OP ARTS AND SCIENCES. 107 



not materially different from 4,000,000° Fahrenheit ; Father Secchi, 

 in his latest research, makes it 133,000° C. ; Sir Wm. Thompson and 

 others estimate 30,000 to 60,000° C. 



These extremely gross discrepancies having drawn general atten- 

 tion, many distinguished French physicists have lately reinvestigated 

 the subject, and, using Dulong's and Petit's formula, have after most 

 elaborate research arrived at the nearly unanimous conclusion that 

 the temperature of the solar surface is altogether lower than any of 

 these, — is in any case not more than 2000 to 2500° C, but is more 

 probably below than above the temperatures which are reached in our 

 furnaces, and in fact is probably less than that of melting platinum. 



It is here to be borne in mind that we really know nothing about 

 the absolute emissive capacity of the solar surface, and that to sim- 

 plify the problem, when we speak of the sun's being at a lower tem- 

 perature than that of a certain lamp-black surface or hot platinum or 

 steel, it is assumed for the purpose of comparison, by myself as well 

 as by the above-named investigators, that both the solar and terrestrial 

 sources of heat have the same emissive capacity. The temperature 

 thus defined has been called the " effective " temperature. 



M. Violle, one of the most distinguished students of the subject, 

 whose experiments bear evidence of intelligent care, found by obser- 

 vations at Grenoble, in March, 1874, that, with an emissive power 

 thus defined, the temperature of the solar surface was 1230° C* In 

 a subsequent memoir, he finds for the same the rather higher value 

 of 1354° C.f After allowing for absorption in our atmosphere, it 

 remains true that the temperature is then much below that of melting 

 platinum, and this seems to be confirmed by his later results, which 

 give about 1550° C. as the highest "effective" temperature. 



All these and other observations involve the use of the empirical 

 formula, well known as that of Dulong and Petit, which has replaced 

 the earlier and simpler one of Newton. 



Now, whatever be the apparent presumption of opposing my opin- 

 ion to that of so many conscientious and recent investigators, I feel 

 there is something yet to be said; and the present paper is an account 

 of experiments of a special character, undertaken at the expense of 

 the Rumford fund, not with the hope of at once solving so arduous 

 a problem, but with the wish, in this confusion of opinion, to contrib- 

 ute one or two incontrovertible facts as material towards the con- 



* Comptes Rendus, vol. lxxviii. p. 1425. 

 t C R. vol. lxxviii. p. 1816. 



