The results of the foregoing experiments may be briefly summarized as follows:
The amount of antitoxic substance obtained by precipitation with magnesium sulphate from the blood-serum of the horse corresponds, as nearly as can be determined by the use of test guinea-pigs, in full to the protective power of the serum from which it is obtained, i. e. the precipitate from 1 cc. of serum will protect against the same amount of toxin as 1 cc. of the serum itself.
Equal amounts of the precipitates by magnesium sulphate from immunized and non-immunized horses act differently toward toxin; i. e. the proportion of protective substance to the precipitate from non-immunized serum is exceedingly small as compared with the proportion of antitoxin to the precipitate from sera of immunized horses.
The average precipitate from the sera of immunized horses, as obtained by magnesium sulphate, is more abundant than the average precipitate from sera of non-immunized horses.
In the case of the same animal before and after immunization, the serum before immunization gives a less abundant precipitate with magnesium sulphate than the serum tested after immunization.
The proportion of increase per unit of antitoxic strength for the same or different horses is not constant. This may be due to an increase of inactive substances (in their relation to diphtheric toxin) or to imperfect methods of determination.
The precipitates obtained by magnesium sulphate give all the reactions recognized as characteristic of globulins, and as distinguishing them from other albuminous bodies. We are not warranted, then, in the present state of our knowledge, in considering any part of these precipitates as other than globulin. But it does seem warrantable to conclude, from the fact that the globulins of normal serum do not protect, or only in comparatively large amounts, against diphtheric toxin, that new globulins are formed, or rather greatly increased in the serum of immunized horses, and that these globulins protect against the toxin. These increased globulins and the inert globulins (which from obvious causes are a very variable factor) are both precipitated by magnesium sulphate.
Every animal has a physiological and pathological history more or less widely diverging from the normal, hence absolute conformity in the results obtained is not to be expected, at least with our present methods of differentiation.