Power of Crystalline Ovalbumin and Serum Albumin. 23 



could never be induced to crystallise, in spite of sowing and repeated agita- 

 tion. They remained for two months in this condition, slowly coalescing into 

 a hard gel formation at the bottom of the beaker. They were then redissolved 

 and reprecipitated by (NH 4 ) 2 S04, but again the precipitate consisted of 

 nothing but globuliths of an enormous size, which again settled to a gel. This 

 observation is of importance in the light of two recently published papers. 

 Bradford (1920) claims to have crystallised gelatin, and Oswald (1915) 

 albumin from human ascitic fluid, yet in both cases only globuliths were 

 obtained. In Table VIII I have collected the amount of acetic acid used in 

 crystallisation of several lots of egg white and expressed it in terms of cubic 

 centimetres per 100 c.c. of filtrate : — 



Table VII L 



Lot. 



Fluid Tolume. 



Ph. 



Add volume. 



Product. 





CO. 





c.c. 





1 



500 



4-9 



3 -32 



Crystals. 



2 



275 



5-1 



2 -94 



Crystals. 



3 



600 



4-8 



3-70 



Globuliths. 



4 



650 



5 -0 



8 -56 



Globuliths. 



5 



325 



4-9 



1 -88 



Crystals. 



6 



325 



4-7 



2-95 



Crystals. 



It is thus shown that the globulilh mother liquors required more acid to 

 adjust them to the conditions of precipitation than the normal. "N"ow it has 

 been observed by both Hopkins and Sorensen that this is necessary for eggs 

 that are not strictly fresh. The phenomenon thus suggests to my mind 

 further evidence toward an explanation of the rdle of acid and sulphate in 

 crystallisation. We are dealing essentially with an equilibrium between 

 albumin, water, salt and acid. The function of the salt is one of dehydra- 

 tion, as has been shown by Chick and Martin (1913). The function of the 

 acid is probably twofold. By virtue of the fact that albumin is an ampholyte 

 and on the alkaline side of its isoelectric point, it is bound to form a salt 

 with free acid. At the same time, the acid ions have dehydrating powers. 

 It is thus an adjustment of the available water molecules between the protein 

 or protein salt, on the one hand, and the ammonium sulphate with the free 

 acid ions, on the other. Sorensen (1917) has brought forward evidence from 

 careful quantitative experiments to show that crystalline ovalbumin is a 

 definite hydrate. The above-mentioned observations would be then compre- 

 hensible if, due to slight autolytic changes, the protein had lost some NH 2 

 groupings, and thus at least some of its power of salt formation with acids, 

 yet retained its ability to form a definite hydrate. The globuliths, on this 

 view, would be egg-hydrate, but not potentially crystalline ovalbumin. 



