98 



Professor H. L. Callendar 



[March 10, 



justifiable method, and may lead to very good results if the empirical 

 law happens to be correct; but if the formula happens to be unsuit- 

 able, it may lead to the most remarkable conclusions. 



The curves shown in Fig. 1 illustrate some of the typical formulae 

 which have either been proposed for the law of radiation, or been 



LAW or RADIATION Bottomley.88*57r»schen 



SCALE OF TCMP. 



Fig. 1. — Formulae of Radiation. Experimental range. 



deduced from the results of modern experiments over the experi- 

 mental range of the gas-thermometer, extending to 1200° C, to which 

 trustworthy determinations of temperature on the theoretical scale are 

 at present restricted. In order to obtain a comparison of the formulae 

 themselves, apart from other issues, the results of different observers 

 are reduced to a common hypothetical value, 10 watts per square 

 centimetre, for the radiation from a black body at 1000° C. 



Excluding the law of Newton, which applies only to small 

 differences of temperature, and also the law of Dulong and Petit, 

 which was founded on observations over a very limited range with 

 mercury thermometers, and is obviously inapplicable at high tempe- 

 ratures, there is a certain family resemblance between the remaining 

 curves ; but the differences between them are still so considerable that, 

 if sufficiently accurate measurements of temperature were available, 

 it should be possible to decide with certainty which of the formulae 

 was the most correct. A fairly close agreement is seen to obtain 

 between the formula proposed by Weber and the curves represent- 

 ing the results of the recent experiments of Bottomley, Paschen and 

 Petavel. But, on the other hand, there is strong evidence, both 

 experimental and theoretical, in favour of the fourth power law 

 proposed by Stefan, which differs materially from that of Weber ; and 

 many supporters may be found, especially among astronomers, for the 

 very different formula of Rosetti. 



