Notes on scientific research. Ill 



low, anyhow the errors can be eliminated to some extent by certain manipulations. 

 The frequent emulsions occurring here are a great drawback. Satisfactory results are 

 obtained by the following chromic acid process: — dilute with sodium sulphate solution, 

 extract with light petroleum, oxidise the alcohol with potassium bichromate and sul- 

 phuric acid to acetic acid, calculate the alcohol present from the amount of bichromate 

 consumed — but only if the portions of water-soluble esters remaining in the sulphate 

 solution are duly taken in consideration, as well as the increase in volume caused 

 by solution of the volatile components. For carrying out this bichromate process, 

 small quantities, say 5 ccm., are sufficient. 



The author finally gives the following improved directions for estimating the 

 alcohol contained in alcoholic preparations together with volatile substances a) by 

 means of the specific gravity, b) by the chromic acid process: — 



a) 50 gr. of the sample, which must be free from glycerin, resins, or other extract substances, are 

 given into a separating funnel, 50 gr. water then added, finally 50 cc. light petroleum ("petroleum benzine") 

 and shaken vigorously several times. The next day, the lower layer is given into a tared flask and brought 

 to 99 gr. The alcohol content as calculated from the density must be multiplied by 2. Special directions 

 are given for removing glycerin, resin, or free acids before carrying ont the estimation, in case these bodies 

 should be present. 



6) 25 or 50 cc. of the sample (according to the preliminary alcohol content as calculated from the 

 density) are given into a tared flask containing about 25 cc. of water, then weighed and diluted up to 500 cc. 

 with a 12.5 per cent, solution of sodium sulphate. The mixture is shaken repeatedly and vigorously, with 

 petroleum benzine, the sulphate solution drawn off when fully clear and shaken with fresh 50 cc. petroleum 

 benzine. The next day 10 cc. of the fully clear sulphate solution are mixed with 25 ccm. 2N-potassium bi- 

 chromate solution and 50 cc. dilute sulphuric acid and cooled down under the tap. After a lapse of two or 

 three hours, the mixture is transferred into a 500 cc. measuring flask and water is added up to the mark. 

 Of this solution, 50 cc. are mixed with 20 cc. potassium iodide solution, then diluted, after 10 to 15 minutes 

 with about 100 cc. water and titrated balk with decinormal sodium hyposulphite solution, using starch solution 

 as indicator. — In order to ascertain whether the sodium sulphate solution used for the oxidation still contains 

 a small amount of water-soluble esters, 200 cc. of the sulphate solution are mixed with 10 cc. standard alkali 

 solution, and the next day the amount of alkali not required for saponifying the esters possibly present is 

 titrated back with standard acid. An amount of 0.2 cc. alkali is not considered. — Each cc. decinormal bi- 

 chromate or hyposulphite solution = 0.0011512 gr. alcohol. — Glycerin, resins, free acids or bases are removed 

 before carrying out the examination, as stated above. 



W. Olczewski 1 ) gives the following directions for estimating the alcohol content 

 in essences: — 25 gr. of the liquid under examination, 25 gr. saturated salt solution, 

 and 20 gr. petroleum ether (b. p. not exceeding 60°) are given into a separating funnel, 

 shaken thoroughly and as soon as the liquid is clear the lower layer is drawn off 

 into a 300 cc. flask. The residual petroleum ether is shaken once more with 15 cc. 

 salt solution, the aqueous layer is given into the flask and the contents, after adding 

 some magnesium carbonate and debris of porous plate, are distilled into a pyknometer 

 of 50 cc. capacity. 



When liquids show a strong tendency to emulsify with petroleum ether, or are 

 rich in essential oil, double distillation is resorted to. From 25 gr. of the preparation 

 and 50 gr. water, with addition of some alkali about 50 cc. are distilled (in case of 

 foaming, the alkali is omitted, and tannin is added) directly into a separating funnel, 

 the distillate is saturated with salt, then shaken out with 25 cc. petroleum ether, drawn 

 off into a 300 cc. flask (the petroleum ether is likewise shaken out once again with 

 salt solution), and the united aqueous liquids are again distilled, as in the first- 

 mentioned process, into a pyknometer of 50 cc. capacity. 



The methods published hitherto for the detection of alcohol in essential oils are 

 discussed by Utz 2 ). We abstain from mentioning them here, since the author likewise 

 has found them to be more or less unreliable. The process described in Gildemeister's 



') Pftarm. Zentralh. 62 (1921), 288. — 2 ) Deutsche Parf. Ztg. 1 (1921), 217. 



