220 



NATURE 



[7 an. ly, 1878 



LETTERS TO THE EDITOR 



\The Editor does not hold himself responsible for opinions expressed 

 by his correspondents. Neither can he undertake to return, 

 or to correspond with the writers of, rejected manuscripts. 

 No notice is taken of anonymous communications. 



The Editor urf^ently requests correspondents to keep their letters as 

 short as possible. The pressure on his space is so great that it 

 is impossible otherwise to ensure the appiarance even of com- 

 munications containing interesting and novel facts.\ 



The Radiometer and its Lessons 

 I AM sorry to have again to correct Mr. Stoaey ; but I cannot 

 allow the statements contained in his letter to pass unnoticed. 



I, There is nothing in my earlier paper that is " admittedly 

 erroneous^'' If there is error in these papers I am not aware 

 of it. 



3. These papers do not " conclude with Prof. Reynolds's own 

 expression of opinion that residual gas is not the cause of the 

 force observed by Mr. Crookes." Nor have I ever held or any- 

 where expressed such an opinion. 



3. In the passage to which Mr. Stoney refers, Clausius does 

 not imply that the law established by himself and Maxwell, viz., 

 that the only condition of thermal equilibrium in a gas is that of 

 uniform temperature, depends on the mean path of the molecules ; 

 and it was this law that I instanced as being at variance wiih 

 Mr. Stoney's assumptions (i) that gas is a perfect non-conductor 

 of heat ; (2) that a layer of gas across which the temperature 

 varies can exist in a state of thermal equilibrium without the 

 passage of heat from the hotter to the colder part. Mr. Stoney 

 has nowhere that I can see given any proof of these assumptions, 

 and I venture to prefer the authority of Professors Maxwell ani 

 Clausius, supported as it is by the whole evidence of facts. 



4. Mr. Stoney says that I have excluded the polarisation of 

 gas from my explanation. Mr. Stoney has not, that I am aware, 

 defined what he means by polarisation, but if he measures the 

 polarisation of a gas conducting heat by the excess of momentum 

 carried across any ideal surface in one direction over and above 

 that which is carried in the opposite, this polarisation is inde- 

 pendent of the length of the],mean path, and forms an essential 

 part of my explanation. 



There is one statement in Mr. Stoney's letter which is not 

 erroneous. He says : — *' I cannot find anywhere in Prof. 

 Osborne Reynolds's writings an explanation of the thing to be 

 explained, viz., that the stress in a Crookes's layer is different in 

 one direction from what it is at right angles to that direction." 



I do not at all admit that this is " the thing to be explained," 

 and I am quite sure that Mr. Stoney would .find no explanation 

 of it in my writing?. 



In the passage quoted above Mr, Stoney has, for the firat time, 

 so far as I know, expressly stated his belief that Mr. Crookes's 

 phenomena depend on such a difference of stress. I have thought 

 all along that his views were based on such an assumption, but 

 I did not like to take it for granted. It is almost a pity, it I may 

 use the phrase, that he did not express himself thus clearly at 

 fir»t, as in that case I might have done before what I am about 

 to do now, viz., prove definitely that such a condition of stress 

 can have nothing to do with the cause of Mr. Crookes's results — 

 that, so Jar from explaining, such a condition of stress is incon- 

 sistent with, these results, and this, not in mere matters of detail, 

 but as regards the fundamental direction in which the force acts. 



Throughout all the experiments that have been made one 

 invariable law as to the direction of motion has bf-en found to 

 maintain, which is that the force always teiads to drive the vanes 

 or bodies in the direction of their colder faces. Thus when a 

 body iifree to move in a sufficiently rarefied medium, if its front 

 be heated it will move backward, while if its front be cooled it 

 will move forward, always moving towards its colder face. 

 There are no exceptions to this rule. 



Let us now suppose that we have two bodies, A and B, free to 

 move in a sufficiently rarefied medium. Suppose A to be initially 

 hot and B cold, while the medium and surrounding surface are 

 at the mean temperature of A and B. Then, owing to the radia- 

 tion of heat between the two bodies, that side of A which is 

 opposite to B will be cooled faster, and hence be colder than the 

 other side of A. Hence according to the law stated above, A 

 must move towards B, and this it is found to do by experiment. 

 On the other hand, that side of B which is opposite to A will 

 become heated by radiation faster, and hence become hotter, 

 than the other side of B, and hence B will move away from A. 

