178 Messrs. E. C. Hort and W. J. Penfold. [Feb. 17, 



marked fall of temperature. The different extent of fever excited by- 

 different injection ratios when one sample of water containing F.P.B. is 

 used, is ,'clearly seen in nearly all the charts here shown. In 1910 one of 

 us (4) advanced the theory that water fever is largely an auto-intoxication 



Ol 2 345678 10 12 



A. Eabbit, 2083 grin., injected subcutaneously witb water containing F.P.B. Injection 



ratio, 1/40. 



B. Eabbit, 2080 grm., injected subcutaneously with water containing no F.P.B. Injection 



ratio, 1/40. 



(Interval between observations, 30 minutes.) 



due to absorption of lytic products liberated locally by the injection of 

 water. The evidence given in support of this view now disappears, since 

 water supposed to be pure, and known to contain only 160 organisms per 

 cubic centimetre, may actually contain F.P.B. 



Salt Fever. 



Charts 4 and 5 show that solutions of salt made with ordinary distilled 

 water give rise on injection to fever, but that when freshly distilled water 

 is the solvent a fall of temperature results. These effects were obtained 

 in all our experiments up to 25-per-cent. concentrations of sodium chloride. 



