>XPERIMENTS WITH 
 VACUUM TUBES 
 
 SIR.D.SALOMONS 
 
A LIST OF WORKS 
 
 LIBRARY 
 
 UNIVERSITY OF CALIFORNIA. 
 
 Class 
 
 naecinc j-,igui riumg, 5*. 
 Alternating Currents, Blakesley, 5*. 
 
 Bedell and Crehore, I2s. 6d. 
 
 Circuit, Maycock, 2s. 6d. 
 
 Houston and Kennelly, 5-y. 
 
 Loppe and Bouquet, los. 6d. 
 
 Steinmetz, 2os. 
 
 Still, $s. 
 
 American and English Lathes, Horner, 
 
 2IS. 
 
 Analysis, Electrolytic, Neumann, 6*. 
 
 Steel Works, Arnold, los. 6d. 
 
 Volumetric, Coppock, 2s. 
 
 Anatomy, Brodie,4 parts, or I vo\.2l.2s. 
 Anderson's Heat, 6s. 
 Apothecaries Hall Manual, Thomson. 
 Arc Lighting, Houston, $s. 
 Arithmetic, Electrical, is. 
 Arnold's Steel Works Anal., los. 6d. 
 Ashworth's Magnetism and Elec- 
 tricity, 2s. 6d. 
 
 Astronomy, Chambers, 2s. 6d. 
 Atkinson's Static Electricity, JS. 6d. 
 Atlantic Ferry, Maginnis, 7-r. 6d. 
 Auto Cars, Farman, 5-r. 
 
 Bales' Shafting and Gearing, 2s. 6d. 
 
 Ballooning, May, 2s. 6d. 
 
 Barker's Engineering Works Manage- 
 ment, 3-y. 
 
 Barter's Drawing, $s. 6d. 
 
 Woodwork, 6s. 
 
 Battery, Storage, Tread well, TS. 6d. 
 net. 
 
 Galvanic, Bottone, 5^. 
 
 joiuyuie, otuciy, vja.na.ii., jo. 
 
 Bjorling's Mechanical Engineer's 
 
 Pocket Book, 5*. 
 
 Pipes and Tubes, 31. 6d. 
 
 Blair's Analysis of Iron, iSs. net. 
 Blakesley's Alternating Currents, 5^. 
 
 Geometrical Optics. 
 
 Bodmer's Hydraulic Motors and 
 
 Turbines, 155. 
 Inspection of Railway Material, 
 
 5 s - 
 Bonney's Electrical Experiments, 
 
 25. 6d. 
 
 Electro-platers' Handbook, 3*. 
 
 Induction Coils, 3*. 
 
 Bookkeeping, Crellin, is. Key, 
 
 is. 6d. 
 
 Bern's Metric Measures, 35. 
 Botany, Massee, 2s. 6d. 
 Bottone's Electric Bells, 3^. 
 Electrical Instrument Making, 
 
 3s.6<t. 
 
 Electric Light Guide, is. 
 
 Electricity and Magnetism, 
 
 2S. 6d. 
 
 Electro-Motors, 3-r. 
 
 Galvanic Batteries, 5^. 
 
 How to Manage a Dynamo, 1^. 
 
 Radiography and X Rays, 3*. 
 
 The Dynamo, 2s. 6d. 
 
 Wireless Telegraphy, 35. 
 
 Boyd's Coal Pits, 7s. 6d. 
 Boyd's Petroleum, 2s. 
 Bricklaying, Maginnis, 7-r. 6d. 
 British Locomotives, Cooke, 7-r. 6ct, 
 
Brodie's Dissections, 2/. 2s. 
 Browne's (Jukes) Geology, 2s, 6d. 
 
 Cabinet Making, Denning, $s. 
 Cables, Electric Light, Russell, ios.6d. 
 Cakes, Marshall, is. 
 Calculus, Diff. and Integral, Wyatt, 
 
 3-r. 6d. [net. 
 
 Carpentry and Joinery, Fletcher, 5*. 
 Catechism, Electrical, Shepardson, 
 
 i os. 
 Central Station Electricity Supply, 
 
 Gay and Yeaman, los. 6d. 
 Chambers' Astronomy, 2s. 6d. 
 Chemical Analysis, Coppock, 2s. 
 
 Composition, Landolt, 45. 6d. 
 
 Evolution, Rudorf, 45-. 6d. 
 
 Manipulation, Hewitt, 4^. 6d, 
 
 Chemist's Compendium, $s. 6d. 
 Chemistry, Physical, Reychler, 4^. 6d. 
 
 Practical, Cooper, 2s. 
 
 Harris, 3 vols. 45. 
 Coal Pits, Boyd, 7s. 6d. [2is. 
 
 Colour in Woven Design, Beaumont, 
 Cooke's Locomotives, 7s. 6d. 
 
 Locomotive Development, 2s.6d. 
 
 Cooking, Jack, 2s. 
 
