174 



NATURE 



[Dec. 27, 1877 



refrangibility, on botk sides 'of which it gradually becomes trans- 

 parent again. ^ Imagine a spectrum containing radiations of all 

 refrangibilities with which we have to deal ; let ponions of th's 

 spectrum on the two sides of the region of powerful absorption for 

 glass be called ivings of that region, and let left to right be the 

 order of increasing refrangibility. Then the spectrum of the 

 radiation from a thin plate of glass, if it could be observed, would 

 be seen to occupy the region of chief absorbing (and therefore 

 emitting) power and its wings. The spectrum of the radiation 

 from the outer stratum of the bulb of the pith radiometer, after 

 transmission through the inner, would consist of two wings, with 

 a blank, or nearly blank, space between ; it would resemble, in 

 fact, a widened bright spectral line, with a dark band of reversal 

 in its middle, save that, instead of being confined to extremely 

 narrow limits of refrangibility, the central space and its wings 

 would be of wide extent. It follows from the experiments that, 

 in the complete radiation from glass, the portions of the spectrum 

 called the wings together act negatively, the portion between 

 positively. It does not, of course, follow that each wing acts 

 negatively, but only that the balance of the two is negative. 

 When the tumbler is heated a little over 2 12° there is a shght 

 positive action from radiation which passes directly through the 

 bulb. The circumstances lead us to regard this as an extension 

 of the right wing ; for it comes from a depth, measured from the 

 inner surface of the bulb in glass, i. e, , not counting the interven- 

 ing air, somewhat greater than the thickness of the wall of the 

 bulb ; and we know that the more a solid body is heated, the 

 higher, as a rule, does the refrangibility of the radiation which it 

 emits extend, and the greater the proportion of rays of high to 

 those of low refrangibility. It is simplest, therefore, to suppose 

 that the action of the right wing, like that of the space between 

 the wings, is positive, and that the observed negative action in 

 the experiment of § 7 is due to the excess of negative action of 

 the left wing over positive action of the right. In the mica 

 radiometer the experiments indicate no such difference of action 

 in the different layers of the bulb as in the case of the pith radio- 

 meter. Hence taking, in accordance with what now appears to 

 be made out to be the theory of the motion of the radiometer, 

 the direction in which the fly is impelled as an indication which 

 is the warmer of the two faces of the discs, and that again as an 

 indication which is the darker with respect to the radiation to 

 which it is exposed, we arrive at the following results as regards 

 the order of darkness of the substances for the three regions into 

 which the spectrum of the incident radiation has been supposed 

 to be divided, the name of the lighter substance being in each 

 case placed above that of the darker : — 



Region of intense 



absorption by Right wing. 



glass. 



Pith. Pith. 



Lampblack. Lampblack. 



Lampblack. Mica. 



Mica. Lampblack. 



Hence, on descending in refrangibility, the order of darkness 

 of the two substances o either pair is at first the same as for the 

 visible spectrum, and at last the opposite ; and the reversal of 

 the order takes place sooner with mica and lampblack than with 

 pith and lampblack. The order falls in very well with that of 

 the chemical complexity of the three substances. 



20. The whole subject of the behaviour of bodies with respect 

 to radiant heat of the lowest degrees of refrangibility seems to 

 me to need a thorough experimental investigation. The inves- 

 tigation, however, is one involving considerable difficulty. We 

 can do little towards classifying the rays with which we are 

 working unless we can form a pure spectrum. A refraction 

 spectrum is the most convenient ; but the only substance known 

 which would be approximately suitable for forming the prism, 

 lens, &c., required for such a spectrum, and for confining 

 liquids, is rock-salt, of which it is extremely difficult to procure 

 perfectly limpid specimens of any size ; and even rock-salt 

 itself, as Prof. Balfour Stewart has shown, is defective in trans- 

 parency for certain kinds of radiant heat. Then, again, the only 

 suitable measuring-instrument for such researches, the ther- 

 mopile, demands a thorough examination with reference to the 

 coating to be employed for absorbing the incident radiation. 

 Hitherto lampblack has been used almost exclusively for the 

 purpose ; and it is commonly assumed, in accordance with cer- 

 tain of Melloni's results, that lamplack absorbs equally heat- 



' It may be noticed that this supposition, which, ss appearing the more 

 probable, is adopted for clearness of conception, is not essentially involved 

 in the explanation that follows, which would hardly be changed if the " left 

 wing " were not terminated on the left. 



Left wing. 

 From pith radiometer ...{Lampblack. 



From mica radiometer ...|^anipblack. 

