APPENDIX. 121 



matter, lilready mentioned, separated on the sides of the bottle. The 



agitation ^th basic lead was continued for a considerable time, until 



colouring matter ceased to he dissolved. The amylic alcohol was 



then allowed to stand for some days, fi]tei'edj and evapoi'ated on a 



water-bath. The extract was ues:t taken up with water, and lead 



removed by H^S. After filtration the solution was again evaporated 



to dryness, the extract reduced to powder, and repeatedly agitated 



and digested with ether. The saponin extracted ia this manner had 



a faintly yellowish colour^ and contained 6 per cent, of ash free from 



lead. After drying over sulphuric acid in vacuo, the following results 



were obtained on ultimate analysis: 



Exp. J. E5:p. 2. Mean, 



Carbon 60-12 60-21 QQ^Uo 



■ • « 



Hydrogen 8'35 8-54 8^445 



Oxygen ... ... 31-53 31'25 31-390 



100-00 lOO'OO 100-000 



Some of the saponin used for the last analysis was subjected to 

 a further process of purification. It was dissolved in amylic alcohol, 

 and the solution repeatedly agitated with aqueous barium hydrate- 

 On evaporating the amylic alcohol solution to dryness, and heating 

 the powdered extract with ether to separate traces of a?nylic alcohol, 

 the saponin was left as a white powder which contained '308 per 

 cent, of ash. On ultimate analysis, the following pei-centages vrere 

 obtained, the saponin being dried m vacuo over sulphuric acid: 



Ex^. 1. Es]). 2. Mean. 



Carbon 59-90 59-82 59-86 



Hydrogen 8-64 8-62 8-63 



Oxjgen 31-46 31-5G 31-51 



liO-CO lOQ-00 100-00 



It seems likely to us that the last sample of saponin isolated was the 

 purest of the three examined, though we are not prepared to definitely 

 assert it was a pure saponin. We have adduced some evidence 

 which tends to indicate that at least two saponins exist in false 

 Bikhma, and it is possible that the method we used for separation 



