IMMUNOLOGICAL RELATIONSHIPS 345 



jection of horse serum alone ^vas sufficient for elicitation of 

 hemorrhagic necrosis, the additional injection of bacterial filtrate 

 did not enhance the reaction. It may he concluded, therefore, from 

 all the data described thus far that the phenomenon of local 

 tissue reactivity should be clearly diff^erentiated from the Arthiis 

 phenomenon, ^vhich may take place in the same animals after a 

 considerably longer period of time of sensitization and upon 

 which the bacterial filtrates seemingly have no influence. 



3. In a fmther ptiblication Gratia and Linz (1932^) bring 

 additional arguments in faxor of their hypothesis of the relation- 

 siiip bet^vecn the Arthus phenomenon and the phenomenon of 

 local skin reacti\ity. They state that the intravenous injection of 

 0.5 c.c. of horse serum into rabbits folloxved eight weeks later 

 by the intradermal injection of approximately 0.1 c.c. of horse 

 serum at intervals of eight days, causes the development of the 

 Arthus reaction, i.e., at the beginning, a xvhite or pink edema 

 folloxved on stibseqtient injections by congestion and later, by 

 hemorrhagic reactions. After these long intervals of time, the 

 reaction finally becomes identical in appearance \\'\t\\ the phe- 

 nomenon of local skin reactivity. Hoxvever, according to the 

 same authors, not all rabbits are susceptible to this hemorrhagic 

 Arthus phenomenon. Five to 20 per cent are very susceptible, 

 while 60 to 70 per cent shoxv average stisceptibility and 15 to 20 

 per cent are refractory. These authors assimie that there exists a 

 suggestive similarity with the Arthus phenomenon inasnrtich as 

 the phenomenon of local skin reactivity occurred in similar nimi- 

 bers of rabbits with the B. coll culture filtrates used in their ex- 

 periments. It may be stated, however, that no parallelism can be 

 draxvn even on piuely statistical grounds. Potent "agar washings" 

 filtrates of meningococcus used in adequate doses prodtice the 

 phenomenon of local skin reactivity of considerable sexerity in 

 100 per cent of rabbits. This is jtidged by experiments in approxi- 

 mately over 5,000 rabbits (p. 28) . 



4. Being unable either to contradict or disprove the observa- 

 tions on the lack of specificity in the elicitation of the phenome- 

 non, Gratia and Linz thought that it remained for them to show 

 that anaphylaxis itself xvas not specific. On the basis of the fol- 

 loxving experiment, they came, then, to the tuiexpected conclu- 

 sion that the Arthus jahenomenon is not specific in rabbits. Rai)- 

 bits xvere sensitized either by an intraxenous injection of 1 c.c. 

 of horse serum or sheep serimi. Two months later the sensitized 



