Power of Crystalline Ovalbumin and Serum Albumin. 29 



Experiment 2. — A 15 per cent, solution was prepared and the globulin 

 removed as previously. Acetic acid was added until the Ph was adjusted 

 to 4'8. Crystals began to form in 2 minutes and by 5 minutes there was a 

 heavy precipitate, consisting entirely of crystalline material. The amount of 

 albumin which was obtained in crystalline form by this procedure was 

 estimated at 70*4 per cent. The crystals were redissolved in water and 

 (NH 4 ) 2 S04 solution added with great care at a very slow rate to a point 

 somewhat short of a permanent precipitate. On standing, the solution 

 deposited crystals in about an hour, and the yield was increased by slow 

 addition of (NH^aSO*. The crystals were again brought into solution and 

 the liquid still showed a slight yellow colour. The rotatory power was 

 observed and the concentration of albumin determined by the Devoto 

 method, as previously described. The albumin was again recrystallised and 

 the determinations repeated. Table X summarises the results, and it is thus 

 evident that the specific rotation is constant after the first crystallisation : — 



Table X. 



Crystallisation. 



Protein 

 concentration. 



«D. 





a-E. 



He. 18 



First 



per cent. 

 0-98 



2 -18 



3 36 

 1 -42 



o 



-1-13 

 -2-74 

 -4-29 

 -1 -78 



-58 -06 

 -62 -94 

 -62 -84 

 -62 -70 



— 1 '36 

 -3-41 

 -5-26 

 -2-22 



-69-26 

 -78 -20 

 -78 -32 

 -78-25 





Third 





Ayerage 







-62 -83 





-78 '26 





Method 2. — The object of this experiment was to discover whether it was 

 possible to obtain an albumin giving the same specific rotation, but prepared 

 by a different method. The second method of obtaining serum albumin in a 

 pure state was evolved from attempts to obtain crystals from serum by the 

 exact application of the method of Hopkins for ovalbumin. This method has 

 been used by several investigators as a means of obtaining relatively pure 

 serum albumin for various purposes, but the procedure has never been care- 

 fully studied, and it was still an open question whether an individual albumin 

 could be obtained from serum. Furthermore, the fundamentally important 

 relations of serum albumin to the lipoids, cholesterol, and cholesterol esters 

 of the blood stream seemed possible of investigation by a contrast of the 

 serum albumin obtained directly and by the isolation method previously 

 described. 



The method was constructed on the basis of a series of preliminary 



