374 INFECTION AND RESISTANCE 



what has been said before, but which is again emphasized because of 

 its very important bearing upon later theoretical considerations. 



Variations in experimental anaphylaxis are, to some extent, de- 

 pendent upon the manner in which the antigen is introduced into the 

 body. It is now well known that sensitization may be accomplished 

 by a first injection given subcutaneously, intravenously, intraperi- 

 toneally, or intrapleurally. At the second or toxogenic administra- 

 tion shock may be probably best induced and with the smallest 

 quantities by the intravenous method. Besredka and Stein- 

 hardt, 49 50 51 52 who began their studies soon after the first publica- 

 tions of Eosenau and Anderson, came to the conclusion that the most 

 effectual and rapid method of producing the anaphylactic shock 

 consisted in direct injection into the brain. Curiously enough, while 

 Besredka and Steinhardt obtained the most violent reactions by 

 injection of the second or toxogenic dose into the brain, they were 

 unable to sensitize by this path, at least with doses of 1/4000 c. c., 

 which sufficed to sensitize by the intravenous method. Rosenau and 

 Anderson, in repeating this work, obtained similar results with very 

 minute amounts, but found that intracerebral sensitization could be 

 accomplished by doses of .0001 c. c., or more. According to them, 

 animals intracerebrally sensitized become anaphylactic more rapidly 

 than those in which the injections were subcutaneous. In the 

 former the incubation time was about 7 days, while in the latter it 

 was never less than 9. Lewis, 53 in his thorough study on the same 

 subject, made extensive use of the direct intracardial method of 

 injection. In other words, any method of introducing the foreign 

 protein into the blood or tissues seems to lead both to sensitization 

 and to toxic effect, and those methods which introduce the substance, 

 on reinjection, directly into the blood stream or the brain induce 

 the most violent symptoms with the relatively smallest dosage. Ac- 

 cording to Otto and others, the subcutaneous method, while followed 

 by less violent symptoms, is the method to be preferred when ques- 

 tions of specificity are involved, for, while the reaction is specific in 

 the ordinary sense, yet it is extremely delicate and therefore, as 

 Eosenau and Anderson put it, "quantitatively specific.' 7 The less 

 violent subcutaneous method, therefore, might be said to have the 

 same purpose here that dilution of the antigen or immune serum has 

 in safeguarding against error when carrying out specific precipitin or 

 iigglutinin reactions. 



49 Besredka and Steinhardt. Ann. de VInst. Past., p. 117, 1907; ibid., 

 p. 384. 



50 Besredka. Ibid., p. 777, p. 950, 1907; ibid., p. 496, 1908; p. 166, 1909; 

 p. 801, 1909. 



51 Also: Bull, de VInst. Past., Nos. 19, 20, 21, 1908; No. 17, 1909. 



52 Also: C. E. de la Soc. Biol, p. 478, 1908, Vol. 65; p. 266, 1909, Vol. 

 67. 



53 Lewis. Jour. Exp. Med., Vol. 10, 1908. 



