96 THE VEGETABLE PROTEINS 



some pronounced manner from other known proteins. All the other 

 facts which we know in regard to these toxalbumins speak strongly in 

 favour of their 'protein nature, and in view of the experiments of Osborne, 

 Mendel and Harris (365) it is difficult to believe that they should not 

 be classed with the other protein substances. 



These experiments were directed to a separation of all the different 

 proteins of the seed of Ricinus, and resulted in showing that the toxic 

 properties are associated only with those proteins which are soluble in 

 water and coagulable by heat. By dialysing the sodium chloride ex- 

 tract of the seed the greater part of the protein separates as a globulin 

 which is free from toxic properties. The aqueous solution from which 

 this globulin separates, when fractionally precipitated with ammonium 

 sulphate, yields the toxic substance sharply concentrated in fractions 

 obtained within comparatively narrow limits of concentration in this 

 salt. The toxicity of these fractions is closely proportional to the 

 amount of coagulable albumin which they contain, and, by further 

 fractionation with ammonium sulphate, preparations result which con- 

 sist largely of albumin. The most toxic of these, as yet obtained, and 

 at the same time the richest in albumin, consisted of a mixture of over 

 70 per cent, of albumin with nearly 30 per cent, of proteose. The tox- 

 icity of this preparation was far in excess of any preparation of ricin 

 before described, for '0005 of a milligramme per kilo of body weight was 

 fatal to rabbits when subcutaneously injected. An examination of the 

 properties, ultimate composition, partition of nitrogen and colour re- 

 actions showed this preparation to have in all respects the properties 

 characteristic of pure protein, and no evidence whatever was obtained 

 which would indicate that the preparation contained more than the 

 most insignificant quantity of any other substance. The conclusion 

 appears, therefore, to be justified that the toxic property belongs to the 

 protein, and in view of the close relation found for the degree of toxicity 

 to the proportion of coagulable albumin contained in various fractions 

 of the Ricinus protein, it seems almost certain that this latter protein 

 has the toxic power and that true toxalbumins occur in seeds. Similar 

 studies of the other toxalbumins have not been made, and therefore 

 definite conclusions in regard to the actual existence of a toxalbumin 

 in the seeds from which these are obtained are at present not possible, 

 but the facts available strongly indicate that these seeds also contain 

 a toxalbumin similar to ricin. 



