Chemical and Physical Papers. 



41 



10, 10 and 5 cc. of chloroform, added in succession to the aqueous 

 solution. The mixed chloroformic solutions are then evaporated 

 to dryness and the resulting crystals of caffeine, if they contain an 

 appreciable amount of coloring matter, are redissolved in a mixture 

 of petroleum, ether, and chloroform, which gets rid of small quanti- 

 ties of color. This latter solution when filtered is evaporated and 

 the crystals are weighed, and from the weight the percentage of al- 

 kaloid computed. The results that we have thus far obtained show 

 concordant results. Samples of roasted coffee as obtained upon 

 the market have given from 0.825 to 1.66 per cent, of caffeine, and 

 of caffeotannic acid (or plant acids estimated as caffeotannic acid) 

 from 10 to 12.8 per cent. 



We desire to append to the present article the analysis of sam- 

 ples of coffee of different varieties which we have collated from the 

 Proceedings of the American Pharmaceutical Association, volume 

 24, page 144, and, for the benefit of those who are interested in as- 

 say, we would call attention to the tables which accompany the 

 most excellent article on the chemical constituents of coffee pre- 

 sented in Bulletin No. 90 of the Department of Agriculture, pages 

 43 to 45, inclusive. 



Proc. Am. Pharm. Ass'n, vol. 24, p. 144. 



Paul and Cownley proved that a small amount of caffeine is lost 

 in the roasting of coffee. To illustrate this, three samples are 

 given : First, one roasted to extreme degree ; second, one just far 

 enough to give berries a fine chestnut-brown color; third, coffee 

 roasted to pale brown color, when aroma was not fully developed. 



Proc. Am. Pharm. Ass'n, vol. 35, p. 142. 



