W. Ferret — Laic of Thermal Radiation. 25 



former values from formulae which are not supposed to hold 

 accurately down to so low a temperature. 



It is seen from the computed values of E from the several 

 expressions of E, as well as from the expressions themselves, 

 that these values increase very rapidly with increase of temper- 

 ature. Also, that while the values of m, as we have seen, as 

 determined from observation for the several laws and expres- 

 sions, are very scattering and uncertain, the values of E, as 

 determined from the several very different expressions, differ 

 but little. 



23. The several preceding laws of radiation here investi- 

 gated pertain to the radiations of a lampblack or a bare glass 

 surface, between which there seems to be but little, but of 

 course some, difference in the rate with which radiation in- 

 creases with increase of temperature. According to Yiolle* 

 this rate is much greater for the short than the long wave- 

 lengths, and the same is to be inferred from Langley's results, f 

 in which it is seen that the maximum intensity with reference 

 to the wave-lengths is thrown toward the end of the shorter 

 wave-lengths in the spectrum as the temperature of the radiat- 

 ing body is increased. The law of radiation, therefore, for the 

 resultant radiation of all wave-lengths must differ very much 

 in different bodies in which the radiativities differ considera- 

 bly from that of a surface of maximum radiativity, according 

 as the predominating wave-lengths in its radiations are toward 

 the one or the other end of the spectrum. The radiativity of 

 bare glass at ordinary temperatures is supposed to be one-tenth 

 or more less than that of a lampblack surface, but if the de- 

 ficiency in its radiation at these temperatures is mostly toward 

 the end of the longer wave-lengths, then it is readily seen that 

 for higher temperatures its relative radiativity, that is, radia- 

 tivity relative to lampblack, must increase and gradually ap- 

 proximate that of the latter, but it could never quite become 

 the same, as it is according to Lehnebach's experiment at the 

 temperature of 100°. For this reason, also, the absolute radia- 

 tivity in glass must increase with increase of temperature a 

 little faster than that of lampblack, as seems to be indicated 

 by the larger numerical coefficient of z in the expression of e 

 in (21) than in (23). 



24. The condition from which the temperature of the sun is 

 determined where the law of radiation is known is 



(34) H =7- A 



4 oo 



in which S : co is the ratio between the whole surface of a 

 sphere and the part which subtends the same solid angle as 



* Comptes Eend., vol. xcii, p. 1204, 1881. 

 f This Journal, vol. xxxi, January, 1886. 



