N^ TURE 



[May 9, 1 90 1 



Forty-four lines occur in Rowland's list only ; thirteen of 

 these have been traced with certainty to impurities of Ca, Mn, 

 Al, Pb or Sr. No origin other than vanadium has been found 

 for the remainder. A list of lines is given which Rowland has 

 previously identified with solar lines, but which, for some 

 reason or other, are missing from his latest list of vanadium 

 lines. 



There are 194 lines which occur in the Kensington list only. 

 No other origin has been found for them from a comparison of 

 the vanadium spectrum with those of the forty-three metals 

 which have been photographed on the same scale. As they 

 appear both in the spectrum of the chloride and o.\ide they are 

 probably genuine vanadium lines. 



" The Growth of Magnetism in Iron under Alternating Mag- 

 netic P'orce." By Ernest Wilson. Communicated by Prof. 

 J. M. Thomson, F.R.S. 



The subject of the shielding effects of induced currents due to 

 changes of magnetic induction in plates of iron has been dealt 

 with theoretically by Profs. J. J. Thomson {Electriciati, vol. 

 .xxviii. p. 599) and J. A. Ewing (Electrician, vol. xxviii. p. 631). 

 The same subject has been attacked by e.\periment in the case 

 of iron cylinders {Phil. Trans. R.S. A vol. cl.xxxvi., 1895, pp. 

 93-121, a.nA Journal Inst. Elec. Engineers, vol. xxiv. p. 195), 

 and the object of the present paper was to carry out further ex 

 periments with alternating magnetic force. 



The magnet consists of a solid iron cylinder 12 inches in 

 diameter and 18 inches long, around which the magnetising 

 coil is wound. The magnet circuit is completed by means of a 



•Q^ 



5; fl 



■5 .0 



1! s 



spoo 



10,000 16,000 



a.ver<s.^e B neia,r Surface. 



ring concentric with the cylinder and a circular slab of iron at 

 each end. 



In order that changes of magnetism may occur at the centre 

 of the cylinder it was necessary to employ alternating currents 

 of long periodic time. Such times vary from 2 '5 to 10 minutes 

 in the experiments. It would be difficult to obtain such 

 currents by means of electro-magnetic apparatus. A liquid 

 reverser consisting of copper plates in dilute CuSOj solution was 

 employed and gave satisfactory results. 



For the purpose of observing the changes of magnetic in- 

 duction at different points of the cylinder a series of holes were 

 drilled in a plane perpendicular to the axis and half way along 

 its length. Insulated copper wires were then threaded through 

 these holes in such manner that each circuit enclosed an element 

 of the cross-sectional area of the cylinder. Four such elements 

 were enclosed between the centre and the circumference, and 

 a fifth coil was wound completely round the cylinder, as it was 

 required to take account of the average change of induction over 

 the whole area. In the circuit of each of these five coils was 

 placed a dead-beat galvanometer, the deflection of which 

 enal>led the E. M.F. to be observed. The epoch for the 

 simultaneous observations of E.M.F. as well as the current in 

 the magnetising coil was determined by the operator at the 

 reverser counting seconds aloud. The E.M.F. curves were 

 ultimately plotted in terms of time and integrated. 



Before dealing with the experiments it may be stated that the 

 results obtained are applicable to cylinders of other diameter 

 .ihan 12 inches. Similar electric and magnetic events will 



happen in cylinders of different diameters, but at times varying 

 inversely as the square of their linear dimensions. Thus a 

 periodic time of 10 minutes with the 1 2-inch cylinder corre- 

 sponds to a frequency of 150 periods per second with a wire 

 I mm. diameter. 



Two variables have been dealt with, namely the frequency 

 of the magnetising force H, due to the current turns in the 

 copper coils of the magnet, and its amplitude. The figure gives 

 one set of results obtained at 15" C. Each curve refers to a 

 definite periodic time, and the number near it gives the fre- 

 quency for a wire i mm. in diameter. The point of interest is 

 that for a given frequency when the limits of the induction 

 density, B, at the surface, are small, that is, the average per- 

 meability is sm.all, the limits of B at the centre of the 

 cylinder do not differ greatly from those at the surface. With 

 larger limits of E at the surface, corresponding to a large average 

 permeability, the limits of B at the centre are very much smaller 

 than at the surface. Finally, when the limits of B at the 

 .surface are great, corresponding to a small average permeability, 

 the limits of v, at the centre are again more nearly equal to those 

 at the surface. 



Referring to the average induction over the whole core, that is, 

 taking account of phase displacement as well as variation in 

 amplitude, it may be stated that as the limits of B at the surface 

 increase from zero, the maximum average B over the whole 

 area grows less than B at the surface and then more nearly 

 approximates to it. The curves are similar to those in Fig. I, 

 but the percentage diminution for a given frequency and a given 

 value of B at the surface is not so great. 



The magnet was heated to 53° C. and 

 the effect was to tend to equalise the 

 limits of B over the whole core. In the 

 figure the points were obtained when 

 the magnet was at this temperature. The 

 .same maximum average B over the whole 

 core is obtained with slightly less ampli- 

 tude of magnetising force. On account of 

 a greater phase displacement of the change 

 of induction as the centre is approached, 

 the maximum average B over the whole 

 section is not greatly altered for a given 

 maximum value of B at the surface. 



Chemical Society, April 18. — Prof. 

 Emerson Reynolds, president, in the chair. 

 — The following papers were read : — 

 Action of alkyl haloids on aldoximes and 

 ketoximes. Part 2. Alkylated oximes and 

 isoximes and the constitution of aliphatic 

 oximes, by W. R. Dunstan and E. Gould- 

 ing. — The supposed existence of two iso- 

 meric triethyloxamines, by W. R. Dunstan 

 Nitrocamphene, aminocamphene and hydr- 

 oxycamphene, by M. O. F"orster. On treating i : l-bromo- 

 nitrocamphane with silver nitrate, i-niirocamphene, 



CH 

 C,H„, li 



-C.NO., 



is produced, and on reduction gives the corresponding i-artiino- 

 On heating the sulphate of this 



and E. Goulding.- 



it is converted into i-hydroxy- 

 This substance is of importance 



camphene, CsHu' 1 



^■-CNir, 



base with potassium nitrite 



,CII 

 camphene, Csllj,,/ || 



^COH 



as being the enolic isomeride of ordinary camphor, into which 

 it is converted by warm mineral acids. — A contribution to the 

 chemistry of the triazoles, by G. Young and W. U. Oates. 

 The authors discuss the possibility of isomerism in the triazole 

 series as compared with the pyrazoles ; they have prepared a 

 number of substituted triazoles from the corresponding semi- 

 carbazones. — Researches on moorland waters. Part 2. On the 

 origin of the combined chlorine, by W. Ackroyd. The author 

 concludes that the common salt in the water of the Widdop 

 reservoir in Yorkshire is derived from the winter rainfall. — 

 Robinin, violaquercitrin and osyritrin, by A. G. Perkin. The 

 composition Ci5H3,0.,„,8H20 is now assigned to the glucoside 



NO. 1645, VOL. 64] 



