From George Gabriel Stokes   24th June 1864

Lensfield Cottage Cambridge | 24th June 1864

Dear Tyndall,

I sent you tonight by book post your original paper,¹ which I found at once, when I thought of looking for it, in a certain box in which the housemaid is in the habit of putting things for tidiness but which I do not myself use and had not thought of looking into. I had previously had a good hunt for it in several rooms. The copy of the MS has not yet been returned.

I was one of the referees. I did not keep the MS after my return to London till I should have completely looked over it, knowing that you wanted it, and therefore in my remarks I must trust a good deal to memory. There were three points on which I wished to make some remarks.

I. By operating on vapours, filling your tube, in tensions varying as you ensure a common height of column of the substance supposed liquefied. The reader would naturally like to know what that common height is. This is got by an easy numerical calculation from the data you have given combined with the ratio of the densities of air and water at a standard barometric height and the value of the height in inches. I make it about 0.00215 inch. I think this calculation should be made once for all by the author instead of being thrown on each individual reader. As it happens the order of opacity comes out the same for the vapours and liquids, because the latter were employed in a cell sufficiently thin to have its thickness comparable with 0.00215 in. so that there was no inversion in passing from the latter to the thickness employed. But to know whether any such inversion if it had occurred would or would not prove an inversion of order referable to the state (liquid or gas) of the substance, we should want to know whether the equivalent liquid thickness of the gas was equal to the actual liquid thickness in the experiments with liquids, or at least sufficiently nearly so to preclude the supposition of a change of order arising merely from that cause.

2. The statement as to the raising (shortening) of the periodic time of vibrations seems to me somewhat too absolute inasmuch as it is quite conceivable, be it even deemed improbable, that it is the vibrations of short period which we know to exist that are converted, without the wire into much longer ones, with the wire also into somewhat longer ones. But be that as it may, there is of course one broad distinction between this phenomenon and fluorescence, that in the latter case it is ethereal in the former molecular vibrations that are so converted.

By the bye fluorescence is not by any means confined to the more refrangible part of the spectrum. In the case of chlorophyll for instance, it begins near the line B and consequently very near the red extremity. Of course you are well aware of this, and I merely make the remark because (if I recollect rightly) your words would seem to imply that I had found it only in the more refrangible rays.

3. This is the most important of the points I have to mention. Your experiment on the quality of the heat radiated by the lamp-black film covering the thin metal face of a cube containing water at 212^o is sufficiently remarkable whatever be the explanation of it. But a difference of quality between the radiations from two different opaque bodies at the same temperature is what cannot be admitted without the clearest experimental evidence, seeing that it is in opposition to the extension B Stewart² and Kirchhoff³ have made of Prevost’s law of exchanges.⁴ Prior to your experiment, had any one asked me I would have assumed with you that the temperature of the lamp-black was sensibly equal to 212^o. But I now believe that so rapid is the loss of heat by radiation from the lamp-black and so bad the conducting power of the substance, thin as the film is its outer portion, from whence almost exclusively the radiation comes, is very sensibly lower in temperature than 212^o. The point is well worth being made the subject of further experiments. If my explanation is right you ought to get different results from thin films of different thickness. Also it ought to make a difference if the face was if I may so speak muffled i.e. made to radiate in all directions but that wanted for the pile into a chamber kept to 212^o, the chamber itself being screened off from the pile.

I hope you will find leisure to take up this subject, for unless my explanation, referring the result to known causes, be the true one (I myself however expect it will turn out so to be) a great deal may come out of the phenomenon. Meanwhile I think it would be well to insert some expression of doubt (I don’t mean as to the accuracy of the experiment but as to the interpretation) in this part of your paper.

The other referee⁵ after speaking very highly of the paper remarks. There is just one important point of theoretical connexion in which what is said in the paper appears to me altogether unsatisfactory…. It appears to me that the first quarter of § XIV is not at all conclusive; and further that it does not present any clear or probable view as to the real relation between radiation and conduction from which a reader would be led to see his way towards a conclusion. I do not think I say too much when I say that after reading it and considering it carefully I could not infer from it as probable that rock salt should be a better conductor than glass; or silver than platinum. That in these cases the worse radiator is the better conductor must, as far as it is proved, be looked on as a merely experimental conclusion and in my opinion the theoretical views in the particular passage referred to are less weighty than a mere statement of the fact, and a natural remark that the explanation is probably to be sought for in some relation between radiation and conduction. A full theory of this relation is conceivable as being within reach, but I can not think that the passage in the paper to which I object brings us nearer it, scarcely does as much as Huxley would call showing a finger post pointing to a way that might lead to it. I therefore think it would be a great improvement in the paper to remove or modify very much the first part of the concluding § XIV. But there is one more important matter noted in that § - the case of copper which of course must occur to any one as an exception when a table of the other metals is given in order of conductivities and reverse order of radiating power. It is very much to be desired that Tyndall should give a short description of the experiments he refers to in which he proved that silver is a better radiator than copper. A result so opposed to supposed results of previous experimenters is very striking, and particulars as to the evidence of it would be received with great interest. If not added to the present paper it is to be hoped Tyndall will take an early opportunity of publishing it.

With reference to the Akin experiment⁶ he says. The platinum wire in the hydrogen flame may be said to be heated by conducted heat of the vapour and gas (whatever conducted heat may mean) and Tyndall is scarcely justified in his facts in saying that radiation from the flame could make the wire incandescent (unless by a concentration of the small amount of violet and ultra violet rays). It would be easy <hereafter LT Typescript only> to try by a good silver ellipsoidal reflector whether the whole radiation—violet rays and all included—could so be got to produce incandescence in a piece of platinum or other solid out of the flame, and I think this experiment would be worth trying. Mean while I think some slight change such as the insertion of a probably would improve Tyndall’s statement, which as it stands is extremely interesting and suggestive, if to some extent open to objection.

Yours very truly | G. G. Stokes

RI MS JT/1/S/235

RI MS JT/1/TYP/4/1390–3