 I Mica. 



rays of all kinds. But the experiments by which Melloni 

 established the partial diathermancy of lampblack prove that 

 rays exist for the absorption of which that substance is un- 

 suitable. 



On calling on Mr. Crookes after the above was written, I was 

 surprised to find that all his mica radiometers behaved towards 

 a heated glass shade in the opposite way to that he had given 

 me, going round positively instead of negatively. Mr. Crookes 

 showed me and gave me a specimen of the kind of mica he 

 employs as eminently convenient for manipulation. It is found 

 naturally in a condition resembling artificially roasted mica. It 

 is not, however, quite so opaque for transmitted light, nor of 

 quite such a pearly whiteness for reflected light as that which has 

 been artificially roasted at a high temperature. The mica radio- 

 meter that Mr. Crookes first gave me, which I will call Mj, was, 

 Mr. Gimingham told me, the only one they had made wi^h 

 roasted mica. 



Mr. Crookes was so kind as to give me, for comparative ex- 

 periment, a mica radiometer, which I will call M.j, made from 

 the natural foliated mica. It revolves a good deal more quickly 

 than Mj under the influence of light ; it also gets more quickly 

 under way, indicating that the mica is thinner. When covered 

 with a hot glass it revolves positively, as already remarked ; 

 there is, however, but little negative rotation when the glass is 

 removed. 



The difference in the thickness and condition of the mica suf- 

 ficiently explains the difference of behaviour of M^ and Mg. 

 Any radiant heat incident on the white face that reaches the* 

 middle of the mica, whether it afterwards is absorbed by the 

 mica or reaches and is absorbed by the lampblack, tends to heat 

 the second or blackened face more than the first, and therefore 

 conspires with the heat incident on the lampblack, and absorbed 

 by it, to produce positive rotation ; and 'the smaller thickness 

 and less fine foliation of the natural mica are favourable to the 

 transmission of radiant heat to such a depth. 



P.S. — It might be supposed at first sight that the change of 

 rotation from negative to positive (in § 7) was due, not to a 

 change in the conditions of absorption, but to the circumstance 

 that the inner surface of the bulb had become waim by conduc- 

 tion, so as to be warmer than the surfaces of the fly instead of 

 colder. For we no sv know that the "repulsion resulting from 

 radiation," as in some way or other it undoubtedly does result, 

 is an indirect effect, in which radiation acts only through the 

 alterations it occasions in the superficial temperatures of the 

 solids in contact with the rarefied gas ; and it might be sup- 

 posed that when the inner surface of the bulb passed from colder 

 than the fly to warmer, the direction of rotation would, on that 

 account alone, be reversed. This, however, is not so. If bulb 

 and fly are at a common temperature, and the instrument is pro- 

 tected from radiation, the fly remains at rest whether the com- 

 mon temperature be high or low. If a small portion of 

 the total surface in contact with the rarefied gas be 

 warmed by any means, repulsion takes place, through the 

 intervention of the rarefied gas, between the warmed sur- 

 face and the opposed surfaces, if not too distant ; if it be 

 cooled, the result is attraction. It does not matter whether the 

 surface at the exceptional temperature belong to the fly or the 

 bulb. The former takes place in the ordinary case of a radio- 

 meter exposed to radiation, the latter in that of a radiometer at 

 a uniform temperature and protected from radiation when a 

 small portion of the bulb is warmed or cooled, in which case the 

 part at the exceptional temperature repels or attracts the disc 

 irrespectively of its colour or the nature of its coating. ^ Sup- 

 pose now that the fly is being warmed by radiation from without, 

 the bitlb being cool, at least at its inner surface. Let A, B be 

 the two kinds of faces of the discs, and suppose A to be the 

 better absorber of the total radiation. Then A will be the 

 warmer, and therefore will be more strongly repelled than B. 

 Suppose now that the bulb is heated till its inner surface be- 

 comes warmer than the fly. Then the fly will still be receiving 

 heat by radiation, to some extent also by communication 

 from the gas ; but this will be the same for both faces. 

 Hence if A be still the better absorber of the two (A, B), 

 A will be the warmer, and being less below the tem- 



' Theoretically there would be a minute difference of temperature, pro- 

 duced, other circumstances being r.like, by the difference in the absorbing or 

 emitting power of the two faces of a disc, as regards the radiation which is 

 the difference between the radiations from or towards the affected portion of 

 the bulb and the same portion at the normal temperature. But this, and the 

 repulsion or attraction corresponding to it, would be only a small quantity of 

 the second order, the main effect being deemed one of the first order. 



