1 84 



Journal of Agricultural Research 



Vol. Xvn, No. S 



Tables I and II show that 95 per cent ethyl alcohol is the only good 

 economic solvent used and that heat has no effect on the extract obtained. 

 It was decided, therefore, to make quantitative extractions of several 

 species of Derris by using hot denatured alcohol, since this solvent is 

 comparatively cheap. 



Table III. — Quantitative extractions of various species of Derris made with hot denatured 



alcohol 



Name of species. 



Part of plant used. 



Percentage 

 of extract 

 obtained. 



Derris sp. (probably D. elliptica) 



D. elliptica, called "tuba" 



D. uliginosa 



D. koolgihberah 



D. scandens 



D. oligosperma 



D. rohusta 



D. robusta 



Roots. 

 Roots. 

 Stems 

 Stems 

 Roots . 

 Stems 

 Roots. 

 Stems 



14.25 

 8.50 

 8.50 

 10.30 

 20. 30 

 22. 50 

 16. 70 

 15-70 



The foregoing table shows that denatured alcohol is a good solvent and 

 that the percentages of extract obtained vary considerably ; this variation 

 is certainly due in part to the fact that the eight powders used varied con- 

 siderably in fineness. Results showing the effectiveness of these extracts 

 are discussed on page 188 and in Table V. 



EXTRACTION OF" TOXIC PRINCIPLE; FROM DERRIS SP. BY TWO METHODS 



As already stated, Greshoff (j), van Sillevoldt (9), and Powers {8) have 

 agreed that the toxic principle in Derris elliptica and. D uliginosa is a resin 

 and have called the active portion of it " derrid." In the present investi- 

 gation it was considered expedient to isolate a small quantity of the resin 

 and to test it on insects and on a few higher animals. 



VAN SILLEVOLDT 'S METHOD 



One kilo of the powdered root of Derris sp. was repeatedly extracted 

 wdth boiling water until the extract was only slightly colored. After the 

 powder had been filtered and thoroughly dried it was boiled under a reflux 

 condenser with successive portions of 95 per cent alcohol until exhausted. 

 The combined alcoholic extracts were mixed with one-fourth their volume, 

 of water, and the alcohol was distilled under reduced pressure. As the 

 alcohol was removed, the material in the flask became milky in appearance 

 and the resinous substance collected in a mass on the bottom of the flask 

 The last portion of the water was removed by transferring the material to 

 an open dish on a steam bath. The residue was a resinous, sticky mass 

 which weighed 1 10 gm., representing 1 1 per cent of the dry root. It was 



