BY JAMES iM. PETKIK. 791 



of an exceedingly active and poisonous alkaloid. In order to 

 obtain an amount sufficient for further investigation, 1000 grams 

 of leaves were exhausted with boiling water faintly acidulated; 

 the solution was evaporated to dryness, and its residue extracted 

 with alcohol in a Soxhlet extractor. The alcoholic solution con- 

 taining all the alkaloid was then distilled in vacuo, the residue 

 dissolved in water, and the solution filtered. This clear aqueous 

 solution was acidified with H2SO4, and shaken out, first with 

 amyl alcohol, then with ether, till nothing further was removed. 

 The residual aqueous solution was then rendered alkaline and 

 shaken out with ether. On evaporation of the solvent 865 

 gram of impure alkaloid was obtained. 



Purification. — The crude material was dissolved in slightly 

 acid water and shaken with ether in alternately alkaline and 

 acid solution; the ethereal layer was separated and the shaking 

 out repeated till no further residue was observed on evaporating 

 the solvent. The total ethereal extract now yielded 0*39 gram 

 of alkaloid. The aqueous solution was further shaken out with 

 chloroform, and from this, after removal of the solvent, was 

 obtained a residue of 0'188 gram. This residue w^as found later 

 to be non-alkaloidal. 



It was noticed, when the solutions were warmed to volatilise 

 the solvent, that a strong odour was evolved resembling that of 

 nicotine or conine, and suggesting the admixture of a volatile 

 alkaloid. 



To test this point the ether residue (0-39 gram) was dissolved 

 in acidulated water, then made faintly alkaline and distilled in 

 a current of steam. The distillate was strongly alkaline; but 

 when neutralised and evaporated to small bulk it gave negative 

 results in every case on testing with all the ordinary alkaloid 

 reagents. The same process was likewise gone through with the 

 chloroform residue (0*188 gram) and again the distillate was 

 found to be non-alkaloidal. 



From this it may be definitely inferred that no volatile alka- 

 loid is present, and that the distillate is strongly alkaline by 

 ammonia formed probably from the amidenitrogen in the plant. 

 63 



