94 THE VEGETABLE PROTEINS 



toxin in the bark of Robinia pseudacacia. This substance he considered 

 to be a nucleoprotein and he named it robin. 



Dunbar (99, 100) in 1903 ascribed the effect which extracts of rye 

 pollen had on hay-fever patients to a toxic protein contained in the 

 pollen. In 1904 Kammann described the proteins obtained from rye 

 pollen and stated that one was a toxalbumin which when administered 

 in extremely small doses produced the symptoms of hay fever in 

 patients afflicted with this disease. Prausnitz (382) described in 1905 

 several other toxic proteins related to the hay-fever toxins. 



Osborne, Mendel and Harris (365) in 1905 made an extensive 

 study of the proteins ofRtdnus with special reference to the isolation of 

 ricin. They found that the globulin, which constitutes the greater 

 part of the protein of this seed, and which can be precipitated by dia- 

 lysing the sodium chloride extracts, is entirely free from toxic pro- 

 perties, and that from the solution from which this globulin had been 

 separated by fractional precipitation with ammonium sulphate the 

 ricin was separated between narrow limits of concentration in this salt. 



Most investigators of this subject have been impressed with the 

 idea that the toxic properties belonged in fact to the protein substance. 

 That the protein substance should have such properties is, however, 

 extremely difficult to believe, and doubt has been frequently raised as 

 to the real nature of these toxins. The principal experimental evi- 

 dence which has been employed to solve this question has involved the 

 effect of proteolytic enzymes on the toxic preparations, it being assumed 

 that, if the toxin were a protein, digestion with proteolytic enzymes 

 would destroy its physiological activity. As long ago as 1889, Still- 

 mark (526) showed that in one experiment the toxic power of his ricin 

 was not destroyed by digesting with trypsin for eighteen hours. In 

 1895 Repin (395) stated that abrin was not rendered inactive by pan- 

 creatic digestion. Miiller (280) in 1899 subjected ricin to the action 

 of very active trypsin, and found that after twenty-four hours it was 

 quantitatively as active as at first. He concluded from his experiment 

 that ricin was not a protein, though he stated that the possibility was 

 not excluded that toxic proteoses might have been formed. Jacoby 

 (173) showed in 1901 that ricin was precipitated between one-tenth and 

 six-tenths saturation with ammonium sulphate, and he digested a pre- 

 cipitate obtained under such conditions for five weeks with active pan- 

 creatic juice, and reported that, although the protein reactions had 

 vanished, the precipitate which he obtained from the digested solution, 

 by six-tenths saturation with ammonium sulphate, contained as much 

 toxin as the original preparation of ricin. The experiments of 



