Inflammation and Related Phenomena 127 



cytosis) . Furthermore, repeated injections of endotoxin lead to a 

 state of tolerance in which further injections fail to cause pyrexia. 

 Following repeated doses of one endotoxin, tolerance develops to 

 endotoxins from other bacterial species and also to the pyrogenic 

 action of infections and trauma. This refractory state lasts for about 

 two weeks and is not an immunological phenomenon. 



Since these polysaccharide pyrogens are present in both bacteria 

 and tissues they might in theory account for fever in both infective 

 and non-infective states. The falsity of this assumption was demon- 

 strated by the extraction from polymorphonuclear leucocytes of 

 another quite different substance that also caused fever on injec- 

 tion. This leucocytic pyrogen, as it is called, is followed on injection 

 by fever, with the intervention of only a very short latent period. 

 In addition, repeated injections do not call forth tolerance and 

 animals tolerant to endotoxin respond to leucocyte extracts with 

 fever. Finally, leucocytic pyrogen does not cause significant leuco- 

 penia. 



The relationship between endotoxin and leucocytic pyrogen 

 was then studied. It was found that injected endotoxin disappeared 

 rapidly from the blood stream and that simultaneously, the level of 

 another circulating pyrogen with the properties of leucocytic ex- 

 tracts rose. The concentration of this endogenous pyrogen was re- 

 lated to the height of pyrexia. As a result of this experiment it was 

 suggested that endotoxin acts by damaging leucocytes which then 

 liberate endogenous pyrogen. Endotoxin is known to cause leuco- 

 cytes to adhere to vessel walls and this might be associated with re- 

 lease of their pyrogen. It now seems more likely, however, that 

 endotoxin acts directly on the leucocytes. Endogenous pyrogen 

 indistinguishable from leucocytic pyrogen has been demonstrated 

 in the blood of rabbits at the height of fever due to influenza virus, 

 pneumococcal septicaemia and non-septicaemic streptococcal cellu- 

 litis. 



At this stage it seemed probable that fever was due largely to 

 liberation of endogenous pyrogen, derived probably from poly- 

 morph leucocytes. This view was confirmed by experiments show- 

 ing that endogenous pyrogen caused fever more quickly if it was 

 injected directly into the arteries supplying the thermo-regulatory 



