March 23, 1899] 



NA TURE 



48s 



LETTERS TO 'I HE EDI'IOR. 

 \The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 7iianiiscripts intended for this or any other part of NATURE. 

 No notice is taken of anonymoiis com/nuiiications.'\ 



Radiation in a Magnetic Field. 



The application liy Prof. Michelson of his interferometer to 

 the study of the structure of the spectral lines has raised two 

 important questions regarding the performance of this instru- 

 ment ; and it is to be hoped that perfectly satisfactory answers 

 to both of them may be forthcoming in the near future. These 

 questions are : 



(i) Is the complex structure of the lines indicated by the 

 interferometer a real structure existing in the light emitted by 

 the source ; or is it imposed, in part or altogether, by the 

 apparatus employed ? 



(2) Supposing the structure referred to in 'i) not to be 

 spurious, but to be real, or partly real and partly spurious, can 

 tile interferometer be relied on as a measuring instrument for 

 the purpose of determining the distribution of light in the 

 complex line by estimation from the visibility curve? 



With regard to (i), I may say that although it has been 

 suggested that the structure indicated by the interferometer is 

 entirely due to diffraction effects (or other unknown instru- 

 mental troubles), yet I personally am of opinion {from the study 

 of Prof. Michelson's work) that the structure indicated is in the 

 main real. It is possible, and indeed probable, that diffraction 

 effects influence the final results in some small degree ; but the 

 main character of the indicated structure agrees, no doubt, with 

 a real structure existing in the light emitted by the source. 



The modifications introduced by diffraction (if any) ought to 

 be detected from the fact that such effects are the same in 

 character for light of all wave-lengths, and their magnitudes for 

 different spectral lines depend on the wave-length only, and in 

 no way on the nature of the radiating substance. For this 

 reason I believe, with Prof. Michelson, that the structure indi- 

 cated by the interferometer as existing in certain spectral lines, 

 even when uninfluenced by the magnetic field, is a real structure ; 

 but as to whether it is all real, or to some small extent spurious, 

 has not yet been placed beyond all doubt. The discovery of 

 this structure adds one more to the already long list of achieve- 

 ments in the advance of science for which we are indebted to 

 Prof. Michelson, and I trust he will place it beyond all doubt as 

 to whether diffraction, or other causes, exert any appreciable 

 influence in the instrument, or in any way mask the true 

 structure. It is not sufficient to reply, as he does on p. 440 of 

 this journal (March 9), that the explanation of this struc- 

 ture by diffraction effects " would be very difficult to accept, in 

 view of the very great constancy of the results, with instruments 

 of different construction and dimensions, with different ob- 

 servers, and with different forms of vacuum tubes employed " ; 

 for whether the effects are due to diffraction or not, they ought to 

 remain the same under those circumstances here related, unless 

 the "instruments of different construction " differ in principle 

 and are not all interferometers. Diffraction cannot be the main 

 cause if the character of the effect differs for different wave- 

 lengths, and Prof. Michelson finds that it does differ for 

 different spectral lines : and in the same way, I think, it might 

 be determined if it intrudes itself as a modifying influence. 



With regard to question (2) above, the charge against the 

 interferometer remains most serious ; nor is it diminished in any 

 way by Prof. Michelson's further explanations given on p. 440. 

 The case is this — the interferometer, when applied to the study 

 of the splitting up of the spectral lines by the magnetic field, 

 yields the law that the magnetic separation of the constituents 

 " is approximately the same for all colours and for all sub- 

 stances." Now the facts of the case are that no such law 

 holds, even as the roughest approximation. The magnetic 

 separation is quite large for some lines, and very small, almost 

 imobservable, in the case of others, and this even in the case 

 of lines of nearly the same wave-length in the same substance. 

 In fact the law yielded by the interferometer is nothing short 

 of preposterous nonsense, and what remains to be done is to 

 determine the causes of error in the instrument, or to standardise 

 it, so that it may be employed as a measuring instrument. Of 

 course, as I have already mentioned (Nature, January 5, 

 p. 22S), the interferometer might have yielded this law without 



NO. 1534, VOL. 59] 



censure if by chance Prof. Michelson had happened to observe 

 lines which suffer approximately the same amount of resolution 

 in the magnetic field. But this is not the case, for in the case 

 of cadmium the separation for the blue line is more than 30 

 per cent, greater than for the green line, yet the interfero- 

 meter gives o'4i for the green line and 040 for the blue. 

