THE PROBABLE MECHANISM OF THE REACTION 117 



simple and in which speculation and conjecture must enter to a 

 larger extent. Perhaps it will be most satisfactory to discuss the 

 status of our information in a definite variety of diseases, each dif- 

 fering in fundamental pathology, and each yielding different results 

 when nonspecific therapy is used. 



(A) In typhoid fever we deal with an infection originating with a 

 gastro-intestinal infection followed by an infection chiefly of the 

 lymphopoietic system and a concurrent bacteriemia. Here the in- 

 toxication is principally due: (1) to native proteins derived from 

 the typhoid bacillus and to which the organism has become sensi- 

 tized during the incubation period of the disease, (2) to very toxic 

 protein split products contained in the body of the bacterium (endo- 

 toxins), and (3) to split products from the bacterial cell when it 

 undergoes proteolysis. The disease is one to which an active im- 

 munity can be established. 



(B) In pneumonia, on the other hand, we are dealing with an in- 

 fection differing from this materially. Leaving undetermined its method 

 of invasion whether primarily a blood invasion that localizes in the 

 respiratory tract or an extension directly along the respiratory tract 

 (which seems most probable), we deal with a localized inflammatory 

 process, probably without sensitization of the body to the native 

 protein of the pneumococcus, a process in which the toxic manifesta- 

 tions are due to an absorption of the higher split products derived 

 from an inflammatory exudate (fibrin, cell detritus, etc.) located 

 actually on the surface of the body (considering the pulmonary alveoli 

 in their actual relation to the surface an invagination, not from 

 their mere anatomical position as part of the internal organs). Be- 

 cause of the vascular supply the alveolar surface is naturally an 

 ideal absorbing structure, so that, while largely isolated during the 

 course of a lobar pneumonia from the general circulation, absorption 

 is still sufficiently great to cause a profound intoxication. The 

 pneumococcus protein itself is not particularly toxic and as already 

 stated the organism not necessarily sensitized to it; the bulk of the 

 toxic manifestations must come from the splitting of the exudate 

 in the pulmonary alveoli. It will be recalled that Kaznelson found 

 the split products obtained from fibrin particularly toxic. Here 

 we deal with recovery frequently by crisis rather than by lysis and 

 with an immunity of low grade and of relatively short duration. 



(C) Finally we might take as an example a pure septicemia of the 

 streptococcus type, without localization, where intoxication is as- 

 sociated with marked virulence on the part of the organism rather 

 than by the production of a soluble toxin from the bacterium or a 

 toxic autolytic product from tissues of the invaded host. 



Typhoid Fever. Our observations in typhoid fever are perhaps 

 most complete and from them we may be able to construct a picture of 

 the processes that follow the nonspecific reaction. Let us assume that 



