322 LOCAL TLSSUE REACTIVITY 



becoiiies poisonous just jjiior to the appearance ol coagulation. 

 The speed of poison prodiK tion corresponds to that of anaphyla- 

 toxin produced by agar. The toxic ity of raljbit serum is influenced 

 greatly l)y the conditions under which defibrination is effected. 

 Rod defibrination gives the least toxic serum. The most toxic 

 serum is obtained by bead defif^rination carried out at low tem- 

 perature. In my experiments, normal ral)bit blood and blood of 

 rabbits injected intravenously ^vith to per cent peptone shortly 

 before bleeding w^is collected into flasks containing glass beads, 

 chilled wdth ice and alcohol (temperature minus 5° c.) , and 

 rapidly dehbrinated by vigorous shaking. The defibrinated blood 

 was then injected an hour later into ral)bits prepared twenty-fcjin^ 

 hours previously by intradermal injection of 0.25 c.c. of potent 

 B. typhosus culture filtrate. No reactions followed the injection of 

 blood in these rabbits. 



Novy and DeKruif made the interesting observaticm that the 

 mixture of distilled w^ater and serum of a rat sensitized to horse 

 serum becomes rapidly anaphylactic. Accordingly, it w^as attempted 

 by me to determine whether distilled \vater induces reacting 

 potency in antibody-containing whole blood and blood serum. 

 In one group of experiments rats received seven ^veekly intra- 

 venous injections of horse serum. One week after the last injec- 

 tion of horse serum the rats were injected twice intravenously 

 ^vith 2 c.c. of distilled water at the interval of twenty-four hours. 

 T^vo hours after the last injection of distilled water the blood 

 obtained from the heart was rapidly defibrinated in chilled con- 

 tainers containing glass beads and injected intravenously into 

 rabbits ^vhich had previously received a single intradermal injec- 

 tion of a potent bacterial filtrate. No reactions resulted from these 

 injections in the prepared skin sites. 



In another group of experiments the sera employed w^ere those 

 which in mixture with the respective antigens were able 

 to elicit strong reactions in prepared sites, namely, anti-horse rab- 

 bit serum, anti-rabbit horse serum and anti-human horse serum. 

 The distilled water was added to the antiljody containing sera 

 in amounts 4 and 9 times the volume oi the serum. The sera 

 thus diluted with distilled water were incubated at 37° c. for one 

 hour and injected into prepared rabl)its in doses of 3 c.c. per kilo 

 of body weight. The addition of g volumes of distilled water usu- 

 ally resulted in considerable turbidity apparently due to precipita- 



