Relatively accurate inductions with regard to the origin of febrile heat were first rendered possible by the experiments of Billroth and Weber. These observers found that the introduction of putrid material into the circulation of animals produced fever. It was afterward shown that various substances, not necessarily of a putrid character, might produce the same result. From measurements with the calorimeter of the heat produced, it was concluded by Wood[9] that in the fever of pyæmic dogs more heat was produced than in healthy, fasting dogs, although less than in high-fed, healthy dogs. An increased production of heat in the fevered animal is thus obvious, as his capacity to receive and assimilate food is considerably less than that of a high-fed, healthy dog. The calculations of Sanderson, referred to by Wood,[10] based upon the analyses of eliminated carbonic {40} acid and urea, show that the febrile human subject produces very much more heat than the fasting, though less than the fully-fed, healthy, man. [Footnote 9: _Op. cit._, 236.] [Footnote 10: _Op. cit._, 239.] An increased production of heat in fever is generally admitted, although it alone is not to be regarded as the essential feature in the elevated range of the temperature. The fasting man or animal under ordinary circumstances is not febrile, and an increased production of heat from full feeding in health, equal to that observed in fever, not being associated with fever, it is apparent that the retention of the produced heat is of importance for the existence of fever. Although it has been shown by various observers that more heat is dissipated during fever than in health, this increased loss is not in proportion to the increased production of heat. A persistent elevation of temperature is the necessary result. This elevation is subject to daily and hourly differences, as is the temperature of the healthy individual. These variations in the range of the febrile temperature are apparently due to an agency like that which dominates the course of normal temperatures--viz. a varying action of the vaso-motor nervous apparatus, as well as of that controlling the secretion of sweat, now permitting, now checking, the dissipation of the produced heat.