12 THE VEGETABLE PROTEINS 



in question is a mixture. We can, therefore, for the present, treat as 

 individual proteins only those products whose extensive fractionation 

 has given no evidence that a mixture is being dealt with, and we must 

 await new methods of study before any one of these proteins can be 

 definitely accepted as a true chemical entity. Many, however, of the 

 proteins with which we are at present familiar have shown such con- 

 stancy of composition and properties that we feel justified in now con- 

 sidering them as substances of reasonably definite character. 



In this connection it is worth while to consider the evidence now 

 available which indicates the possibility of separating these proteins 

 from one another, as well as from associated substances, for it has been 

 generally assumed that the difficulties presented in making such sepa- 

 rations are so great that there is little hope for success. Most of the 

 seed proteins now known are globulins which are precipitated by 

 dialysis in the form of crystals or spheroids which are relatively large 

 when compared with the minute particles of an amorphous precipitate. 

 These separate from solution very slowly, and are deposited as dense 

 precipitates which adsorb associated substances to a far less extent than 

 do the amorphous, bulky precipitates of the proteins of animal origin. 

 The best evidence of this is the fact that the globulin of the castor 

 bean, as Osborne, Mendel and Harris (365) have found, could be freed 

 by a single reprecipitation by dialysis from all but minute traces of any 

 toxic substance, though the solution from which it separated on the 

 first dialysis contained a large quantity of extremely active ricin. 

 Ricin, if not a protein, is, at least, so intimately associated with the 

 albumin that it is practically quantitatively precipitated therewith. 

 The conclusion is, therefore, justified that the separation of the castor- 

 bean globulin from the water-soluble albumin by a single reprecipita- 

 tion by dialysis is in a high degree complete. This fact is good 

 ground for considering it possible to separate proteins very completely 

 from one another under such conditions. 



As to the separation of globulins from the non-protein substances 

 with which they are associated in the seed extracts, the fact that many 

 of these globulins can easily be obtained so free from carbohydrates 

 that their preparations give no trace of reaction with the Molisch test, 

 is strong evidence that they can be obtained entirely free from non- 

 protein contaminations. Although such evidence is indirect and does 

 not prove that preparations of these proteins can be obtained in a state 

 of purity and unmixed with other proteins, nevertheless, it strongly in- 

 dicates that these can be obtained in a higher state of purity and of 

 more definite character than is commonly supposed to be possible in 

 preparing proteins of animal origin. 



