30 ANTHONY McGILL ON 



by Blyth (p. 351) for guidance of analysts as to the percentage of chicory present in a 

 mixtvire of cotfee and chicoiy, from the percentage of soluble ash. In the table referred 

 to, 3 per cent, of soluble ash is supposed to indicate entire absence of chicory ; whereas, I 

 have found mixtures (Table V) which gave more than 8 per cent, of soluble ash, and yet, 

 undoubtedly, contain more than 50 per cent, of chicory. 



I am of opinion that the silica — by which I mean that portion of the insoluble ash 

 which remains after treatment with hydrochloric acid — found in commercial ground 

 coffees, is a very unsafe guide to the amount of chicory present. While the absence of 

 any marked amount of silica pn-ves the absence of chicory, the presence of silica does not 

 justifiy the assumption that chicory is present. (Nos. 1, 2 and 5 in Table V.) 



Fat. — 1 find the thorough drying and pulverisation of the sample a very important 

 consideration in the estimation of the fat. My mode of proceeding is as follows : — 



The equivalent of 5 grammes of the dry substance is weighed out and heated for half 

 an hour in an air-bath at 120° C. It is then thoroughly pulverised in a stone-ware mortar 

 and placed upon a filter which is loosely folded over it, and, so enclosed, is introduced 

 into a Soxhlet extraction apparatus, lined with glass wool, which should be closely 

 packed into the opening of the syphon to act as a strainer and prevent any particles of 

 coffee from being carried over with the liquid. 



The solvent which I use is petroleum of B. P. 60° to *70° C. I find this to extract a 

 much clearer fat than ether. The extraction is allowed to proceed for three hours, when 

 the fat is recovered by evaporation of the petroleum. 



Mattees Oxidisible by Permanganic Acid. — I have used indigo as an indicator, 

 and the process is essentially Lowenthal's well-known one; but I have not used it 

 primarily for the estimation of tannin. 



I hoped that the estimation of oxidisible matters, without precipitation of the tannin, 

 might furnish a clue to the adulteration of coffee ; but I have obtained no result worthy 

 of being noticed here. I hope yet to investigate this subject more fully. 



Caffeine. — 50 cc. of the 10 per cent, extract, corresponding to 5 grammes of dry 

 coffee, is clarified by precipitation with basic lead acetate, and the filtrate freed from 

 lead by the addition of a strong solution of sulphur dioxide in water. 



The filtrate from lead sulphite is evaporated to dryness on the water bath, with 

 calcined magnesia. I find it convenient to transfer the residue, while still slightly moist, 

 to a filter placed upon a large watch glass. The dish is wiped perfectly clean with 

 pieces of filter paper, and the whole thoroughly dried on the watch glass in the water 

 oven. The dry filter, with its contents, is placed in a Soxhlet tube lined with glass 

 wool, and exhausted with chloroform for three hours. The chloroformic extract is nearly 

 pure caffeine. 



I have found a markedly larger yield, on extracting with a mixture of one part 

 chloroform and three parts ether. This I attribute partly to the greater number of siphon- 

 ings in a given time, owing to the lower boiling point of the mixed liquids used in 

 extraction. 



Sugar. — The exact estimation of the amount of sugar present in commercial coflees 



