— ii7 — 



For carrying out the determination there are required: 

 i. Fuch sin sulphurous acid: 0,5 g. fuchsin (magenta) are dissolved 

 in ioocc. water, and a solution of sulphurous acid added which contains 

 16 g. S0 2 . This is left standing until discolored, and then made up 

 to 1 litre. The solution must be renewed every 2 or 3 days, as it 

 deteriorates very rapidly. 



2. Alcohol of 95 per cent, by volume, absolutely free 

 from aldehyde. In order to free the alcohol from aldehyde down 

 to the smallest trace, it is left standing for several days over alkali, 

 then distilled off, and the distillate boiled for several hours in a reflux 

 condenser with m-phenylenediamine hydrochloride, 25 g. per litre; 

 the alcohol is next again distilled off, and is then ready for use. 



3. A standard citral solution, i. e., a 0,1 per cent, solution 

 of citral in 50 per cent, alcohol free from aldehyde. 



All these, and also the following solutions, are used exactly at 15 , 

 and all determinations are also carried out at that temperature, by 

 placing the vessels, before making the comparative tests, for 10 minutes 

 in a water bath of 15 °. A higher temperature must be particularly 

 avoided. 



The procedure now is as follows: 2 g. of each of the lemon oils 

 to be examined, or 20 to 30 g. of the lemon extracts, are diluted 

 with alcohol free from aldehyde to 100 cc. Of each of these solutions, 

 4 cc. is placed in vessels of exactly equal capacity; 20 cc. alcohol 

 free from aldehyde are added, and next 20 cc. fuchsin sulphurous 

 acid, and finally alcohol is added up to 50 cc. and the whole well 

 mixed. By means of the above-mentioned standard solution, com- 

 parative samples of known citral-content are prepared in the same 

 manner. All solutions are left for 10 minutes in the water bath, and 

 they are then compared, either direct, or by means of a colorimeter. 



Chace proved by tests with mixtures of known citral-content that 

 this method gives very exact results, both with mixtures of citral and 

 limonene, and with lemon extracts, in which the values found differed 

 on the average by not more than 0,2 °/ from the actual content of 

 citral. With lemon oils the results were less favourable, but in spite 

 of this the limit of error was here also mostly below 0,5 °/ . 



In testing this method, where the only object we had in view was 

 to ascertain its usefulness for lemon oils, we did not employ a color- 

 imeter, but determined the shades of colour by direct comparison. 

 We proceeded as follows: by means of the normal solution we pre- 

 pared in the above -described manner with fuchsin sulphurous acid 

 and alcohol, eight different samples, of which each following sample 

 contained 1 cc. solution = 1 mg. citral more than the previous one; 

 the first vessel contained 1 cc. normal solution = 1 mg. citral; the eighth 

 8 cc. solution = 8 mg. citral. We then found that the individual 



