ANAPHYLAXIS 399 



that the property of sensitiveness in anaphylactic animals may be an 

 attribute of the cells, independent of the substances circulating in the 

 blood. Can we reconcile these apparently opposed facts ? 



It would seem to us most rational to look upon the problem in the 

 following way: The antibodies produced by the body in response 

 to the injection of an antigen are, of course, the products of the cells, 

 and it is likely, on the basis of experiments and data considered in 

 another chapter, that many or all the cells of the body with which the 

 antigen comes in contact may participate in their production. At 

 different periods during the process of immunization the antibodies 

 are therefore present, in varying ratio, both in the circulation and 

 within the cells. During the earlier stages of the response to the 

 antigen, i. e., during the period of hypersensitiveness, it is likely 

 that the reaction changes (antibodies) are more plentiful in the 

 cells than in the blood stream. When antigen is injected into a 

 sensitized animal, it is likely that it goes into reaction with both the 

 circulating and the sessile antibody. The latter is probably most 

 important in the ordinary reaction of anaphylaxis, the former is 

 probably of great importance in such cases as the sudden death which 

 can be seen in highly immunized rabbits, when a fourth or fifth 

 injection of a foreign serum is given (at a time at which the blood 

 already strongly reacts with injected antigen). Both the cellular 

 and the intravascular reaction probably occur in all cases, although 

 we are inclined to believe that the cellular reaction must be taken 

 to dominate ordinary anaphylaxis, as observed experimentally. That 

 the intravascular reaction, however, may also have importance is 

 testified by the experiments which we have cited (pp. 383, 384 and 

 398) and which cannot be ignored. 



Now, granted that the reaction takes place in both the cells and 

 the circulation, varying in relative intensity in each location accord- 

 ing to the stage of antibody formation, and the relative concentration 

 of the antibodies in cells and in blood stream, in how far are we 

 justified in assuming that the complement or alexin of the circu- 

 lating blood is necessary for the production of anaphylactic shock? 

 The experiments of Friedberger and many others have shown beyond 

 doubt that the action of complement upon antibody-antigen complexes 

 may produce poisons, and much evidence has been gathered to show 

 that complement is diminished in animals during shock. This consti- 

 tutes reasonable ground for assuming that the complement partici- 

 pates at least in that phase of anaphylaxis which takes place in the 

 blood stream. Whether or not the circulating complement acts upon 

 those antibody-antigen complexes which are formed on the sensitive 

 cell, is hard to decide. Experimentally it cannot be absolutely de- 

 termined, since it would be quite impossible to remove all traces of 

 complement from any cell. Moreover the complement is, after all, 

 also a cell product, and it is more than likely that the cell disposes 