 Thus if both bodies were free to move, we should have B 

 running away from A, and a running after B. 



This aspect of the phenomenon is perhaps the most paradoxical 

 that presents itself; it is nevertheless in strict accordance with 

 experiment, and it was by instancing this cas2 that I was 

 enabled to show that the force could not by any possibility be 

 directly due to radiation (see /'^?7. Trans., vol. 166, p. 728). 



The same reasoning now enables me to show, just as con- 

 clusively, that the force which causes ths motion in the b jdies 

 canno!; be due to the stress, in the layer of gas which separates 

 the bodies, being greater in the direction joining the bodies than 

 it is at right angles to this direction. For the only effect of such 

 a difference in the stress would be to cause the bodies to separate ; 

 therefore, instead of A following B, it would be forced back in 

 the direction of its hottest side, or in a direction opposite to that 

 in which it is found experimentally to move. 



This case, therefore, shows the fundamental error of Mr. 

 Stoney's view. Although he allows that the intervening gas is 

 the medium of communication, he assumes, none the less, that 

 the force acts directly between the two bodies (the heater and 

 cooler), in which c^se action and reaction must be equal between 

 the two bodies. Experiment, on the other hand, shows con- 

 clusively that the force acts independently between each body 

 and the gas which surrounds it ; the pressure being always 

 greatest on the hottest side. The force which acts on the body 

 reacts on the gas, cau->ing it to move in the opposite direction, 

 and the wind thus caused tends to carry all opposing obstacles 

 with ir. Hence, in the case above, the motion given to the air 

 at the one body must to some extent affect the opposing surface, 

 but this surface forms only one obstacle, while the action of the 

 wind is distributed throughout the entire chamber, in which it 

 acts in the manner so beautifully shown by Dr. S.:huster's plan 

 of suspending the vessel. A simple analogy to what happens 

 in the case of a and B is furnished by two steamboats, the one 

 following the other. The water thrown back by the screw of 

 the first would stop the second, but only to a small extent. 



When answering Prof. Foster in a former letter, I siid " that 

 it is contrary to tne kinetic theory that the increase resulting 

 from rarefaction in the mean path of the gaseous molecules 

 should favour the action." In making this statement all I meant 

 to imply was that the action was independent of any relation 

 between the mean path and the distance of the hot surface from 

 the cold surface, which was the only point in question. Although 

 my statement was strictly true in this sense, it appears to me, on 

 further consideration, that it might include more than I intended. 



I hope that nothing I have said, either in my earlier papers or 

 in this controversy, has led any one to suppose that I regarded 

 my explanation as entirely complete. I suggested, and to some 

 extent established, the true source of the force, namely the heat 

 communicated to the residual gas, and although nosy my sugges- 

 tion appears to have been universally accepted, it may be 

 remembered that at the time my first paper was written the only 

 other suggestions as to the cause of the motion observed by Mr. 

 Crookes were of a widely different character. As regards the 

 working out of the detail of my explanation, there has been one 

 point which I could not quite see through, viz., the influence 

 which the hot molecules receding from the surface might have on 

 the rate at which the cold ones would come up, and although I 

 have been trying to satisfy myself on this point ever since my 

 first paper was published, it is only within the last three months 

 that I succeeded. 



Now, however, I have arrived at a result which, although 

 somewhat unexpected and striking, will, I hope, be found to 

 reconcile what fias hitherto appeared to be anomalotis in the 

 phenomena already known, and to have suggested certain hither- 

 to unexpected phenomena which now only await experimental 

 verification. Osborne Reynolds 



January 15 



Sun-spots and Terrestrial Magnetism 



Precisely because the article (Nature, vol. xvii. p. 183) 

 on "The Sun's Magnetic Action at the Present Time," is by 

 so able a mathematical physicist as Mr. John Allan Broun, 

 and because of all sides of the solar problem there is none 

 wherein he is so facile princeps as the magnetic, I venture to 

 think this a good opportunicy for asking a question which has 

 troubled me much of late, and which is this : — 



The sun-spot cycle and the terrestrial magnetic diurnal oscil- 

 lation cycle are looked on now generally as being, if not actual 

 cause and effect, nt least as equally both of them effects of one 

 and the same cajse, and necessarJy, therefore, synchronous. 

 Yet if we inquire of the sun-spot observers the length of their 

 cycle, they declare it (as see Prof. Rudolph Wolf's admirable 



