or an immune serum, can be diminished by the use of alcohol.
"2. I have studied the action of alcohol in drinking and abstaining persons on the hæmolytic power of blood-serum over heterogeneous red blood-corpuscles (rabbits). I have studied not only the haemolytic power of the human blood-serum, but also its power of precipitation in the presence of rabbit-serum, with a view to ascertain if the reaction between a known dilution of rabbit-serum and a certain dilution of serum of alcohol-users and non-drinking persons is different or not, and if the reaction is more apparent with the former or with the latter.
"3. The resisting power of serum obtained both from alcohol-drinking and from non-drinking persons was further tested by human blood, with the object of discovering whether any difference in reaction existed between the same immune serum and the two kinds of human sera above mentioned.
"4. I have studied the problem as to whether the hæmolytic complement in the blood-serum of alcohol-drinking and non-drinking persons is altered in any way by alcohol.
"5. The bactericidal power of blood-serum from both alcohol-drinking and non-drinking persons was determined by some experiments.
"The above experiments have given the following results:
"1. The normal resistance of human red blood-corpuscles appears to be somewhat diminished against a heterogeneous normal serum or an immune serum by the consumption of alcohol, provided that tolerably large equal, or nearly equal, numbers of drinkers and abstainers of both sexes be examined, and the average of resistance be taken on both sides: this last-named precaution being necessary because the resistance of red blood-corpuscles from different human beings varies largely. The difference is often greater when using weaker solutions than when using stronger dilutions of lysin.
"2. These experiments have shown the normal hæmolytic power of human blood-serum to be less in the case of alco-