 Marshall, is. 
 
 Cooper's Chemistry, 2s. 
 Coppock's Volumetric Analysis, 2s. 
 Crapper's Electrical Measurements, 
 
 25. 6d. 
 
 Crellin's Bookkeeping, is. Key, 
 
 is. 6d. 
 Cullyer's Land Measuring Tables, 
 
 25. 6d. 
 
 Davis' Builder's Quantities, 35. 6d. 
 Dawson's Electric Railways, 42^. 
 Denning's Cabinet Making, $s. 
 Designing, Leland, 2s. 
 Designs, Electrical, los. 
 Dictionaries, Technological. 
 Dictionary of Medical Terms, Hoblyn, 
 
 los. 6d. 
 Dissections Illustrated, Brodie : 
 
 Part I., Ss. 6d. Part II., los. 
 
 Part III., los. Part IV., los. 
 
 In One Volume, 2/. 2s. 
 Dobbs' Telephone Work, 3*. 6d. 
 Drainage, Beardmore, 2s. 6d. 
 
 Nadi&ne, is. 
 
 Draughtsmen's Work, is. 6d. 
 Drawing and Designing, Leland, 2s. 
 Drawing and Designing for Marine 
 
 Engineers, Roberts, 6f. 
 
 Drawing, Woodwork, Barter, $s. 6d. 
 Dynamo, Bottone, 2s. 6d. 
 
 Building, Parkhurst, $s. 
 
 Gibbings, is. 
 
 Hawkins and Wallis. 
 
 How to Manage, Bottone, is. 
 
 Strange History, is. 
 
 Testing, Parham, 12s. 6d. 
 
 Machinery, Hopkinson, 5^. 
 
 Wiener, 15^. 
 
 Edgcumbe's Electrical Engineer's 
 
 Pocket-Book, 5*. 
 Education, Practical, Leland, 6s. 
 Electric Batteries, Bottone, 5^. 
 
 Bells, Bottone, jr. 
 
 Heating, Houston, $s. 
 
 Influence Machines, Gray. 
 
 Lamps, Maycock, 6s. 
 
 Light Cables, Russell, IQJ. 6d. 
 
 Light Fitting, Allsop, Ss. 
 
 Light Installation, Salomons': 
 
 Vol. I., Accumulators, 5-r. 
 
 Vol. II., Apparatus, fs. 6d. 
 
 Vol. III., Application, 5^. 
 
 Lighting of Colliery, 2s. 6d. 
 
 Lighting, Guide to, is. 
 
 Lighting Specifications, Js. 6d. 
 
 Lighting, Maycock, 6^. 
 
 Lighting for Marine Engineers, 
 
 Walker, 5*. 
 
 Railways, Hering, 5^. [15^. 
 
 Railway Handbook, Herrick, 
 
 Railways and Tramways, Daw- 
 son, 42^. 
 
 Traction, Rider. 
 
 Transformers, Kapp, 6s. 
 
 Transmission of Energy, Kapp, 
 
 los. 6d. 
 Transmission of Intelligence, 
 
 Houston, 5-s-. 
 
 Wiring, Maycock, 6s. 
 
 Details Forms, 2s. 6d. net. 
 
 Tables, Maycock, $s. 6d. 
 
 Electrical Catechism, Shepardson, 
 
 Designs, lOr. [los. 
 
 Energy, Kapp, los. 6d. 
 
 Energy, Plante, 12s. 
 
 Engineering, Steinmetz, 125. 6d. 
 
 Engineer's Pocket-Book, Edg- 
 
 cumbe, 5^. 
 
 Experiments, Bonney, 2s. 6d. 
 
 Instrument Making, Bottone, 
 
 3J. 6d. 
 
 Measurements, C rapper, 2s. 6d. 
 
 Measurements, Houston, 5^. 
 
( 3 ) 
 
 Electrical Terms, Houston, 3U. 6d. 
 Electricity, 100 years ago, $s. 
 
 in its application to Telegraphy, 
 
 Herbert, 3*. 6d. 
 
 in our Homes, Walker, 6s. 
 
 Static, Atkinson, "js. 6d. 
 
 Supply, Gay and Yeaman, los. 
 
 6d. 
 
 Municipal, Gibbings, los. 6d. 
 
 and Magnetism, Ashworth, 2s. 6d. 
 
 Bottone, 2s. 6d. 
 
 Gerard, 12s. 6d. 
 
 Houston, 3-r. 6d. 
 
 Maycock, 2s. 6d. 
 
 Electro-motors, Bottone, 35. [3^. 
 Electro-platers' Hand- Book, Bonney, 
 Electrolytic Analysis, Neumann, 6s. 
 Elliott's Gas and Petroleum Engines, 
 
 2S. 6d. 
 
 Industrial Electricity, 2s. 6d. 
 
 Emmet's Alternate Current Wiring, 55. 
 Engineer Draughtsmen's Work, I s. 6d. 
 
