February 20, 1896] 



NA TURE 



365 



letism (and the ultimate theory of magnetism is • undoubtedly 

 kinetic) ^ involves quite a different class of motions from those to 

 \shich the beautiful phenomena discovered by Mr. Guthrie are 

 !ue. He rather wished to point out the close connection that 

 \isted between the laws of some of these actions and the laws of 

 lu^gnetism, which, while involving some remarkable coincidences, 

 nvolves certain contrasts decisive against any hypothesis, 

 iich as the ingenious one- of Euler, explaining magnetism by 

 luid motion directly comparable with that which forms the sub- 

 cct of the present communication." Kelvin. 



The University, Glasgow, February 13. 



The Stress in Magnetised Iron. 

 I AM glad to see from Dr. Taylor Jones' letter that he and 

 Mr. Nagaoka contemplate a discussion of the niagnittide of the 

 iiagnetic stress really existing in iron under, I hope, natural 

 mditions. A complete discussion of the sort would, I think, 

 lie of much value. It seems to me somewhat doubtful what is 

 the true nature of the assumptions latent in Maxwell's work, 

 vol. ii. arts. 630-646. I am uncertain whether his conclusions 

 lie strictly applicable to any case other than that of an infinite 

 iiimogeneous medium in which the permeability is unity. In 

 !ie ordinary case of a magnetic bar, some lines of force traverse 

 he surrounding air, and complications also arise from the 

 ■ free " magnetism at the ends. When, as in previous ex- 

 ^leriments on the magnitude of the stress, a magnetic bar or 

 ellipsoid is cut in two, the stress meas- 

 ured is presumably that exerted across a 

 minute air-film separating the opposed 

 surfaces. Under such circumstances the 

 formula B'/Sir is, according to Dr. Taylor 

 Jones' own experiments {Phil. Mag. for 

 1895, P- 254), a close approximation to 

 the stress when the permeability of the 

 bar is large. 



Prof. Ewings letter also calls for some 

 comment at my hands. The view I ad- 

 vocated is hardly that he attributes to me. 

 I expressed no opinion as to the correct- 

 ness of the measure B-ZStt, assumed by 

 Dr. More and others, for the intensity of 

 the magnetic stress. Neither did I say 

 the stress system reduced to a simple 

 longitudinal tension. It is, I believe, in 

 general considerably more complicated, 

 and is accompanied by strain perpen- 

 dicular to, as well as along the lines of 

 force. Avoiding complications unessential 

 to the point at issue, I simply supported 



the view that the magnetic stress along the lines of force is a 

 tension, and that the associated strain in the metal along this 

 direction is an extension. 



Again, Prof. Ewing says that the case illustrated by me in 

 p. 270 is a " special one." The illustration in question is, how- 

 ever, just as applicable to the case of an endless ring, to which 

 he specially refers, as to that of a long rod. It is really based on 

 three hypotheses or assumptions, as valid in one case as in the 

 other : ( I ) that the magnetisation is uniform ; (2) that the air gaps, 

 real or imaginary, are narrow compared to the length of the 

 element ; (3) that the existence of an indefinitely thin air gap, 

 such as may be found Vjetween two attracting pole faces in con- 

 tact, does not affect the sign of the stress. 



Now ( I ) is merely a definition, and (2) a mathematical expedient 

 to simplify the proof, so that (3) alone remains. As regards (3), I 

 merely followed Prof. Ewing in § 145 of his " Magnetic Induc- 

 tion," except in so far as I did not assume the existence of a 

 narrow gap to be wholly immaterial. As Prof. Ewing admits 

 that the stress in the case of my illustration is a tension, we 

 should, if he adhered to his original views regarding {3), be now 

 on the same side. He has, however, apparently completely 

 altered his views, for he says : " Make the iron continuous by 

 closing up the gaps, and the tensile stress disappears." This 

 seems an explicit declaration that there is no such thing as a 

 " Maxwell stress," and that consequently neither extension nor 

 •' impression can be assigned to such a cause. Prof. Ewing's 



' " Electrostatics and Magnetism," ti 290 (Roy. Inst. Friday Evening 

 1 t-cture, May 18, i860) ; '.' Electrostatics and Magnetism," § 546 (Proc. Roy. 

 Soc., May 10, 1856). 



2 '' Electrostatics and Magnetism," § 573. 



NO. 1373, VOL. 53] 



new view implies a sudden large discontinuity between the state 

 of close juxtaposition of magnetic material, when the stress he 

 admits is large, and that of absolute continuity, when the streiss 

 he says is nil. 



It would, I think, be premature to do more than call attention 

 to this apparent discontinuity until the arguments in favour of 

 the new view are produced. 



Prof. Ewing's remarks about the pulling and pushing of 

 magnetised rings, indiarubber bands, &;c., suggest that he 

 experiences a difficulty in imagining any complete, homogeneous, 

 isolated ring under conditions such that the normal stresses over 

 all cross-sections are tensions. The consideration of- the elastic 

 stress system produced in an anchor ring by rotation about a 

 perpendicular to the plane of its circular axis (line of centres of 

 cross-sections) through the centre will be found, I think, to aid 

 such a conception. 



I must decline, I fear, to follow Prof. Ewing into the clothes- 

 basket he refers to, though doubtless the experiment must be a 

 fascinating one — for spectators. C. Chree. 



Kew Observatory, February 11. 



Experiments with Soaring Machines. 



At the end of the article on Herr Lilienthal's soaring experi- 

 ments, in your issue of January 30, reference is made to my 

 trials, with somewhat similar apparatus, at Cardross, and it is 

 said that so far I have not been very successful. 



I do not consider this to be the case. I think that, so far as I 

 have gone, my work has been rewarded with as much success 

 as could possibly have been hoped for. 



I only commenced to build my first machine just a year ago, 

 and being very busy with other things, was not able to lake the 

 machine out to practise till June. 



At first my wings were, as is stated, very much raised, so 

 that the machine should have transverse stability of its own in 

 flight, just as a kite is stable sideways when it is of this shape, 

 or as a sheet of paper folded to a V-shape will always fall on the 

 folded edge. 



There is, however, a great practical difficulty with the raised 

 wings, and that is, that if one stands not exactly head to wind, 

 the machine tends to "i>ayoff" from the wind, which then 

 comes more on one side, and rolls the machine over. 



With the wing tips lowered more as a gull's wings are, if the 

 wind (virtual or real) is not ahead, through a squall or any other 

 cause, the machine has very little tendency to raise on the wind- 

 ward side, and so be turned over by the wind. 



Not appreciating this, certainly caused me some delay and 

 breakages ; but when I realised how great is the objection to the 

 raised wings in such a squally place as Cardross, I curved the wings 

 as shown in the second photo which I enclose, and with this 

 original apparatus have been picked up by a puff of wind over 

 and over again, sometimes as much as 12 feet above the spot I 

 was lifted from, and put down again on exactly the same place. 

 At other times, when there has not been so much wind, I ran to 

 meet the wind with the front of the wings depressed somewhat, 

 so that my Afeight is only partially taken by the wings, and in 

 this way T' am, ' strange as it n.ay appear, able to run very 



