538 BACTERIOLOGY. 



cerned in the production of the so-called " hemorrhagic 

 septicsemias." When running their normal course the 

 specific organisms of these diseases cause typical septi- 

 caemias in susceptible animals ; but often, from causes not 

 entirely clear, the animals die with only local lesions, or 

 with but very few organisms in the internal viscera. We 

 see here conditions analogous to those observed in the 

 two experiments with anthrax, viz., we find a group of 

 diseases that are properly classed as septicsemias, be- 

 cause of the usual general invasion of the body by the 

 organisms concerned in their production, but which 

 frequently assume a purely local character — in both 

 instances proving fatal to the animal infected. From 

 what we have seen it is manifestly probable that, 

 whether these diseases be designated as septicsemias or 

 as toxaemias, death is produced in all instances by 

 poisonous substances that are generated by the infecting 

 bacteria. In the case of typical anthrax, and other 

 varieties of septicemia, the production of this poison 

 is associated with the general dissemination of the 

 organisms throughout the body ; while in those infec- 

 tions often referred to as toxaemias, of which diphtheria 

 may be taken as a type, the poison is produced by the 

 organisms that remain localized at the site of invasion, 

 and is thence disseminated throughout the body by the 

 circulating fluids. Infection thus far, then, appears to 

 be a chemical process. 



Through special investigations that have been made 

 upon the products of growth of certain pathogenic bac- 

 teria this opinion receives further confirmation ; it has 

 been found possible by the use of appropriate methods 

 to isolate from among the mass of material in whicli 

 certain of these organisms have been firtificia,lly cujti? 



