September 27, 1906J 



NA TURE 



539 



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 intended for this or any other part of Nature. 

 No notice is tahen of anonymous communications.] 



The Recent Radium Controversy. 



I SHALL be obliged by your making tlie following correc- 

 tion on an accidental misstatement of mine, quoted by 

 Mr. Soddy {p. 517, line 3 from foot of column one) in 

 his very interesting article on " The Recent Controversy 

 on Radium," in your issue of September 20 (p. 516). In 

 a letter to the Times, from which the quotation is correctly 

 made by Mr. Soddy, I had written accidentally four 

 instead of five. The corrected statement is that Prof. 

 Rutherford had suggested that radium might be a com- 

 pound of one atom of lead and five of helium. This is a 

 suggestion wholly in harmony with chemical science. 

 Rutherford is scientifically cautious in naming lead as 

 possibly one of the ingredients of radium, but he names 

 helium as demonstrated experimentally to be an ingredient, 

 and considers five atoms as rendered somewhat probable 

 by elaborate and important e.\perimental investigations, of 

 which he gives careful descriptions and very complete 

 references in his book on " Radio-activity." 



Netherhall, Largs, September 21. Kelvin. 



Stress in Magnetised Iron. 



In Nature, .'\ugust 2 (p. 317), I observe under the above 

 title a letter from Dr. Shelford Bidwell re-opening a 

 question discussed in your columns ten years ago. As the 

 originator of the discussion I feel indisposed to let Dr. 

 Bidwell 's letter pass unnoticed. 



In my first letter (Nature, vol. liii., i8g6, p. 269) I 

 directed attention to the fact that certain writers, including 

 Dr. Bidwell, Dr. J. A. Ewing, and Dr. More, had stated 

 e.xplicitly or implicitly that the material of a magnet is 

 subjected to a longitudinal compressive stress approxi- 

 mating to B'/Stt when H/B is small, whilst other authori- 

 ties, including Kirchhoff and Prof. J. J. Thomson, postu- 

 lated a tensional stress of like amount. I remarked on 

 the apparent inconsistency, and explained the reasons which 

 led me to regard the last-mentioned view as the more 

 plausible. My letter led to others. Dr. Ewing explained 

 that he had changed his views. He left it, however, un- 

 certain whether he believed that no such stress as B'/Stt 

 exists in iron, or whether he took the view that tensional 

 and compressive stresses both exist, but in adjacent por- 

 tions of the magnet. The latter view seems indicated by 

 the illustration he advanced, viz. that of a man sitting in 

 a clothes basket and pulling the handles. The medium in 

 this case is obviously not free from stress, having a tensile 

 stress in his arms and a corresponding compressive stress 

 elsewhere. Prof. E. Taylor Jones wrote favouring a 

 tensile stress, and referring to work by himself and Prof. 

 Nagaoka on the subject. Prof. L. R. Wilberforce dwelt 

 on the fact that the stress B'/Stt most properly associated 

 with Maxwell's name has its seat in a hypothetical ethereal 

 medium, not in the material iron. Dr. Bidwell did not 

 then give his views to the public, so far as I am aware. 

 From his late letter I infer that they have remained un- 

 altered since 1896. 



Dr. Bidwell advances illustrations to explain his ideas. 

 His arguments, however, seem really to amount only to 

 this, that if two masses of iron, whether bars or spheres. 

 close to one another be capable of bodily movement (e.g. 

 if they rest on a smooth table or be suspended by long 

 threads), and have between them some compressible non- 

 magnetic medium (e.g. a finger), this interposed material 

 will be squeezed if the iron becomes suddenly magnetic. 

 This result is, however, equally consistent with either of 

 the above theories ; also it throws no light on the nature 

 of the stress In the iron. 



