156 Intelligence and Miscellaneous Articles. 



same as that of the sun, I obtain an excess of temperature t—B of 

 0°*5 only. The temperature of the disk is at least equal to, and 

 probably exceeds, that at which platinum fuses ; it may therefore 

 be estimated at 1900 or 2000 degrees. 



Father Secchi, a few months ago, communicated to the Academy 

 a similar experiment, made with the electric light, the temperature 

 of which he estimates at 3000°; and he obtained an intensity of ra- 

 diation from 34 to 36 times less than that of the sun. I am quite 

 disposed to admit the accuracy of this experiment • but one might 

 perhaps object that there is sufficient uncertainty about the tempe- 

 rature of the electric lamp, and that it is not quite certain that the 

 whole extent of the carbon visible to the thermometer is equally 

 heated. I therefore dwell on my own experiment only, against 

 which those objections cannot be urged with the same force. 



If Newton's law were strictly accurate, to calculate the tempera- 

 ture T of the disk heated at the oxyhydric lamp, we should have 

 the formula 



*-0=aT, 



a being equal to y^qqq' Now 



t-e=o°'5, 



whence 



T=91980°, 



an absolutely inadmissible result ; for it is certain that the tempe- 

 rature of the disk is near 2000°, and at any rate does not exceed 

 2800°. Newton's law, then, is not verified. 



The Rev. Father Secchi, in discussing either some of my experi- 

 ments or his own experiment, infers for the sun a temperature of 

 more than 100,000 degrees. His reasoning maybe summed up by 

 saying that, since the intensity of the solar radiation is 44 times 

 that of the electric light, the temperature of the sun must' also be 

 44 times that of the lamp. This conclusion does not appear admis- 

 sible • for it is equivalent to supposing that Newton's law is accu- 

 rate from 2000° or 3000° upwards ; while experiment manifestly 

 proves that it is inaccurate between zero and 2000°. 



Let us pass to the lav/ of Dulong and Petit. From this law M. 

 Vicaire has deduced, for the case we are considering, the formula 



in which a== — — — — as before, and « = 1*0077. From this we 

 183960 



get 



T 



log (a* — a 9 ) + log - 



log a 



Now, if we apply this formula to my experiment, in which 

 teifi°-45 and = 14°-95, we find T = 870, a figure evidently too 

 low, as Father Secchi has already remarked. And if the converse 

 calculation be made by seeking the value of t for a temperature 



