The generally held view has been that in any immune serum only a single antibody would be induced by and react with a single antigen. Were this true the various manifestations of antibody activity should show a quantitative parallelism. It has already been shown (1), however, that with antipneumococcus horse serum the mouse protective potency does not parallel the maximum amount of specifically precipitable protein except within certain well defined groups of antisera. The simplest explanation of this situation is that different horses form antibodies differing in specific protective capacity, but from our studies it seems probable that in any immune serum there may occur a mixture of antibodies which, while directed against the same antigen, possess different protective capacities, different avidities, etc. It would now appear that this latter hypothesis is the more tenable since the experiments here reported indicate the existence of antibodies of various protective potencies in horse antisera. It would not be unreasonable to hold that the antibodies of a single serum represent a series of substances with varying properties. On the basis of the present immunological fractionation experiments, the following deductions seem permissible.
1. Antipneumococcus horse sera must contain at least three, possibly many, antibody substances which react with the capsular polysaccharide. These are: (a) A substance which precipitates upon the addition of a relatively small amount of polysaccharide. This antibody possesses a low protective potency. (b) A substance which is precipitated with intermediate amounts of polysaccharide and which possesses an extremely high protective value. (c) A third substance which is precipitated only with the addition of relatively large amounts of polysaccharide. The protective value of this antibody is very low It may represent a degraded form.
2. With antipneumococcus rabbit serum the situation is somewhat simpler. This is in accord with the fact that with Type I antipneumococcus sera from this species there is a direct proportionality between the amount of specifically precipitable protein and the protective potency of the serum (1). The results with antipneumococcus rabbit serum indicate the existence of at least two antibody substances: (a) An antibody with high protective value which makes up the greater proportion of the total content. (b) A second substance which is precipitated only upon the addition of relatively large amounts of capsular polysaccharide. The existence of this second antibody is not clearly demonstrated by the present findings but the lower protective ratios obtained as greater amounts of antibody are removed probably indicate its existence. This may also represent degraded material.
The observations on the antibodies of both horse and rabbit antisera will be supported by experiments with immunochemical fractionation which will be reported in a subsequent paper.