NO. II BODY RADIATION ALDRICII 5 



The instrument has recently been recalibrated. In mounting for 

 calibration and comparison, the constant temperature junction was 

 fastened against the bulb of a mercury thermometer. The ther- 

 mometer was then inserted in a metal tube and lowered into a stirred 

 kerosene bath, surrounded by a vacuum flask. In calibrating, the 

 thermoelement device was placed in a well-stirred kerosene bath whose 

 temperature was measured with a second mercury thermometer. A 

 sensitive-type Leeds and Northrup D'Arsonval galvanometer was used 

 with the instrument. The calibration curve, plotting galvanometer 

 deflections against temperature differences, is nearly a straight line. 



As certain systematic discrepancies had been noted between skin 

 temperatures observed with this thermoelement and corresponding 

 ones computed from observations of body radiation with the me- 

 likeron, it was desired to ascertain whether the thermoelement was in 

 any considerable degree influenced by air temperatures in making such 

 measurements. The instrument was accordingly tested in the follow- 

 ing manner by measurements on a skinlike membrane of known 

 temperature. 



In the vertical copper calorimeter previously used (see Smithsonian 

 Misc. Coll. vol. 8i, no. 6, p. 15) three holes were made in the side at 

 equal altitude, each 6 cm. in diameter. These holes were closed with 

 rubber diaphragms, cemented in with waterproof cement. The thick- 

 ness of the diaphragms was as follows (determined with micrometer 

 gauge) : 



0.18 mm. (thinnest dental dam) 



0.36 mm. (sheet rubber) 



1.20 mm. (composite sheet rubber used for gaskets) 



Rubber was chosen because it is pliable, simulating the surface pre- 

 sented by the skin or clothing. The calorimeter as before was filled 

 with water kept thoroughly stirred and a record of its temperature 

 determined by a mercury thermometer. 



It is evident that the surface of the thickest diaphragm will be 

 appreciably lower in temperature than the water in the calorimeter, 

 and that the thinner the diaphragm the more closely the surface tem- 

 perature approaches the temperature of the water. By obtaining a 

 series of surface temperatures of the various diaphragms, a curve may 

 be plotted and extrapolated to zero thickness. The more nearly cor- 

 rect the thermoelement temperatures, the more closely the zero dia- 

 phragm value will approach the calorimeter water temperature. 



A series of comparisons is summarized in Table i. Each value in 

 the table is the mean of three separate determinations. Air motion was 

 produced by a fan in the same manner as in the body-radiation experi- 



