If L is the coefficient of induction per turn of primary, the total induction is

for N2/2 turns per linear centimeter, and the resistance of this coil and its circuit, we may compare any two coils by the equation

e2 and e'2 being the electromotive forces, and is and i'2 the currents induced in the two secondaries in question. Thus

it being assumed that the resistance r in the secondary circuit is made so large that the inductive resistance vanishes.

The coil tester was now thrust through a variety of helices, differing in shape and construction.

Figure 7 gives the results read off for these coils at the vibrator (s), when a high resistance was added to the circuit and Ri constant. The relation of Ho and s as seen in figure 7, is linear for coils W, B, G, which were about of the same length and wound on non-conductors or split brass. The result for coil D, which was about 1.8 times longer is low, probably owing to mutual induction, as this coil was wound on a thick brass tube without fissure. Similarly it made very little difference whether the two helices wound side by side throughout G, were used in parallel or in series. Finally a number of single layer coils of lengths 10, 20, 30, 40 cm. were wound on stout glass tubing and compared with B. Provided care was taken to prevent the induced currents from heating the telephone, the band width s increased throughout proportionally to the number of turns of wire in the secondary and under good conditions at a rate of about s = 0.06 scale parts per turn of wire.

THE TEMPERATURE OF THE HUMAN SKIN

BY F. G. BENEDICT, W. R. MILES, AND ALICE JOHNSON NUTRITION LABOratory, CarnegiE INSTITUTION OF WASHINGTON, BOSTON, MASSACHUSETTS Read before the Academy, April 28, 1919

The clinical importance of records of body-temperature, as taken usually in the mouth and occasionally in the axilla or rectum, has quite obscured the physiological significance of the skin temperature. Extensive researches have shown that the temperature of the human body, deep in the body trunk or in any of the natural cavities, remains reasonably constant, although there is a diurnal ryhthm, with a minimum value at about 4 a.m. and a maximum at about 5 p.m. Simultaneous observations of temperature at different parts of the body show that there is almost always a parallelism in curves for temperature. Thus, the temperatures in the rectum or vagina, the well-closed groin or axilla, and the mouth, show similarly shaped curves, although at markedly different levels.

The technical difficulties in recording skin temperature have undoubtedly retarded extensive study in this phase of body-temperature research. The temperature of the human skin is the resultant of several factors, as heat is supplied from the subcutaneous tissues and lost from the surface of the body by radiation, conduction, and the vaporization of water. An attempt to secure accurate records of skin temperature by the application of an ordinary mercury thermometer is obviously useless, for but a small proportion of the bulb of such a thermometer comes in actual contact with the skin. Even thermometers constructed with a special bulb providing a large surface to apply to the skin have a like surface exposed to the environmental temperature. If, as is occasionally done, this outer surface is covered with non-conductive material, there is almost immediately a disturbance in the temperature of the skin due to the fact that there is a retardation of the normal loss of heat, with a conse

quent accumulation of heat from the subcutaneous tissue. The true tempera-}

ture of the skin should therefore be recorded by an apparatus which is nearly instantaneous in action and sufficiently protected from the environment to insure a true record of the surface temperature, and not a resultant of skin and environmental temperature.

A method recently used at the Nutrition Laboratory consists of two copperconstantan junctions, one located in a constant-temperature bath-a Dewar flask—and the other applied to the skin. The resulting current which can be measured on any one of several types of galvanometers is directly proportional to the temperature difference between the two junctions. All thermometric lag is thereby eliminated and it is only necessary that the junction which is applied to the body should be properly protected from the environmental temperature. It was found that when this junction was backed with

a light tuft of cotton batting and installed rigidly in a piece of hard rubber, it was possible to apply it to the body and have it assume the surface temperature inside of a few seconds. Thus the period of application was so short that the protecting material did not sensibly affect radiation, conduction, or the vaporization of water.

In our ordinary skin temperature measurements, the junction in the constant temperature bath is kept at 31° or 32°C. When the other junction is exposed to usual room temperature, there is a large deflection of the galvanometer. On applying the junction to the surface of the skin, the amplitude of deflection decreases rapidly and with negative acceleration. The deflection level is reached in about six seconds after application of the junction. Thereafter the rate of movement of the galvanometer is extremely slow, being but a millimeter or two for each successive minute. The explanation of this is that the junction almost instantly assumes the true skin temperature. Thereafter the skin temperature gradually rises, as the protection of the junction with cotton batting and hard rubber stops the loss of heat by radiation, conduction, and vaporization of water, and there is then a 'building up' of the temperature due to the subcutaneous source of heat. As a matter of fact, observations taken for some time show that this rising temperature continues, the length of time depending in large part upon the amount of hard rubber and cotton batting used. For all of our work we have assumed that under these conditions the true skin temperature is obtained at the end of about six seconds after application of the junction.

In connection with an extended series of metabolism experiments on the influence of temperature environment upon the metabolism, it was found that the subject used (a professional artist's model) could withstand exposure to temperatures as low as 14°C. for several hours without shivering. These conditions presented unusual opportunities for studies of skin temperature which would give evidence first, as to the rapidity of the change on exposure of the body to the environmental temperature, and finally as to the absolute level which the skin temperature reaches after prolonged exposure to an environment varying from 14° to approximately 30°C.

When the subject arrived at the Laboratory, the clothing was loosened and temperature measurements were made at numerous points on the skin under the clothing. Prior to these records, the subject clothed and wearing a heavy coat had been exposed to an external temperature of about 17°C. The results of one series of measurements are recorded in table 1. The extreme range in the skin temperature on this particular day was 6.6°C. Special attention is called to the relatively high value of 31.6°C. on the exposed forehead, the record of approximately 34°C. at the waist, a part well protected by clothing, and values of 30°C. or below on the buttocks, shin, and calf. This series is typical of the skin temperature of a normally-clothed woman.

The method of using isolated points, such as those given in table 1, can be much improved upon by taking advantage of the fact that the thermal junction, when not excessively protected by the cotton batting and hard rubber, almost instantaneously assumes the skin temperature. This is particularly the case if the junction is moved from one part of the skin to another after the hard rubber and cotton have assumed approximately the temperature of the skin. When recording the values of isolated points, we have employed a good grade, direct reading galvanometer, thus making the apparatus compact and portable. The response of the thermal junction is so rapid, however, that when connected with a sensitive aperiodic instrument, i.e., a string galvanometer, it is perfectly possible to move the junction at a moderately rapid rate over the sur

face of the body and record photographically a continuous and true temperature curve for an infinite number of points on the skin.

To illustrate our many results obtained by this latter method, two typical curves are given in figure 1. These were obtained after the nude subject had been exposed to an environmental temperature of 14.6°C. for two and onehalf hours, the greater part of the time in the standing position. The upper curve follows the left mammillary line, while the lower curve was taken in a corresponding position on the back. An extraordinary difference in skin temperature at the different points may be noted, the extreme range on the front and back showing differences of 10°C. or more. The rectal temperature taken simultaneously was 36.7°C.

The pronounced influence of the environmental temperature upon the curve for the skin temperature was especially studied by making temperature photographs over exactly the same line at intervals from the moment of disrobing

Photographic records from a string galvanometer for skin temperatures at different parts of the nude body following a 2 hour exposure to an environmental temperature of 14.6°C. The thermal junction was moved at constant tempo from shoulder to ankle. Time regIstered at bottom of each curve is in 2-second intervals.

to the end of the period of prolonged exposure to cold. As would be expected, these show a progressive increase in the amplitude of the curve as the time of exposure is prolonged; marked changes occurring even inside of the first fif