Numerous investigators have shown that if the blood serum of a patient suffering from carcinoma be mixed with normal human red blood cells hemolysis occurs. The reaction takes place in from 50 to 80 per cent, of cases of malignant disease. It occurs with considerable frequency in tuberculosis, and more rarely in other diseases.
It occurred to me that, by the injection of red blood cells under the skin of the carcinoma patient, it might be possible to produce a local reaction at the site of the injection. Logically, a local reaction should take place. Theoretically, such a reaction might prove to be a delicate one; it might give more positive and definite results than the test-tube method. In the technic which is used for the test-tube method the presence and degree of hemolysis is indicated by the amount of laking of the red cells—that is, by the amount of