130 JOURNAL OF THE 



standing and washed with alcohol. The alcohol was then dis- 

 tilled off from the filtrate, the watery residue was diluted with 

 water and filtered. Petroleum-ether, benzol and chloroform 

 were successively used to extract the alkaloidal principles, if any 

 were present in the acid liquid. Then, after rendering alkaline 

 with ammonia, the liquid was again extracted with the solvents 

 mentioned. 



As, even with water hut slightly acidified with sulphuric acid, 

 some risk of the destruction or change of the alkaloids was run 

 during the long evaporation, a second method was made use of 

 as follows: 



The leaves were digested for ten hours with 70 per cent, alco- 

 hol, the alcohol distilled off and the residue treated with lead 

 acetate and soda. The excess of lead was removed by means of 

 sulphuretted hydrogen and the filtrate from this evaporated to a 

 thin syrup. This was then treated with strong alcohol, filtered, 

 and the excess of alcohol distilled off. Bismuth-potassium 

 iodide and sulphuric acid was next used to precipitate any alka- 

 loid present. The presence of albuminoid matter rendered it 

 necessary to decompose this by means of soda, neutralize with 

 dilute sulphuric acid, and re-precipitate with mercuric chloride. 

 The solutions to which mercuric chloride had been added were 

 allowed to stand several days. The results may be tabulated as 

 follows : 



I. opaca, leaves No alkaloid. 



I. opaca, berries 



I. dahoon, leaves 



I. dahoon, berries , 



I. cassine, leaves Caffeine. 



I. cassine, berries No alkaloid. 



I regard these analyses as conclusive, at least, of the absence 

 of the known, well characterized alkaloids. It is, of course, 

 possible that other methods might reveal the presence of some of 

 the more elusive ones. 



a a 

 a u 



