258 PHYSICS. 



the facts better than that of Dulong and Petit. From this, results were 

 calculated for temperatures beyond 300° and up to 2000°, which agreed 

 well with those derived from experiment. The apparatus was then ap- 

 plied to the determination of the solar temperature. The thermopile 

 was placed where the direct rays of the sun could fall normally upon its 

 face. Four readings of the galvanometer were made during each obser- 

 vation : the first with open circuit to fix the zero point ; the second, with 

 closed circuit, the case of the pile being closed ; the third, after opening 

 the case to allow the heat rays to act; the fourth, with the case ojien and 

 the solar rays intercepted by a double-walled screen placed one meter off. 

 This last reading was generally negative, the iiile radiating heat into 

 space. Calculated from these results it appeared that the thermic effect 

 which wouhl be produced by the solar radiation falling normally on the 

 face of the pile, if the atmosphere did not exist, or if the pile were above 

 it, is represented by 323 scale divisions. Since resistance was used in the 

 circuit in the last experiments, this value must be multiplied by 5.G921 

 to make it comparable to the others ; this giv^es 1838.5 divisions. Calling 

 the effective temperature of the sun, that which an incandescent body 

 of the same size placed at the same distance should have to produce the 

 same effect, its emissive jjower being at the maximum, the formula gives 

 as the absolute effective temperature of the sun 10238O.4, or in degrees 

 centigrade, 99G5.4. The author concludes that if the absorption of our 

 atmosphere only be taken into the account, the true temperature of the sun 

 cannot be far inferior to ten thousand degrees. If the absorption pro- 

 duced by the solar atmosphere be also considered, and Secchi's estimate 

 of .88 be allowed for it, then this temperature is higher but not very far 

 superior to twenty thousand degrees. — {A7in. Chim. Fhys., Y, x^i, 177, > 

 June, 1879.) 



Langley has compared the intensity of the solar-heat radiation with 1 

 that of the incandescent surface of Bessemer steel in fusion as it flows 

 from the converter. The temperature of this molten metal is above that 

 at which platinum melts, a wire of this metal fusing in the gas stream 

 issuing from the mouth of the converting vessel ; hence it is from 1800° 

 to 2000° C. The heat radiated from the liquid stream fell on one face of 

 a thermopile, while a bundle of solar rays was reflected on to the other 

 face. Though the apparant diameter of the opening of the converter 

 was many times that of the sun, so much greater was the energy of the 

 latter that a lens was interposed to form a magnified image of the sun 

 on the pile, and it was diaphragmed to bring the needle to zero. Calcu- 

 lating for equal diameters, the solar energy w as found to be 86 times that 

 of the molten metal at least. Photometric comparison showed the solar- 

 light radiation to be 5,300 times that of the metal. Crova, in noting 

 these experiments, calls attention to the rapid increase of light emitted 

 by melted silver as the temperature rises, being 147 times greater at 

 1200O than at 916^, and hence 28,900 times at 1500°, as evidence that the 

 sun's temperature is quite moderate compared with some of the calcu- 

 lations made on the subject. — [J. Phys.., ix, 59, February, 1880.) 



I 



