THE PROTEIN SUBSTANCES. 65 



sition of the complex albuminous molecule is extremely diffi- 

 cult, if not impossible. The formulas which have been pub- 

 lished are interesting chiefly in showing roughly how complex 

 the structures certainly are. For serum albumin Hofmeister 

 has given this minimum formula : 



CoHnoNu.S.Oi40, 



while for egg albumin he has given this: 



C*.H.N..S,O. 



Even more complex formulas are given, for example this : 



These formulas are in a measure based on an assumption as 

 to the number of sulphur atoms present, about which some- 

 thing will be said later. 



The usual methods of fixing organic formulas by aid of a 

 molecular weight determination can not be successfully applied 

 in these cases. In the cryoscopic method, for example, the 

 traces of mineral impurities present have possibly as much 

 influence on the result as the whole weight of dry protein. 

 Because of changes in composition at a high temperature the 

 boiling point method, even if otherwise reliable, can not be 

 applied, and methods based on osmotic pressure observations 

 lead to results of no value. 



CLASSIFICATION OF THE PROTEIN BODIES. 



The substances thus far studied have been divided into 

 groups or classes dependent on chemical composition or struc- 

 ture. With the protein compounds this is only partially 

 possible because of our lack of full knowledge in this direction. 

 Of the structural relations of the protein molecules nothing 

 whatever is known, while of composition only a few general 

 facts are clearly enough established to be available in a scheme 

 of classification. The first efforts at classification, which 

 we owe largely to the work of Hoppe-Seyler, were therefore 

 essentially empirical. Other schemes were later proposed as 

 6 



