40 Kansas Academy of Science. 



Dr. Wilhelm Autenrieth, in his work on the detection and esti- 

 mation of alkaloidal substances, page 185, gives a r6sum6 of the 

 literature on the estimation of caffeine of caffeine-bearing plants, 

 and gives six different methods. We have tried these various 

 methods, and have found the Katz method to be the most satis- 

 factory. This method is as follows : 



Shake 10 grams of powdered coffee or tea for thirty minutes with 

 200 grams of chloroform and 5 grams of ammonium hydroxid solu- 

 tion. When the liquid has settled filter 150 grams of the chloro- 

 form solution, which should be perfectly clear and free from water. 

 Distil the chloroform and dissolve the residue with gentle heat in 

 about 6 cc. of ether. Add 20 cc. of 0.5 per cent, of hydrochloric 

 acid and, in an assay of coffee, also 0.2 to 0.5 grams of solid paraffin. 

 Evaporate the ether and filter the cold, aqueous solution. Wash 

 the flask and filter-paper a few times with small portions of 0.5 

 per cent, hydrochloric acid. Finally extract the total aqueous hydro- 

 chloric acid solution four times with 20 cc. portions of chloroform. 

 Distil the filtered chloroform extracts, dry the residue, and weigh. 

 This residue will consist of nearly pure caffeine. 



The one difficulty in all of these processes is that the caffeine 

 recovered in the process contains some coloring matter or some 

 empyreumatic matter imparting color to the product. It has been 

 suggested that when the yield is thus contaminated the crystals be 

 subjected to the Kjeldahl's nitrogen determination. The quantity 

 of anhydrous caffeine is calculated on the basis of this analysis. 

 One cc. of ^ oxalic acid represents 0.00485 grams of anhydrous 

 caffeine. Tetrachlormethane has been suggested by some analysts 

 as a proper substitute for chloroform in the extraction of the alka- 

 loid. We have found that an almost colorless caffeine can be ob- 

 tained by properly using this solvent. It has been our endeavor 

 to obtain a simple process by which we could estimate the caffeo- 

 tannic acid and the caffeine at the same time. We have thus 

 far had quite accceptable results by using a modification of 

 the Krug method. Instead of using 2 gm. of the coffee, 7 gm. of 

 the finely powdered coffee are taken. After digestion with the 

 water and alcohol (as required by the formula) for twenty-four 

 hours, 100 cc. of the filtrate, which is made to represent 5 gm. of 

 the coffee, are taken. This is precipitated by the solution of lead 

 acetate, and 60 cc. of the filtrate, representing 3 gm. of the coffee, 

 are taken and treated with an excess of diluted sulfuric acid. The 

 filtrate from the lead sulfate, which has been thoroughly washed, 

 is carefully evaporated to remove traces of alcohol, and the residue 

 is filtered into a separatory funnel and thoroughly washed with 15, 



