108 PROCEEDINGS OF THE AMERICAN ACADEMY 



struction of future theory. I hope to show convincingly that the sun's 

 " effective " temperature is at any rate far above that of any ordinarily 

 attained in the arts (very much above that of melting platinum for 

 instance), and incidentally that the law of Dulong and Petit is untrust- 

 worthy precisely where we need to apply it. 



If we have no formula by which to infer the temperature of the 

 sun, there remains the comparison of its radiation with that of a ter- 

 restrial source of high known temperature. Thus the late Father 

 Secchi has measured the radiation from the electric arc, and M. Violle 

 that from a Siemens-Martin's furnace ; but, by comparing these only with 

 others made at other times on the sun, discrepant results appear also. 

 Were we, however, to compare the sun directly with a terrestrial 

 source of high temperature, and bringing them face to face find one 

 giving more heat than the other, there could (with equal emissive 

 powers) be no question but that the body radiating more heat was 

 also the higher in temperature. Strange to say, this simple test has 

 never, that I know, been applied to this problem* until in the experi- 

 ments I am about to describe. 



We have in the arts one process which gives what we want ready 

 to hand in the production of a vertically disposed surface of several 

 square feet of a liquid metal, hotter than melted platinum itself. I 

 refer to the Bessemer process now in use in several places in this 

 country, among others at the Edgar Thompson steel-works about 

 twelve miles from Pittsburg. I have received every j>ossible assist- 

 ance from the managers of this great establishment, and owe my 

 acknowledgments here for their kindness. 



As the Bessemer process may be as vaguely known to some as it 

 was till lately to me, I will first briefly describe so much of it as con- 

 cerns the present purpose. 



An enormous egg-shaped vessel called the " Converter," capable of 

 holding 30,000 to 40,000 pounds of melted metal, is swung on trun- 

 nions so that it can be raised by an engine to a vertical position, or 

 lowered so as to pour its contents into a caldron. First, the empty 

 " converter " is inclined, and into its mouth is poured about 15,000 

 pounds of fluid pig-iron, whose temperature as it flows in from an 

 adjacent furnace, where it has previously been melted, is about 1400° C. 

 Then the " converter " is lifted to an erect position, and an air-blast 

 from a powerful blowing-engine is forced up through its liquid con- 



* Experiments with the lime and electric lights made for other purposes are 

 not here in question. 



