A Critical Study of Experimental Fever. 



175 



Doulton filters or by the use of the centrifuge, as shown by subsequent 

 injection of the water. The actual presence therefore of organisms at the 

 time of injection of sterilised saline did not appear to be the sole cause of 

 salvarsan fever, nor of " salt fever " in general. 



Since reporting these results we have studied the question more fully. 

 Injection of the centrifugalised deposit from 250 c.c. of water shown to be 

 pyrogenetic, and to contain 73,000 organisms per cubic centimetre, produced 

 no fever. This was also true of a deposit from 75 c.c. of saline shown to be 

 pyrogenetic, and to contain before the use of the centrifuge 950,000 organisms 

 per cubic centimetre. In both cases the medium employed for injection of 

 the organisms was first shown to be free from the heat-stable fever-producing 

 body referred to. On the other hand injection of suitable quantities of water 

 containing this body, but only 40 to 160 organisms per cubic centimetre, 

 produced fever whether salt were subsequently added or not. This body is 

 held back by Martin's* gelatine filter, a fact which shows that it is 

 a colloidal substance in a fine state of dispersion. We conclude, therefore, 

 that contamination by this body of water before mixture with salt is a more 

 important factor in salvarsan fever than we had realised. 



The unexpected presence of this hitherto unrecognised body must to some 

 extent vitiate deductions drawn from previous work on the causation of 

 fever which is the sequel to injection of a variety of substances dissolved or 

 suspended in water. 



We have therefore determined to what extent the fever which had been 

 known to follow injection of salt, sugar, fibrin ferment, and tissue extracts 

 was due to this contamination. 



Experimental Methods. 

 Before dealing with our evidence, it is necessary to refer to the current 

 belief that it is impossible to study experimental fever with any approach to 

 accuracy. The number of variants is admittedly great. Full experimental 

 data dealing with the most important will be published in due course. 

 They include the effect on body temperature of disease, age, sex, weight, 

 breed, food supply, environment, exercise, rest, repeated thermometric 

 observation, and other factors. Prolonged study has taught us that to 

 establish broad effects, such as the ability of any given substance to produce 

 fever on injection, knowledge of all the variants is necessary but sufficient. 

 On the other hand, no comparative observations can be relied on unless the 

 ratio of the volume of an injection (including weight of material dissolved or 

 suspended in injection fluid) to body weight is kept constant. In every 

 * C. J. Martin, " Gelatine Filter," ' Journ. Physiol.,' 1896, vol. 20. 



