We have tested these preparations in vitro and in vivo for inactivating or neutralizing or, to use perhaps a better term, antiviral effects against a constant, potent, filtrate virus of poliomyelitis.

The preparations exhibited these effects when combined in vitro. Their action in this respect appears to be greater and more constant than that found by Stewart and Haselbauer for the Pettit antipoliomyelitic horse serum.

On the other hand, in vivo tests carried out by us were less successful. In comparison with the constancy of action, under given conditions, of convalescent monkey and human sera, the antipoliomyelitic horse serum displayed striking irregularity, and certain preparations were devoid of protective power.

The precise nature of the inactivating substances in the horse antiserum and their relation to the corresponding substances in convalescent sera have still, to be determined. As far as one absorption test carried out by us indicates, precipitin does not play a major rôle in the inactivating process.

When an active globulin concentrate was filtered through Berkefeld candles, it lost its in vitro inactivating power. This is not true of convalescent sera in the native state. No tests have, however, been made with globulin concentrates from such sera.

The experiments described in this paper raise the question whether, therapeutically considered, the antipoliomyelitic horse serum should be regarded as an exact equivalent of, and hence employed as a perfect substitute for, convalescent serum. This question can only be answered by further experiment and observation.