of dark Heat-rays by Gases and Vapours. 3 



Without entering into details, we may remark that Tyndall 

 has shown that, with the source of heat in the experimental 

 space, currents of air and conduction of heat are unavoidable; 

 and this source of error is obvious. Moreover, so far as the 

 literature of the subject is known to us, Magnus does not 

 return to this question; so that apparently he has himself 

 recognized this source of error. 



Buff endeavoured to avoid this error by rapid heating of 

 the surface which serves as source of heat ; but the duration of 

 the experiment is always much too great, and, especially, cur- 

 rents of air and conduction are not avoided. The most striking 

 proof of this is found in Buff's tables, since the greater deflec- 

 tion always takes place after the heating has been interrupted 

 for some time, which could not have been the case if the heat 

 radiated by the heated surface had been the sole cause of the 

 deflection, as is required by the conditions of the experi- 

 ment. 



Garibaldi moreover employed a concave mirror to concen- 

 trate the heat-rays ; and we suspect that in this is to be found 

 the reason of the incredible result which he obtained of an 

 absorption of 92 per cent, by aqueous vapour. 



Tyndall*, as is well known, first employed the second 

 method. Since the space containing the gas is closed and 

 yet must be open to heat- rays, the tube employed must be 

 closed on both sides by diathermanous substances. Tyndall 

 employed plates of rock-salt. Dr. Franz f had previously 

 employed glass plates, which, in consequence of the great 

 absorption which glass exerts upon dark heat-rays, must 

 clearly lead to very fallacious results. Tyndall's method was 

 also employed by Wild X in determining the absorption of 

 aqueous vapour ; but since the arrangement was precisely the 

 same as Tyndall's, the results must be open to the same 

 objection. 



Since the perfect diathermancy of the plates which close the 

 tube can never be attained, there must always remain a source 

 of error in this arrangement. And even if the original objec- 

 tion of Buff's, that rock-salt and air are of similar colour for 

 heaWays, is not confirmed, it is at any rate clear from Tyndall's 

 own researches § that rock-salt absorbs 16 per cent, of the ra- 

 diation from a lampblack surface, and indeed, as he expressly 

 maintains, in opposition to Melloni and Knoblauch, that rock- 

 salt does not transmit all rays in equal proportion. 



Buff has maintained that Tyndall found no absorption pro- 

 duced by air, simply because air and rock-salt absorb the 

 * Loc. cit. Memoir I. t Pogg\ Ann. xciv. 



X Ibid, cxxix. § Loc. cit. p. 322. 



B2 



