Th.e benzol extraction of the 20-gram sample of ground Tecoma mollis 

 leaves was made in a continuous extraction apparatus which allowed 

 the condensed vapors to percolate through a layer of the plant material 

 about three inches thick. About 200 cc of benzol were used and the 

 extraction continued four days. On evaporating the benzol an almost 

 black residue was obtained. The portion of this residue recorded as 

 soluble in water and in dilute acid was amorphous and gave no indica- 

 tions of alkaloids or glucosides. It resembled very much the resin 

 found in the alcohol-soluble portion of this residue. 



The statements given in the usual outlines of the Parsons method 

 found in text books are not clear as to just how the alcoholic extraction 

 of the plant should be accomplished. It is not certain whether the 

 required 12 to 14 hours' continuous extraction, as stated, is to be per- 

 formed in the cold by percolation, or in a closed flask using a definite 

 ratio of plant material to solvent, or by some other convenient method. 

 In the present case it was decided to use the same continuous extrac- 

 tion apparatus as employed for the benzol extraction, placing 80 per cent 

 alcohol in the flask of the apparatus. It was of course recognized that 

 by this plan the extraction would actually be made with alcohol of a 

 concentration higher than 80 per cent. In order to determine approxi- 

 mately just what strength of alcohol this would be, a blank experiment 

 was made, using 83 per cent alcohol in the flask, and a test tube in the 

 position occupied by the sample. In this way it was found that the 

 condensed alcohol collected in the test tube had a specific gravity of 

 0.8226 and thus corresponded to 91 per cent alcohol. Therefore by 

 using 80 per cent alcohol in the flask the plant was actually extracted 

 with approximately 90 per cent alcohol. 



The examination of the residue obtained by 'evaporating the alcoholic 

 extract was made by treating this residue with absolute alcohol, thus 

 separating it into two portions, each of which was further separated by 

 treatment with water into a soluble and an insoluble portion. This pro- 

 cedure proved rather unsatisfactory, since the absolute alcohol separa- 

 tion yielded two portions which appeared almost identical in all respects 

 except their solubility in absolute alcohol. The time required, there- 

 fore, to make the separation was almost doubled, without a correspond- 

 ing improvement in the results. 



COMPARISON OF THE FIGURES OBTAINED BY THE TWO METHODS OF ANALYSIS. 



A comparison of the figures obtained by the two methods of analysis 

 shows that the quantity of material extracted by benzol in the Parsons 

 method corresponds in a measure to the amount removed by both the 

 petroleum ether and ethylic ether extractions in the Dragendorff method. 

 In the latter case, however, the two solvents give a separation which is 

 clearly advantageous from an analytical standpoint. It will be observed 

 that the alcoholic extract of the Parsons method yields a result which 