 Similar remarks apply to the corresponding lines of zinc and 

 magnesium ; and what person, who has had even the slightest 

 survey of these effects, can have any doubt as to the great 

 difference in the magnitudes of the magnetic effects in the case 

 of the green lines of magnesium and the green lines of copper ? 

 — and so on ad infinitum. 



In conclusion, it may be well to mention that the relative 

 intensities of the light in the components of the magnetically 

 resolved lines, as observed by the eye in a good spectroscope 

 (2i'5-foot grating), are not by any means the same as those 

 indicated by the interferometer. Thus in the figures repro- 

 duced on p. 440, the central components A (Fig. I) are shown 

 by the interferometer as possessing greater intensity than the 

 lateral components 11. But when the resolved line is observed 

 by the eye (or photographed), it is at once seen that the illumin- 

 ation in the lateral components B (types II. and III.> is very 

 much greater than that in the central components A. Type II. 

 shows as a quartet (in which A is double) if the field is not very 

 intense ; but this quartet becomes resolved into a sextet, owing 

 to the side lines B splitting up into doublets when the field be- 

 comes very intense. There is no trace of the further little 

 "humps" pictured by Prof. Michelson, but the lines are clear 

 and sharp — and it is possible that these little humps maybe due 

 to diffraction effects (?). Similar remarks apply to its relative 

 illumination in type III.; and in this connection I may mention 

 that although I did not give an illustration of the general type 

 on p. 226, viz. that in which each constituent of the normal 

 triplet is itself a triplet (figured by Prof Michelson, p. 441, 

 f^'g- 3^' yfi' I stated in the text of my article (p. 226) that "all 

 the variations so far noted may be embraced in the general state- 

 ment that each line of the normal triplet may itself become a 

 doublet or a triplet." Indeed, these various types of effect were 

 observed by me as early as November 1897, and have been com- 

 municated to the Royal Dublin Society from time to time. 



Thomas Preston. 



Bardowie, Orwell Park, Dublin, March 16. 



The Phenomena 01 Skating and Prof. J. Thomson's 

 Thermodynamic Relation. 



I.N connection with Prof. Csborne Reynolds's " Notes on the 

 Slip))eriness of Ice," read before the Manchester Literary and 

 Philosophical Society (Nature, March 9, p. 455), the follow- 

 ing extract from a brief paper, read by me before the Royal 

 Dublin Society in 1886 {Proc. R.D.S., vol. v. p. 453), may 

 not be without interest. 



" To the many phenomena which have found an explanation 

 in Prof J. Thomson's thermodynamic relation connecting 

 melting-point with pressure, might be added those attending 

 skating, i.e. the freedom of motion and, to a great extent, the 

 ' biting ' of the skate. 



" The pressure under tlie edge of a skate is very great. The 

 blade touches for a short length of the hog-back curve, and, 

 in the case of smooth ice, along a line of indefinite thinness, 

 so that until the skate has penetrated some distance into the 

 ice the pressure obtaining is great ; in the first instance, theo- 

 retically infinite. But this pressure involves the liquefaction, 

 to some extent, of the ice beneath the skate, and penetration or 

 ' bite ' follows as a matter of course. As the blade sinks, an 

 area is reached at which the pressure is inoperative, i.e. in- 

 adequate to reduce the melting-point below the temperature of 

 the surroundings. Thus, estimating the pressure for that posi- 

 tion of the edge when the bearing area has become 1/50 of a 

 square inch, and assuming the weight of the skater as 140 lbs., 

 and also that no other forces act to urge the blade, we find a 

 pressure of 7000 lbs. to the square inch, sufficient to ensure the 

 melting of the ice at -3'5°C. With very cold ice, the pres- 

 sure will rapidly attain the inoperative intensity, so that it will 

 be found difficult to obtain ' bite '—a state of things skaters 

 are familiar with. But it would appear that some penetration 

 must ensue. On very cold ice, 'hollow-ground' skates will 

 have the advantage. 



"This explanation of the phenomena attending skating 



