74 PHYSIOLOGY 



of 32CO. In this way the following molecular weights have been arrived at 

 (Abderhalden) : 



Sulphur per cent. Molecular weight. 



Edestin . . 0-87 . . 3680 



Oxyhsemoglobin . . 0-43 . . 7440 



(horse) 



Serum albumin . . . 1-89 .. 1700 



(horse) 



Egg albumin . . . . 1-30 .. 2460 



Globulin 1-38 .. 2320 



The greater part at any rate of the sulphur in the protein molecule occurs 

 as a constituent of a substance, cystine, each molecule of which contains two 

 atoms of sulphur. Each molecule of protein must also contain two atoms of 

 sulphur, and we must regard double the molecular weight given in this Table 

 as the minimum molecular weights of these various proteins. Some idea of 

 the molecular complexity represented by these weights may be gained by 

 writing out the empirical formulae of the various proteins, e.g., 



Egg albumin C 204 H 322 N 52 66 S 2 



Protein in haemoglobin (from horse) . . C 68 oH 10 9 g N 210 02 4 iS2 



Protein in haemoglobin (from dog) . . . C 725 H 1171 N 194 214 S 2 



Crystallised globulin (from pumpkin seeds) . C 202 H 481 N 20 83 S 2 



With some proteins we may make use of other elements to arrive at an 

 idea of the approximate molecular weight. Thus oxyhaemoglobin contains 

 between 0*4 and 0*5 per cent. iron. If we assume that each molecule of 

 oxyhaemoglobin contains one atom of iron, its molecular weight must be 

 from 11,200 to 14,000. 



Attempts have been made to solve the same question by studying the compounds 

 of proteins with inorganic salts or oxides. Thus, the crystals of globulin from pumpkin 

 seeds prepared with magnesia contain 1-4 per cent. MgO. Assuming that one mole- 

 cule of protein has combined with one molecule MgO, the molecular weight of the 

 protein must be about 2800. 



(If x be the molecular weight 



5. = 100 ~ 1 ' 4 

 40 1-4 



. ' . x - 2817) 



Harnack has shown that many proteins are precipitated from their solutions as 

 copper compounds by the addition of copper sulphate. Harnack found that this pre- 

 cipitate of copper contained either 1-34 1-37 Cu. or 2-48 2-73 per cent. Cu. The 

 smaller percentage would correspond to a molecular weight of 4700, while the second 

 number might be accounted for on the hypothesis that each molecule of protein was 

 combined with two atoms of copper. Similar attempts have been made by determining 

 the amount of acid or alkali necessary to keep certain types of protein in solution. 

 We shall see later on however that the amounts vary largely with the physical con- 

 dition and previous history of the colloidal substance. We are dealing here not with 

 compounds in the strict chemical sense of the term, but with adsorption compounds, 

 where the quantities taken up are determined not only by the chemical nature of the 

 protein itself, but by the state of aggregation of its molecules. It is therefore impossible 

 to lay any great stress on the determinations of the molecular weight which have been 

 effected in this way. 



