531. 



REPORT — 1902. 



be --0000069. No absolute measurements were made, but tbe value obtained by 

 Regnault was assumed. 



The formula proposed by "Winhelmanu is shown in the figure (W). This reads : 



S, = S„ - -0000060 t. 



The remarkable a5:reement with the average change from our determinations 

 is sbowji Tei-y cleaily in the llirure, a total deciea.se of 2 per cent, between 

 0° and 100° being given both by A\'inkehnanu's results and our own. 



The position of the curve in tbe figure shows that the temperature at which 

 Winkelmann ast'umed IJegnauh/s value, which in one casewas the mean from 1(3° to 

 20°, and in the second case I'rom \'-i° to 25°, is very nearly equal to the tempera- 

 ture at which our value is exactly equal to Kejrnault's, i.e., \.°-\ C. 



Pettersson and Hedelius ' oljtauied 0;J.'j2()6 between and •>° , -033202 between 

 5° and ]H°, 03:3800 between 5° and 26°, and -033290 between 5° and 36° C. These 

 results were all expressed in terms of a thermal unit for water at 0°. 



The measurements were repeated more accurately by J. Miltbaler,- who used a 

 similar method to tbe previous investigators, i.e., the method oC mixture. This 

 author verified Winkelmann's results and showed that the specific heat apparently 

 decreased with rise of temperature, tbe average decrease from his measurements 

 being -0000092, which is somewhat, greater than AVinkelmaun obtained. 



'ihe formula proposed by Milthaler is shown in the figure (M). The slope of 

 the curve is about tlie same as the slope of our curve at the low temperature ; 

 but since Milthaler supposed this formula to apply as high as 200°, its agreement 

 for the low temperatures is more a matter of coincidence. The position of the line 

 in tbe figure showt* that the value of l'e1ters.son and Hedelius wbich he assumed 

 is too small at 0° compared to our values, but this value will be increased when 

 corrected to a thermal unit for water at 15°-5 instead of 0°, and will then be brought 

 into much closer aiireement. 



2. Some Experiments on Radiation and Absorption : a Preliminary Study 

 for a Standard of Light. By J. E. Petavel. 



The first jiart of the paper refers to some determinations of the intrinsic 

 brilliancy of iridium and platinum-iridinm at their temperature of fusion. 



It has been previously shown' that if t be the temperature centigrade and b 

 the intrinsic brilliancy in candle-power per square centimetre 



Fonmda I. 

 ^ = 400 + 880-G''^^» 



The photometric observations and calculated temperatures of fusion are given 

 in tbe following table: — 



The temperature of a radiating body may also be estimated by a second 

 method. 



' Jour, fur 'prah. Cheiii.. vol. xxiv. p. 135 (1881). 

 - Wied. Ann., vol. xxsvi. p. 807 (1889). 

 ' See Phil. Trans., cxci. 501. 



