REACTIVITY OF MALIGNANT NEOPLASMS 253 



void ot lethal effect and yet retained their power ot eliciting 

 hemorrhagic necrosis in the skin ot a high percentage ot pre- 

 pared rabbits. This may i)e illustrated by the tollowing example: 



T^venty-fi\e reacting iniits of B. fyphosus "agar washings" fil- 

 trate T.1938 injected intraxenously killed approximately 90 per 

 cent of rabbits tested; 200 reacting units of B. fypliosiis "agar 

 washings" filtrate T.1938 mixed with 200 antitoxic units of horse 

 serum H.682, gave no reactions in prepared rabbits and had no 

 lethal effect; a mixture of 500 reacting units of B. tyj)lws}Ls "agar 

 washings" filtrate T.1938 with 200 antitoxic units of horse serum 

 H.682, elicited severe hemorrhagic and necrotic reactions in 8 

 otit of 10 rabbits tested and had no lethal effect; a mixture of 

 750 reacting units of B. typhosus "agar washings" filtrate T.1938 

 with 200 antitoxic units of horse serum H.fi82 produced reactions 

 in 7 and killed 2 out of 8 rabbits tested. 



From a small series of preliminary experiments it was seen that 

 a decided protection was afforded to normal mice by the antisera, 

 inasmuch as 200 reacting" units of B. tyljJiosus filtrates mixed "with 

 200 units of antitoxin produced no effect when injected intrave- 

 nously. A mixture of 500 reacting units of the filtrate plus 200 

 units of the serum killed 2 out of 4 mice while it failed to 

 kill any rabbits. It could appear, therefore, that the mixtures were 

 more toxic to normal mice than to rabbits. This, however, re- 

 mained questionable because the dose per ^veight \\'as much 

 larger in mice than in rabbits (i.e., 1.25 c.c. per kilo of body 

 ^veight of rabbits and 1 c.c. per each mouse) . 



Diagram 4 represents results obtained ^vith B. I\IjIiusus "agar 

 washings" filtrates and mixtures thereof with homologous neu- 

 tralizing horse serum (H.682) in 132 tumor-bearing mice. 

 Se\enty-nine ^vere injected intravenously and 53 mice intraperi- 

 toneally. Inasmuch as no appreciable difference Avas obtained in 

 the mice treated, the figures were added together for simplicity 

 of presentation. 



As is seen from Diagram 4, the addition of serum afforded an 

 unquestionable protection against the lethal effect of the filtrates. 

 With certain amounts there was no mortality obtained Avithin 

 the first t^venty-four hours and only a low mortality Avithin two 

 to t^venty days following the injection (Groups 1 and 2) . As the 

 amount of toxin in the mixtiue Avas increased, the mortality rate 

 ^\as proportionately raised (Groups 3-5) . There was obser\ed, 

 however, a discrepancy in Groups 1 and 2 ^vhere a higher late 



