Further studies of specific precipitin antiserums for the protein of cancer tissue; relation of the proteins of different malignant tissues to each other.

When rabbits are injected with tissue proteins, precipitin antibodies for blood serum proteins appear in the serum in addition to precipitins for the tissue proteins (5, 6, 2, I). There are three methods by which the former can be removed; namely, absorption in vitro, absorption in vivo, and allowing sufficient time to elapse following the injection for the exhaustion of blood proteins from the depots of tissue fixed on aluminum cream. Spinka and Weichselbaum (6) attempted to remove the blood serum protein precipitins from kidney and liver antiserums by absorption in vitro, using solutions of blood albumin and pseudoglobulin. They were able to rid the serum of the blood protein antibodies, but the tissue protein antibodies were also removed partially or entirely and invariably the potency of the tissue protein precipitins was definitely reduced. The absorption in vitro method failed to give us satisfactory results in similar experiments with cancer tissue. Hektoen and Welker (4) observed that the negative phase in precipitin production was specific. In rabbits immunized to many antigens, the injection of a single antigen resulted in the disappearance of the antibody for that antigen from the blood stream, with little or no effect on the titers of the rest of the antibodies. Spinka and Weichselbaum (6) were unsuccessful with the in vivo absorption technic in their work on precipitins for liver and kidney proteins. These investigators found that if sufficient time is permitted to elapse after the injection of tissue fixed on aluminum cream, the blood protein precipitins begin to drop in titer and finally disappear altogether. The presumption is that the blood proteins, being present in much smaller amounts than the tissue proteins, will disappear from the depots long before the tissue antigens become exhausted. This observation had been made previously by Hektoen and Welker (3) in their work on multiple antigens fixed on aluminum cream. When the serums of these animals no