74 



NATURE 



[November 26, 1891 



been for the res angitsta domi of his later years, we might 

 not have had his " Annals of Electricity " or his " Scientific 

 Researches." As it is, it seems well established that he 

 was the pioneer of electro-magnetic discovery, the maker 

 of the first electro-magnet and the first magneto-electric 

 machine ; and Dr. Thompson has done a simple act of 

 justice in bringing Sturgeon's claims in this respect before 

 the great public now interested in the progress of electrical 

 science. 



In chapter ii. we have generalities concerning electro- 

 magnets and electro-magnetism, descriptions of typical 

 forms of electro-magnets, and materials of construction. 

 First are discussed such topics as the uses of electro- 

 magnets, magnetic polarity, magnetic units, and element- 

 ary propositions in electro-magnetism. Here we would 

 remark, as a merit in this work, the fact that the author 

 has not loaded what is intended to be a thoroughly prac- 

 tical treatise with long discussions of purely theoretical 

 matters, however important in relation to practice. The 

 sketches of theory given in many works — which are really 

 only collections of tables of numerical data, lists of 

 formulae, and workshop recipes — would be much better 

 left out, and their places supplied by other matter, or the 

 books lightened by their absence. 



An excellent chapter follows, on the properties of 

 iron. The various methods of measuring permeability 

 are well explained, and for the space devoted to it a good 

 account (with tables) is given of the results obtained by 

 Hopkinson and Ewing in their researches. On p. io8 

 the effect of opening gaps in a magnetic circuit of iron is 

 discussed, and a reference is given to an experiment 

 described later in the work. On turning to this experi- 

 ment (p. 212), it is found to be a description of an explora- 

 tion of the effect produced at different parts of a horse- 

 shoe steel magnet by pulling off the keeper. A narrow 

 coil of a few turns of wire is wound on a light frame 

 capable of being slided round the magnet, and is con- 

 nected with a ballistic galvanometer. The effect of pulling 

 off the keeper is then tested with the coil in different posi- 

 tions on the horseshoe, and is shown by the deflection of 

 the needle of the ballistic galvanometer. The theory of 

 the author is, that putting on or taking off the keeper of a 

 permanent steel magnet does not affect the magnetization 

 at the middle of the horseshoe ; that by putting on the 

 keeper, and so diminishing the magnetic " reluctance " of 

 the circuit, the lines of magnetic force are only collected, 

 not altered in amount. Hence, if this theory be true, the 

 coil, when placed at the middle of the magnet, should 

 show no effect on the removal of the keeper. It is stated 

 that careful and repeated experiments made at Finsbury 

 gave an effect at the middle of the magnet which did not 

 amount to 1/3000 of that found when the coil was close 

 up to either end of the magnet. 



Some time ago, when informed by a friend of this state- 

 ment and result, the writer, feeling extremely doubtful of 

 their general truth, had a magnet constructed for the 

 purpose of repeating the experiments. As a large magnet 

 was required for other purposes, one was constructed of 

 eight steel bars of a mean length of about 3 feet. Each 

 bar was 2 inches broad by \ inch thick, so that when put 

 together they formed a large horseshoe of square cross- 

 section 2 inches in side. A keeper, made of a block of 

 soft iron, fitted between the ends of the horseshoe. The 

 NO. I 152, VOL. 45] 



steel, which was tool steel obtained from a local firm, 

 took magnetism readily, and an excellent horseshoe 

 magnet was obtained. 



A coil in circuit with a ballistic galvanometer was used 

 in the manner described above, to test the effect of 

 removing the keeper. Careful experiments made by 

 students, and repeated by the writer, gave an entirely 

 different result from that obtained by Prof. Thompson. 

 A very large throw was obtained by placing the coil close 

 to either end of the magnet and detaching the keeper ; 

 but with the coil as nearly as possible at the middle of 

 the horseshoe, the throw was about one-eighth of the 

 maximum. It was verified, moreover, that the minimum 

 throw was obtained at the middle. 



This result was exactly what the writer had expected 

 would in general happen. The so-called free magnetism 

 at the extremities of the magnet, in the absence of the 

 keeper, produced a demagnetizing effect throughout the 

 magnet, and thus diminished the induction through the 

 coil, even when at the middle. This action was counter- 

 acted by the magnetizing effect of the keeper when in 

 position, and therefore itself inductively magnetized, but 

 had full play as soon as the keeper was removed. The 

 much greater deflections near the ends were undoubtedly 

 due to the cause to which Dr. Thompson would assign 

 the whole effect — the alterations of the arrangement of the 

 lines near the ends which accompanied the removal and 

 replacement of the keeper. 



It is- certain that this effect will depend on the permea- 

 bility of the magnet steel, which is a function of the 

 magnetization ; but that such an effect will in general be 

 produced there does not seem to be any room for doubt. 



The following chapters deal with specially designed 

 electro-magnets, such as, for example, those used in relays 

 and clockwork, electro-magnetic mechanism, alternate- 

 current electro-magnets, electro-magnetic motors and 

 machine tools, and the purely electro-magnetic part of 

 the book winds up with a very interesting chapter on the 

 electro-magnet in surgery. The last chapter of all is 

 devoted to permanent magnets. To give a satisfactory 

 account of these chapters is here impossible ; but it may 

 be mentioned that the electro-magnetic mechanism fully 

 described and figured includes no less than nine classes, 

 beginning with the different forms of magnet with 

 moving armature or plunger, magnets with armatures 

 moving against counterpoises of different kinds, polar- 

 ized devices, electro-magnetic vibrators, magnetic brakes, 

 &c. 



In chapter xi., on alternate-current electro-magnets, 

 the modes of laminating magnets for the prevention of 

 eddy-currents are described ; then follows a discussion of 

 effects of alternating electro-magnets, depending on the 

 difference of the phase relations of the magnets and the 

 eddy-currents excited in conductors in the form of disks 

 and rings placed near the extremities of the iron cores. 

 Thus we have a very interesting account of Elihu Thom- 

 son's remarkable experiments. The throttling or im- 

 pedance effect of electro-magnets included in circuits is 

 next treated. It would have been worth noticing, where 

 the relations of maximum current, maximum electro- 

 motive force, mean current, mean electromotive force, 

 and impedance are given, that the true mean value of the 

 total electrical activity in an alternating circuit, in which 



