On the basis of solubility measurements, Sørensen1 postulated that certain soluble proteins, including serum albumin and serum globulin, are not chemical entities but exist as a system of reversibly dissociable components held together by secondary valence forces. Recent investigations, following this lead, have thrown much doubt on the significance of the albumin-globulin ratio as employed in clinical practice. Ultracentrifugal studies2 have indicated that a sensitive equilibrium exists between the molecular species constituting the serum protein complex. Thus, the ratio of small molecules to large molecules in a single sample of serum can be changed by aging, by simple dilution with water or Ringer's solution2a and even by the presence of other protein molecules in the solution.2b Although normal native serum appears to contain three molecular species and additional dispersion states are evident in pathologic serums, these in no way correspond with the albumin and globulin fractions obtained by "salting out."
MELNICK D, FIELD H, PARNALL CG. SIGNIFICANCE OF THE ALBUMIN-GLOBULIN RATIO OF SERUM. Arch Intern Med (Chic). 1940;66(2):295-305. doi:10.1001/archinte.1940.00190140003001