1899.] On Measuring Extreme Temperatures. U7 



WEEKLY EVENING MEETING, 



Friday, March 10, 1899. 



The Hon. Sir James Stirling, M.A. LL.D., Vice-President, 

 in the Chair. 



Professor H. L. Callendar, M.A. F.E.S. 



Measuring Extreme Temperatures. 



The measurement of extreme temperatures is a subject of groat 

 theoretical interest, especially in connection with the determination 

 of the laws of radiation and of chemical dissociation and combination. 

 The temperature in each case is the factor of paramount importance, 

 and without means of measuring the temperature there is no pos- 

 sibility of formulating any rational theories. The subject possesses, 

 in addition, a powerful fascination for the experimentalist, on account 

 of the difficulty of the observations involved, and of the extremely 

 conflicting nature of the results obtained by different observers and 

 different methods of research. 



Attempts have frequently been made to estimate the temperatures 

 of the electric arc and of the sun, which may be taken as examples of 

 the most extreme temperatures known to science, and afford an 

 illustration of the difficulties to be encountered, and of the methods 

 available for attacking these problems. A brief consideration and 

 comparison of the results will also serve to explain the causes of the 

 remarkable discrepancies existing in the estimates of such temperatures 

 by different observers and different methods. 



In the case of the sun it is at once obvious that no terrestrial 

 thermometer can possibly be directly applied. The only available 

 method is (1) to measure the intensity of the solar radiation, and 

 (2) to endeavour to deduce the temperature by determining the law of 

 radiation at high temperatures. The measurement of the intensity 

 of the solar radiation is in itself a sufficiently intricate problem, 

 containing many elements of doubt and difficulty ; but by far the 

 greatest source of uncertainty lies in the solution of the second part 

 of the investigation, the determination of the law of radiation. The 

 origin of the discrepancies thus imported into the results may be 

 summed up in the word " Extrapolation." 



The method of investigation necessarily consists in taking a series 

 of observations at temperatures within the laboratory range of ther- 

 mometry, from which to calculate an empirical formula representing 

 as closely as possible the results of experiment. It is then assumed 

 that the formula may be " extrapolated," or used to estimate the 

 temperature of a radiating source of known intensity beyond the range 

 of the observations on which it was founded. This is a perfectly 



Vol. XVI. (No. 93.) h 



