32 The Philippine Journal of Science 



any guiding information concerning the rather feeble < 

 cial preparations, it is probable that quantities have been used 

 which fall considerably short of both the tolerated and the 

 effective doses. In our work we have prepared some concen- 

 trated and highly toxic extracts of two of these plants. No 

 attempt has been made to determine chemically the content 

 of the active principle. The fatal dose for rabbits was estimated 

 in order to secure a general guide for commencing the adminis- 

 tration in man. 



Method of preparation. — The bitter principle of these plants 

 is soluble in either alcohol or water. Elimination of the greater 

 portions of the gums, resins, chlorophyll, and starch can be 

 readily effected by extracting first with alcohol, evaporating 

 almost to dryness, and then extracting the residue with water. 

 For Castela nicholsoni the smaller twigs were ground to a fine 

 powder. The bitter principle of this was extracted by boiling 

 for several hours with methyl alcohol, in the proportion of 1 

 kilogram of the powder to 5 liters of alcohol. One extraction 

 removed the bitter principle almost completely. After filtration 

 the alcohol was evaporated at low temperature, and a tarry 

 residue was left behind. A quantity of residue representing 

 12 kilograms of the powdered plant was extracted with 200 

 cubic centimeters of water divided into small portions. This 

 extract was found to be very toxic for rabbits, 0.1 cubic centi- 

 meter injected subcutaneously producing death overnight. For 

 convenience, this quantity of 200 cubic centimeters was diluted 

 to 1 liter. Injected subcutaneously, 1 cubic centimeter of this 

 solution per kilogram killed rabbits within twenty-four hours; 

 0.5 cubic centimeter produced no symptoms. 



The extracts of Harrisonia perforata and Brucea amarissima 

 were very kindly prepared by Dr. H. I. Cole, of the division of 

 organic chemistry of the Bureau of Science. The extraction of 

 these two drugs was carried out more thoroughly than was done 

 in the case of Castela; the toxicity of these extracts represents, 

 roughly but not accurately, the relative toxicity of the corre- 

 sponding plants. Ethyl alcohol was substituted for methyl. 

 The Harrisonia perforata material was shade dried, and the 

 branches and leaves were ground to a coarse powder. Only the 

 seeds of Brucea amarissima were used. A rather dilute prep- 

 aration of Harrisonia was employed, of which 2 cubic centi- 

 meters represented 1 gram of the original plant. Rabbits were 

 killed in from twenty-four to forty-eight hours by the sub- 

 cutaneous injection of 15 cubic centimeters of this solution, but 



