On the whole, these experiments show that in fever the increase in metabolism and of body temperature occur simultaneously. Further experiments planned along these lines, using other fevers and perhaps even shorter periods, will, it is hoped, give clearer evidence whether there is any reason for an increased heat production other than a rise in body temperature; or whether there is usually a preliminary stimulus to increased metabolism before the rise in body temperature occurs.

Nebelthau has shown a fall in temperature and heat production in a rabbit whose cord was divided between the sixth and seventh cervical vertebrae, and has also demonstrated that under these circumstances infection with erysipelas of the pig had no influence on temperature or heat production. The inference is that the febrile toxins act through the higher vasomotor centers, whose regulatory control is lost in the above experiment.

A kindred interpretation may be placed on the experiments of Mendelson,1 who was unable to produce fever through pus injections when the dog was under the influence of chloral or morphin, although such treatment in a normal animal caused a rise in temperature of from 36.30 to 39.90 in forty-five minutes. Mendelson also finds a constant constriction of the renal blood-vessels in fever.

1 Mendelson: "Virchow's Archiv," 1885, c, 274.

Further experimentation convinced Sawadowski1 that fever cannot be produced after the mid-brain has been severed from the medulla, whereas if the mid-brain be left intact, but the cerebrum be sectioned from it, fever may be induced in the ordinary course. Citron and Leschke2 have found that destruction of the median portion of the 'tween brain on the boundary between the optic thalamus and the corpus quadri-geminum anterius, the "'tween brain puncture," converts a rabbit into the equivalent of a cold-blooded animal. Under these circumstances it is impossible to produce infective or non-infective fevers of any kind. The toxic substance must therefore act on nerve-cells in the mid-brain, which, in turn, stimulate the medullary centers.

At times during high fever the skin may be red and the peripheral blood-vessels distended. Although there is no sufficient explanation for the continuance of fever when the radiation and conduction of heat from the surface of the body are thus increased, Krehl3 suggests that the quantity of blood flowing through the vessels at the time may be inadequate to reduce the body's temperature.

The second means of physical regulation of the body temperature is through the evaporation of water from both the lungs and the sweat-glands. It might be surmised that the activity of this mechanism was reduced in fever. Nebelthau4 has shown that the heat lost by evaporation of water and by radiation and conduction bear exactly the same ratio to each other in normal and in fever-infected rabbits. Since Rubner (p. 140) has proved that the elimination of water in normal animals greatly increases at high temperatures, the mere maintenance of the usual water evaporation during fever would of itself be abnormal.

1 Sawadowsky: "Centralblatt fur medizinische Wissenschaft," 1888, xxvi, 161.

2 Citron and Leschke: "Zeitschrift fur ex. Path, und Ther.," 1913, xiv, 379. 3 Krehl: "Pathologische Physiologie," 1904, p. 453. 4 Nebelthau: Loc. cit.

Lang1 has shown that the elimination of sweat is reduced during the rise of temperature in man, but at the height of fever is the same as the normal, while there is some increased evaporation from the lungs. He has also shown that the secretion of sweat is increased 50 per cent, after the ingestion of food as against an increase of 70 per cent, in the normal individual.

Recent experiments by Schwenkenbecker and Inagaki2 show that the "insensible perspiration" in fever is as great as in health, and that although the urine may decrease in quantity there is no actual accumulation of water in the body as was believed by von Leyden (see p. 522).

Calculations made by Soderstrom3 from the work of Coleman and DuBois show that in typhoid fever, when the body temperature is rising, the heat lost by water vaporization through the skin and lungs bears a lesser relation to the total heat elimination than occurs in normal individuals. On the contrary, a fall in body temperature is accompanied by a relative increase in the water elimination. These experiments confirm the ideas of von Leyden,4 which were published in 1868.

The production of heat in fever may be greatly increased during a chill, and a rapid rise in temperature may follow. This was shown by Liebermeister5 in a case of malaria. The temperature rose from 36.90 in the first half-hour to 39.50 at the end of another hour, while the carbon dioxid expired rose from 14.85 grams to 34.20 grams per half-hour. Barr, Cecil, and Du Bois (unpublished experiments of 1917) have found that during the chill in malaria or the chill following intravenous injection of typhoid vaccine the extra heat produced was retained in the body, causing the sudden rise in body temperature. The amount and manner of heat loss was essentially unchanged from level of the previous normal, suggesting that the heat retention was due to a failure of the vasodilator system to respond normally. The chill ensued even when a patient was surrounded with hot-water-bags at a temperature of 420 C, indicating that the phenomenon was not due to the mechanism of chemical regulation as held by Krehl.1

1 Lang: "Deutsches Archiv fur klinische Medizin," 1904, lxxix, 343. 2 Schwenkenbecker and Inagaki: "Archiv fur ex. Path, und Pharm.," 1906, liv, 168.

3 Soderstrom: Unpublished.

4 von Leyden: "Deutsches Archiv fur klinische Medizin," 1869, v, 273.

5 Liebermeister: Ibid., 1871, viii, 153.

Infectious fevers are characterized by a toxic destruction of body protein. Sometimes, as in the earlier stages of tuberculosis, this tissue destruction may be present in the absence of fever itself. Such a toxic action on body protein is also observed in cancerous cases, as was described by Fr. Muller.2 He writes: "In the 7 cases (of carcinoma) cited, the nitrogen excretion was larger than the nitrogen ingestion, and consequently the body lost protein. In 2 cases the protein loss was no greater than in healthy individuals with similar insufficient nourishment. In all the other cases the protein metabolism was decidedly above that of healthy men under the same conditions. Even an ample dietary was not able to establish nitrogen equilibrium. As more food was given the nitrogen elimination rose higher and higher, but the point of nitrogen equilibrium seemed unattainable." Muller compared the cachexia of carcinoma with that found in febrile processes and believed them to be analogous.