Thermal Radiation in Absolute Measure. 549 



it was on being taken from the hank ; the other was sooted. 

 The two wires were then fixed in the glass tubes. The wires 



are designated Pt^Pto, Pt 3 ,Pt 4 , Pt 5 ,Pt 6 . The first of each 

 pair is the bright wire ; the second is the sooted wire. The 

 diameters of the wires are as follows : — P^ and Pt 2 , 0*0542 

 centim., Pt 3 and Pt 4 , 0*025 centim. ; Pt 5 and Pt 6 , 0*015 centim. 

 In Tables I. and II. specimens are given of the results 

 obtained, in the manner described above, by observation and 

 calculation. The remainder of the results are embodied in 

 the curves appended, which it is hoped will be found self- 

 explanatory. At the head of each table the particulars as to 

 the wires referred to in the table are given. 



In the following tables, III., IV., V., the loss of heat per 

 square centimetre of surface for the several pairs of wires, 

 bright and sooted, at various temperatures, is compared; and 

 the ratio between the radiation from the sooted wire and the 

 radiation from the bright wire is calculated. It will be seen 

 that the numbers are in fair agreement. What may be the 

 causes of divergence from exact agreement it is impossible to 

 say at the present stage of the inquiry; but it may be con- 

 jectured that part of it at least is due to the difficulty, or 

 impossibility, of keeping the vacuum which surrounds the 

 wires in these experiments unchanged. When the pressure is 

 very low, the accession of the smallest quantity of gas to the 

 surrounding space causes an enormous change in the rate of loss 

 of heat, as has been shown in a previous part of this research, 

 and as the temperature rises it is always found that the vacuum 

 becomes deteriorated, owing to the expulsion of gas from the 

 body of the wire itself. This gas must be removed by a fresh 

 application of the pump, and, in fact, during the experiments 

 the pump must be kept always at work. Thus the vacuum is 

 incessantly changing; and, moreover, as the indications of 

 the McLeod gauge lag very much behind, it is not even 

 possible to know the exact state of the pressure at the instant 

 when it is desired to make an observation as to current 

 passing and resistance. The consequence is that owing to 

 the constantly altering state of the vacuum an irregularity 

 is introduced in the loss of heat, and the irregularity tells 

 more in the case of small wires than in the case of larger 

 sizes. 



In the case of the bright wire, Pt 5 , the loss of heat was 

 somewhat abnormal. It is probable that the surface was 

 lacking in polish. 



It will be seen that the loss from the sooted platinum wires 



