G MM. Lecher and Pernter on the Absorption 



Tin's last number (180) is interpolated from Tyndall's expe- 

 riments, and intentionally taken as unfavourably for our case 

 as possible — that is to say, as small as possible. In order that 

 sulphuric ether should stand before formic ether, the deflection 

 of the galvanometer must have fallen from 180 below 133. If 

 we take the total radiation as (unfavourably) small as possible 

 (say at 300), then we find that even with an absorption of 50 

 per cent, the result may vary about 30 per cent, of its value, 

 accordingly as we employ blackened or polished tubes. 



The numbers given above are values actually measured, and 

 moreover with the full accuracy of which the method is capable. 



Tyndall believes that the reversal of the order may perhaps 

 be explained by the impurity of the substances employed with 

 the blackened tube; but it is clear that we have nothing to do 

 with impure substances and their absorption. But that this 

 explanation is not correct is seen from later experiments, where 

 Tyndall employs glass tubes instead of tubes of brass. Now, 

 since glass is " much weaker in reflecting-power than brass" *, 

 the less-powerfully reflecting glass tube ought to behave nearly 

 as a partly blackened metallic tube. And, in fact, Table III. 

 of the second memoir gives exactly the same order for the 

 quantities of heat absorbed as we have previously quoted for 

 the half-blackened metallic tube. 



Exact calculations in this direction are not possible, since 

 unfortunately Tyndall only very seldom gives the total radia- 

 tion. We believe, then, that ice have shown, from TyndalVs own 

 numbers^, that the absorption of radiant heat found ichen a 

 polished metallic tube is employed is different from that obtained 

 when partly-blackened tubes, or tubes of glass are employed. It 

 follows, then, at once that neither of the above cases yields 

 correct results, since in both reflection takes place. 



How important a source of error vapour-adhesion was, can 

 be seen from a closer examination of Tyndall's method of 

 measuring the vapour admitted into the tube. The arrange- 

 ment to fill the large experimental tube with vapour was 

 shortly as follows : — A glass tube was cemented onto the ex- 

 perimental tube, which carried a small chamber (which we will 

 call M) closed by two taps. One tap allowed communication 

 to be made with the large tube, the other with a space in which 

 the fluid to be experimented upon was contained without air. 

 Evidently, when this tap is open and the upper one closed, M 



be a misprint (see Phil. Mag. (A) xxii. p. Y.)2, where the right number is 

 found). Jt should he 157; for this value corresponds to a deflection of G0° 

 (see Tvndall, p. 29). Also p. L57: "For the usual pressure of 0-5 inch, 

 alcohol absorbs exactly twice as much heat as benzol." Benzol = 78 

 (Table XII. p. 31), and 2x78 = lo0. 

 * Loc. eit. p. 71. -f Loc. cit. p. 78. 



