6 



agrees fairly well with the absolute alcohol and water extract together of 

 the Dragendorff method. Here again the use of the two solvents has 

 resulted advantageously, in that a very clear separation of the tannin 

 on the one hand and the glucoses, extractives, etc., on the other, has 

 been effected. 



With regard to the remaining steps of the two analyses, the conditions 

 vary materially in the two cases and the wide differences obtained are 

 naturally to be expected. 



CONCLUSIONS. 



The results of the analytical work described in the preceding pages 

 show that Tecoma mollis contains no alkaloid or other well character- 

 ized plant constituent of medicinal importance. It is probable that the 

 virtues attributed to it are associated in some way with the persistent 

 bitter taste which the plant is found to possess. The bitter ingredient 

 appears principally in the alcoholic and aqueous extracts. Plants such 

 as gentian, taraxacum, quassia, etc., possessing a similar bitter taste are 

 recognized as having a stimulating and tonic effect on the human sys- 

 tem and it is probable that the physiological action of Tecoma mollis 

 may be of a similar character. 



In regard to the advantages possessed by the two methods of analy- 

 sis, it appears that in the case of Tecoma mollis at least the Dragendorff 

 method is to be preferred. With this method the solvents employed 

 effect a better separation of the constituents than do those used in the 

 Parsons method. 



O 



