The results of this investigation show that a gelatin solution consists of a considerable number of constituents. At a particular temperature, certain gelatin constituents tend to aggregate and to flocculate from solution. When these particular gelatin constituents have completely flocculated, no further change occurs in the system and an apparent equilibrium exists. This is not a dynamic equilibrium between the gelatin flocculate as a whole and the gelatin remaining in the solution but a steady state determined for that system by the temperature.
It is also shown that gelatin can be separated into fractions in which the gelatin constituents are more nearly uniform and tend to flocculate over a much narrower temperature range. It should be possible to obtain a number of fractions in which all of the gelatin would flocculate at a definite temperature. The aggregation of the various gelatin constituents is presumably due to loss of thermal energy, and the temperature at which this occurs must be some function of the mass of the constituent. It is natural to assume, then, that the constituents which flocculate at a given temperature are larger than those which remain in solution at that temperature. Recently, Krishnamurti and Svedberg (1930) have obtained evidence with the ultra-centrifuge that the constituents of a gelatin solution are heterogeneous as to mass, even at a pH value at which there is no tendency toward aggregation.
There is much reason to suppose that the gelatin constituents do not differ very greatly chemically since different fractions have the same refractive index and the same isoelectric point.
The data as a whole are best explained by considering the gelatin constituents to be different degrees of association of the same or very similar molecular structural units. This is in agreement with Sheppard and Houck (1930), who consider that "the molecules of gelatin are fundamentally identical with those of collagen, the difference being only in the degree of association and orientation". Meyer and Mark (1928) have interpreted the x-ray data obtained from collagen as indicating that the micelles of the collagen fiber are built up of main valency chains of anhydro-amino acids. It may be supposed that during peptization of these fibers, the amino acid chains become separated, disorientated, and partially broken up, so producing the heterogeneous system which we know as gelatin. It is evident that the manner in which this breaking-up proceeds depends upon the chemical treatment previous to the peptization process and the gelatin produced from lime-treated collagen would be expected to differ from that from acid-treated collagen.
From the results herein reported it seems evident that the technique of isoelectric flocculation of electrolyte-free gelatin offers a profitable method for the study of gelatin and an extended investigation along these lines should yield much valuable information concerning the nature of gelatin. It is possible that this method may also be extended to other hydrophilic colloids.