420 



NA rURE 



[September 2, 1897 



F. Osmond ; a plea for uniform international specifications 

 for iron and steel, by Mr. Ast and Mr. Barba ; and an 

 account of the advantages to be derived from the forma- 

 tion of an international laboratory for investigating 

 the methods of analysing iron and steel, by Dr. H. 

 Wedding. In the last-mentioned paper it was pointed 

 out that the relations of buyers and sellers in the 

 iron and steel trades would be put on a much safer 

 basis if standard methods were worked out so that 

 any chemist of ordinary skill could be trusted to carry 

 them out. This could best be done at a central labora- 

 tory, where the work done in various countries could be 

 collated and reduced to a common standard. A com- 

 petent director for such a laboratory had been found in 

 Baron Jiiptner, and ample accommodation had been 

 granted at Zurich by the Swiss Government. Subscrip- 

 tions towards the cost of maintenance for ten years had 

 been promised by some of the large British and contin- 

 ental ironworks, and it was consequently decided to 

 begin operations in January next. 



In addition to the discussions bearing upon the testing 

 of iron and steel, several important memoirs were dis- 

 cussed in the Section, dealing with artificial building 

 materials, the principal ones being on the relation be- 

 tween the chemical composition of natural building stones 

 and their resistance to weathering, by Mr. A. Buess 

 (Hamburg) ; on the testing of earthenware pipes, by Mr. 

 Gary (Berlin) ; on the hardening process of calcareous 

 cements, by Mr. D. W. Michaelis (Berlin) ; and on the 

 determination of the quality of hydraulic cements, by 

 Mr. M. Meyer (Malstatt). 



The governing body, consisting of Prof. Tetmajer, Prof. 

 Martens (Berlin), Prof Debray (Paris), Mr. Berger 

 (Vienna), and Prof Belelubsky (St. Petersburg), was re- 

 elected ; and, in view of the fact that the whole of the 

 Council of the Iron and Steel Institute had joined the 

 Association, it was decided to have representatives of 

 the English-speaking countries on the governing body, 

 Mr. R. A. Hadfield (Sheffield) and Captain Carter (U.S. 

 Army) being those chosen. It was further decided that 

 in future the proceedings of the Congress will be published 

 in English, as well as in French and German: 



THE RADIATION OF LIGHT IN THE 

 MAGNETIC FIELD. 

 TOURING the past iew months some interesting ex- 

 -L>' periments have been made regarding the partial 

 polarisation of radiations emitted by certain luminous 

 sources when they are placed under the influence of a 

 magnetic field. Important investigations in this direc- 

 tion have been made by Dr. Zeeman, of the University 

 of Leyden, who has shown that the perturbations ex- 

 perienced by the ions, under the influence of magnetic 

 forces, produced new periods of luminous vibrations. 

 Continuing this work, Messrs. Egoroff and Georgiewsky 

 [Comptes rendus, April 5, 1897), with the aid of a Row- 

 land grating and a Ruhmkorff coil, have observed a 

 feeble broadening of the lines Dj and D, in the spectra 

 of both axial and equatorial radiations. In investigating 

 the appearance of coloured flames of polarised rays by 

 using the Savart analyser, it was observed that the 

 partial rectilinear polarisation of rays directed towards 

 the equator of the magnetic field was easily observed, 

 not only in flames of sodium, lithium and potassium, 

 but in induction sparks between magnesium electrodes. 

 In the cases of carbon, aluminium, mercury, zinc, bis- 

 muth and iron, the Savart analyser showed no indication 

 of rectilinear polarisation. In a second communication 

 to the same journal {Comptes retidiis for May 3), the 

 results there enumerated may be summed up as follows. 

 The relative quantity of equatorial radiations emitted 

 by a sodium flame, and polarised rectilinearly, varies 

 with the intensity of the magnetic field according to a 



NO. 1453, VOL. 56] 



particular curve. Under the influence of a magnetic 

 field of given intensity, the quantity of the light polarised 

 rectilinearly, and emitted equatorially by the sodium 

 flame, varies with the temperature of the flame. In 

 studying the change of spectra of metals in a field of 

 consideraI)le intensity, a modification of the method of 

 procedure was adopted, and resulted in the discovery 

 of the rectilinear polarisation of the equatorial radiations. 

 Nearly all the metals employed — namely, Cu, Tl, Zn, 

 Cd, In, Mg, Ca, Ba, &c.— showed polarisation ex- 

 clusively for those rays that are easily reversed. The 

 phenomenon, the authors state, " is observed in a very 

 instructive manner for the copper in the green part of 

 the spectrum (the change is very large for the ray 5105, 

 most feeble for 5153, and zero for the long waves 5217 

 to 5292). For indium, the change only occurred for the 

 violet ray at 4510, while the others (6193, 5230, 5900, 

 4680, 4616 and 4638) were not influenced by the field 

 at all." In experiments with Geissler tubes containing 

 hydrogen and helium, no definite results up to the present 

 have been obtained. 



In the experiments just referred to, it may be men- 

 tioned that all the observations were made with the eye. 

 It would, however, be interesting to inquire whether 

 the photographic plate would register these small vari- 

 ations, for then we should have a permanent record of 

 a phenomenon which is not so very easy to observe, or 

 which, at any rate, might be subject to " personal " error. 

 The application of photography to show such eff"ects has 

 been accomplished by Mr. Alexander Anderson, assisted 

 by Mr. Adeney. They employed a Rowland grating of 

 21-5 feet radius, and obtained photographs of the cadmium 

 spectrum, the source of light being a spark between cad- 

 mium electrodes from the secondary of a large induction 

 coil. Mr. Anderson's account of his experiments is 

 briefly summarised as follows. 



" Three lines are very distinct in all the photographs, 

 namely, those of wave-lengths 5086, 4800, and 4678 tenth- 

 metres. Photographs were taken, both without and 

 with a magnetic field, the time of exposure being exactly 

 the same in both cases. This field was produced by a 

 large elertro-magnet excited directly by a dynamo giving 

 a pressure with open circuit of 70 volts, the flat poles of 

 the electro-magnet being covered with ebonite, so that 

 they could be brought very close to the cadmium elec- 

 trodes without interfering with the sparking. The field 

 thus produced was practically uniform, and its intensity 

 was found to be about 17,000 C.G.S. units." 



In the first photographs the slit width was about '001 

 cm., but in the later photographs about twice this width 

 was used. An examination of the photographs showed 

 that there was no evidence of any eflect of the magnetic 

 field, though the definition of the lines was all that could 

 be desired, namely, very clear and sharply defined. 

 When it be remembered " that a length of one centi- 

 metre in the photograph corresponds to a change of 

 wave-length of 26 tenth-metres, and since an increase of 

 breadth of a line of one-tenth of a millimetre (and prob- 

 ably much less than this) could easily be seen on the 

 photographs, there could not have been a change in the 

 period of oscillation of as much as one part in 20,000." 



To account for this negative result, Mr. Anderson 

 suggests two possible reasons : viz. that perhaps the 

 magnetic field was not of sufficient intensity, or that the 

 exposure (thirty minutes for the narrow slit) was not long 

 enough. He states, however, that with an eyepiece in 

 place of the camera, he " saw (or fancied he saw) a 

 widening of the lines." 



The whole phenomenon of the widening of the lines 

 in the spectra of metallic substances in a magnetic 

 field is, however, of great interest to both physicists 

 and astro-physicists, and it is important that both eye 

 and photographic results should be obtained when 

 possible. 



