The intensity of Arthus reactions was dependent upon the relative amounts of precipitating antitoxin and toxoid used. Severe lesions were caused by intravenous sensitization with 0.48 mg. precipitating antitoxin N and intradermal challenge with 0.05 or 0.17 toxoid N. Reactions of lesser intensity were caused by smaller amounts of antitoxin and toxoid.

The nature of the antibody used for sensitization was of importance in the severity of Arthus reactions. In contrast to the behavior of precipitating antitoxin, amounts of non-precipitating antitoxin equivalent to 0.48 mg. N did not cause severe Arthus reactions when 0.05 or 0.17 mg. toxoid N was given intradermally.

Precipitating antitoxic whole serum is altered by heating at 56°C. for 5 hours so that its precipitability is lost without any appreciable loss of antitoxic strength. This modified antitoxin produced Arthus reactions of only intermediate severity in guinea pigs sensitized with 0.48 mg. antitoxin N. When fractions of precipitating antitoxic serum were obtained using a cold ethanol technique described by Deutsch (16), a mixture of the purified ₂-globulin and the crude albumin fraction heated together at 56°C. for 5 hours behaved similarly to whole serum. However, ₂-globulin alone was not affected by the heating procedure, remained precipitable by toxoid, and was able to cause a severe Arthus reaction following sensitization with 0.48 mg. antitoxin N.