852 METHYL L.EV0-IN0SITOL IN A POISONOUS PLANT, 



mained. The alcoholic fluid was allowed to stand for some days, 

 when there was formed a considerable quantity of clear glassy 

 crystals. On evaporation at 35°C., and again treating with 

 alcohol, a further separation of crystals took place. This evapo- 

 ration and treatment with 95% alcohol was repeated three times, 

 the mass of crystals was drained on a Buchner funnel, and 

 washed with alcohol. The whole was then recrystallised from 

 dilute alcohol, and the crystals dried. The weight of this sub- 

 stance, with the addition of a smaller amount obtained in the sub- 

 sequent treatment of the solution, was 15gms., equivalent to — 



0*65 per cent, of the dried (at 100°) leaves. 



0*60 per cent, of the air-dried material. 

 Purificatio7i of the crystals. — The whole of the substance was 

 now dissolved in water, in which it was exceedingly soluble, and 

 alcohol was carefully added to the point of incipient precipitation. 

 On cooling the solution to 0°C., the substance slowly separated 

 in fine transparent crystals. During this separation, the super- 

 fluid was decanted at intervals, till finally there were obtained 

 twenty separate fractions. The first, tenth, and twentieth frac- 

 tions, when dried, gave melting-points between 188° and 189'5°C. 

 (uncorrected), thus proving the presence of a single substance 

 onl^^ The combined fractions were recrystallised three times, 

 and dried in a desiccator. 



Properties of the Crystals. 



The following tests are described in the order in which they 

 were performed, and show the method by which the constitution 

 of the compound was gradually elucidated. 



Preliminary tests. — (1) On fermentation with a very active 

 preparation of emulsin, no hydrocyanic acid was evolved. The 

 compound is, therefore, not the active principle of the plant. 



(2) When heated, the substance melted, charred, and burned 

 entirely away without residue. 



It consisted of the elements carbon, hydrogen, and oxygen 

 only. The crystals possessed a very sweet taste, and were ex- 

 (;essi\'ely soluble in water; from which facts it may be inferred 

 that the compound contains a nunlbei' of hydroxyl groups. When 