If, however, we suppose — as Dr. Bidwell ostensibly does 

 — the two hypothetical masses of iron to be consecutive 

 members of an infinitely long series, which seems the only 

 hypothesis likely to represent the interior of a magnet, 



NO. 1926, VOL. 74] 



and suppose them to be separated by air or by a non- 

 magnetic solid, there is obviously no reason why their 

 centres should approach one another when the masses all 

 become longitudinally magnetised. Number n mass is 

 urged to the right, say, by the attraction of mass n+i, 

 but to the left by the equal attraction of mass n— i, and 

 there is no reason to move bodily one way rather than 

 the other. If we regard the centres of the masses as 

 fixed, then it is obvious if air gaps separate the mass n 

 when unmagnetised from the masses n—i and n+i, that 

 after magnetisation it will — whether of extensible or highly 

 inextensible material — suffer a tensional stress in conse- 

 quence of the attractions exerted by its two next neigh- 

 bours. If the intervals between n and its neighbours be 

 filled by, say, lead, the lead will suffer no compression 

 unless the mass n lengthens when magnetised. If the mass 

 n lengthens, the lead will suffer compression ; but the 

 longitudinal stress in n, though less than if the lead were 

 non-existent, will still be a tension. If we suppose that 

 the poles of the hypothetical elementary magnets are not 

 quite at their ends, and that the elements lengthen when 

 magnetised, so that adjacent ends either meet across an 

 air gap or else exert pressure on an interposed non- 

 magnetic layer, then a compressive stress might be looked 

 for, not merely in the layer, but also in the short terminal 

 parts between the poles and the ends ; tensional stress 

 would, however, exist throughout the longer central por- 

 tions. Whether these various hypothetical cases will do 

 more than serve to show the inconclusiveness of Dr. 

 Bidwell 's illustrations is, I realise, very doubtful. 



In practice there are usually complications from free 

 ends and want of symmetry. We may, at least theo- 

 retically, avoid such complications by taking some endless 

 solid, the simplest being an anchor ring, preferably of 

 small section but large aperture. Is there a " hoop " 

 stress in such a ring when magnetised which did not exist 

 prior to magnetisation, and, if so, what is its sign? If 

 there is a " hoop " tension, then the case is so far 

 analogous to that presented by a ring rotating about an 

 axis through its centre perpendicular to the plane of its 

 aperture. If we imagine a short element of the rotating 

 ring bounded by planes through the axis, the tensions 

 across the end faces will give an inwardly directed radial 

 resultant which is balanced by the "centrifugal force." 

 .'\ stationary Saturn ring of continuous material, under the 

 attraction of a planet at its centre, similarly gives a case 

 of a compressive "hoop" stress; the pressures over 

 terminal faces of an element give an outwardly directed 

 resultant, balanced by the planet's attraction and that of 

 the ring on itself. In the .Saturnian as in the rotating 

 ring (when reduced to a statical problem), the hoop stresses 

 are really excited by a radial action. Is there anything 

 equivalent to this in the magnetic problem? If there is no 

 radial action there will naturally be a change of aperture, 

 unless, like Dr. Bidwell, we suppose the magnetic material 

 absolutely " rigid." With change of aperture there will be 

 change in the intermolecular distances, and so in the inter- 

 molecular forces. It would obviously be difficult to dis- 

 tinguish between the stresses due directly and those due 

 indirectly to magnetisation. 



I might add that Dr. Bidwell 's remarks on the 

 " uniformly magnetised rod " divided transversely seem to 

 me to confuse tensile with compressive stress. He deduces 

 a stress from iron to air gap, which seems really a tension 

 on the iron. If he supposes the two fragments of iron 

 held so as to prevent them doing more than just touch, 

 he will, I think, realise this. Again, his remarks in 

 reference to his spherical model do not seem to draw a 

 sufficientiv clear distinction between stress and strain. A 

 " rigid " body, if such an entity could be realised, might 

 be under stress though exhibiting no strain. On the other 

 hand, in an elastic body the signs even of the stress and 

 strain in a given direction may differ. 



During the ten years that have elapsed since the con- 

 troversy began I have been too busily engaged in other 

 matters to follow the developments of magnetic and elec- 

 trical theory. I hope that the recognised leaders in these 

 developments will not turn a deaf ear to Dr. Bidwell's 

 appeals for further light. C. Chree. 



September 15. 



