of volume produced in a given quantity of chloroform, when the alcohol in question is shaken with the chloroform under strictly defined conditions. Calvet1 gives the following modification as used in France: the acetates, the volume of which is measured. The propyl alcohol is determined by oxidation with potassium permanganate under prescribed conditions.

Such a quantity of the sample is taken as will contain exactly 100 c.c. of absolute alcohol, and this is diluted to exactly 400 c.c. at 15°. Fifty c.c. of chloroform at 15° are measured into a "Rose tube." (This is a vessel of 700 c.c. capacity, the upper part of which resembles a pear-shaped separator, and the lower part is a bulb of 49 c.c. capacity, joined to the upper part by a narrow, graduated stem.) The diluted alcohol is added to the chloroform, followed by 2 c.c. of sulphuric acid (D = 1.286). After being placed in a water-bath kept at 15° for half an hour, the tube is shaken vigorously for five minutes in a mechanical shaker, and again placed in the bath. When the chloroform has all collected together its volume is read off on the graduated stem, and the increase noted.

Under these conditions, pure alcohol gives an increase of 0.45 to 0 50 c.c. in the chloroform layer. With 1 per cent. of amyl alcohol the increase is 19 c.c., or a net value of 14 c.c. on deducting 0 5 for the "blank" experiment. One per cent. of isobutyl alcohol gives a net increase of 06 c.c., and 1 per cent. of propyl alcohol gives 025 c.c. net increase; these homologues are counted as amyl alcohol unless it is known that one or other of them is the chief alcohol present.

Bardy's method.2 - The principle of this process is the extraction of the amyl and butyl alcohols with carbon disulphide from the sample (saturated with salt), their removal with sulphuric acid from the carbon disulphide, and their ultimate conversion into

1 "Alcohols," p. 97. 2 Compt. rend., 1892, 114, 1201.

(1)

Mix 100 c.c. of the alcohol in a stoppered separator with 450 c.c. of saturated salt solution, and about 50 c.c. of water to prevent the salt crystallising out. Add 60-70 c.c. of pure carbon disulphide.

(2)

Shake vigorously for 1 minute; loosen the stopper a moment to equalise the pressure, then shake again for five minutes.

(3)

Run off the carbon disulphide into a 300 c.c. separator, taking care to avoid the entrance of any of the aqueous liquid.

(4)

Make two further extractions in the same way, adding the carbon disulphide extracts to that in the 300 -c.c. separator. All the amyl and butyl alcohols are thus removed. Retain the saline liquid for the propyl alcohol determination (see below).

(5)

Shake the mixed carbon disulphide extracts with enough sulphuric acid (D = 1.45) to allow of the acid collecting at the bottom of the separator after shaking (about 2-3 c.c. suffice). Twirl the separator a little to assist settling, and let stand till the acid has separated.

(6)

Draw off the acid into a 125 c.c. flask. Wash the carbon disulphide twice with 1 c.c. of sulphuric acid each time, and add the washings to the first quantity of acid in the flask.

(7)

Expel any carbon disulphide from the acid extract by passing a current of air on to the surface, warming, if necessary, to about 60°.

(8)

Now add enough crystallised sodium acetate to decompose nearly all the sulphuric acid (1. grams for each 1 c.c. of acid will suffice), fit the flask with an air-condenser 1 metre long, and heat the flask on the water-bath during half an hour.

(9)

After cooling, add to the contents of the flask 100 c.c. of salt solution, and then remove the whole to a separator of 300 c.c. capacity, the lower part of which is drawn out and graduated into tenths of a c.c.

(10)

Allow the liquid to rest for some hours, then draw off the salt solution so as to bring the amyl and butyl acetates into the graduated stem. Place the instrument in water at 15°, and when the temperature is constant read off the volume of the acetates. The number of c.c. multiplied by 0 8, gives the quantity of amyl and butyl alcohols (together) in 100 c.c. of the spirit taken.

(11)

To determine the propyl alcohol, take the extracted liquid retained at (4), and filter it through moistened paper to remove any carbon disulphide. Then distil it until the strength of the distillate shows 50 per cent.of alcohol (at 15°). At this point all the alcohols have passed over.

Fill a burette with the distillate, and run out the alcohol drop by drop into a beaker containing 50 c.c. of water and 10 c.c. of KMnO4 solution (1.0 gram per litre) until a copper-red tinge is obtained similar to a standard tint. The latter is given by 20 c.c. of fuchsine solution (01 gram per litre) mixed with 30 c.c. of K2CrO4 solution (0 5 gram per litre), and made up to 150 c.c. with water.

Under these conditions, the standard tint is given by approximately 2.5 c.c. of ethyl alcohol at 50 per cent. strength containing 1 per cent. of propyl alcohol; whence the percentage of propyl alcohol in the sample is deduced.

If the sample contains so much fusel oil as to give a distinct layer of supernatant oil when 100 c.c. are shaken with 500 c.c. of salt solution, the volume of this layer is read off in a suitable vessel, and the remainder of the liquid is then extracted with carbon disulphide as described above.