 Fitting, Homer, $s. 
 
 Locomotive, McDonnell, is. 
 
 Works Management, Barker, $s. 
 English and American Lathes, 
 
 Horner, 2is. 
 Explosives, Guttmann, 2 vols., 2/. 2s. 
 
 Farman's Auto Cars, 55. 
 
 Farrow, Specifications for Building 
 
 Works, 3-y. 6d. net. 
 
 Stresses and Strains, $s. net. 
 
 Field Work and Instruments, 
 
 Walmisley, 6s. net. 
 Findlay's English Railway, 7-y. 6d. 
 Fitting, Horner, $s. v 
 
 Electric, Maycock, 6s. 
 
 Electric Light, Allsop, 55. 
 
 Fletcher's Architectural Hygiene, $s. 
 
 Steam-jacket, Js. 6d. 
 
 Foden's Mechanical Tables, is. 6d. 
 
 Forestry, Webster, 3*. 6d. 
 
 Foster's Central Station Bookkeeping, 
 
 I2s. 6d. 
 Friction and its Reduction, Wheeler. 
 
 Galvanic Batteries, Bottone, $s. 
 Garratt's Modern Safety Bicycle, 3^. 
 Gas and Petroleum Engines, 2s. 6d. 
 Gaseous Fuel, Thwaite, is. 6d. 
 Gay and Yeaman's C. S. Electricity 
 
 Supply, los. 6d. 
 Gearing, Helical, Horner, $s. 
 Geometrical Optics, Blakesley. 
 Geology, Jukes-Browne, 2s. 6J. 
 
 Gerard's Electricity and Magnetism, 
 1 2s. 6d. 
 
 German Technological Dictionary, 
 Webber, 4*. net. 
 
 Gibbings' Dynamo Attendants, is. 
 
 Electricity Supply, los. 6d. 
 
 Gray's Electrical Influence Machines. 
 
 Greenwell and Elsden, Road Con- 
 struction, 5^ net. 
 
 Griffiths' Manures, fs. 6d. 
 
 Guttmann's Explosives, 2 vols.,2/. 2s. 
 
 Hanchett's Electric Railway Motors, 
 IOS. 
 
 Harris'Practical Chemistry, 3 vols. 4?. 
 
 Hatch's Mineralogy, 2s. 6d. 
 
 Hatton's Mathematics, 2s. 6d. and 3^. 
 
 Hawkins' and Wallis's Dynamo. 
 
 Heat Engines, Anderson, 6s. 
 
 Helical Gears, Horner, 5^. 
 
 Herbert's Telegraphy, 3^. 6d. 
 
 Telephone System of the Post 
 
 Office, 3J. 6d. 
 
 Hering, Electric Railways, 5-r. [15^. 
 
 Herrick's Electric Railway Handbook, 
 
 Hertzian Waves and Wireless Tele- 
 graphy, Bottone, $s. [tion, 45. 6d. 
 
 Hewitt's Organic Chemical Manipula- 
 
 Hobbs' Electrical Arithmetic, is. 
 
 Hoblyn's Medical Dictionary, los. 6d. 
 
 Hopkinson's Dynamo Machinery, 5^. 
 
 Horner's Helical Gears, 5-r. 
 
 Iron-founding, 4^. 
 
 Metal Turning, 45. 
 
 Pattern Making, 3^. 6d. 
 
 Principles of Fitting, $s. [2is. 
 
 English and American Lathes, 
 
 Hospitaller's Polyphased Alternating 
 Currents, 3^. 6d. 
 
 House Framing, Maginnis, $s. 
 
 Houston's Electrical Terms, 31*. 6d. 
 
 Electrical Measurements, 5J-. 
 
 Electricity and Magnetism, 3 s.6d. 
 
 Transmission of Intelligence, 5*. 
 
 Hurter's Alkali Makers' Handbook, 
 los. 6d. 
 
 Hutton's Mathematical Tables, I2s. 
 
 Hydraulic Motors and Turbines, 
 Bodmer, 15*. 
 
 Incandescent Lighting, Houston, 5.5-. 
 Indicator Handbook, Pickworth, 2 
 
 vols., 3^. each net. 
 Induction Coils, Bonney, 3^. 
 Induction Motor, Behrend, 71. 6J. 
 Industrial Electricity, 2s. 6d. 
 
YOURS FAITHFULLY, 
 
 DAVID L. SALOMONS, BT. 
 
EXPERIMENTS WITH 
 VACUUM TUBES 
 
 BY 
 
 SIR DAVID L. SALOMONS, BART., M.A. 
 
 PAST VICE-PRESIDENT OF THE INSTITUTION OF ELECTRICAL ENGINEERS, 
 
 FELLOW OF THE ROYAL ASTRONOMICAL SOCIETY, 
 
 ETC., ETC., ETC. 
 
 WITH 54 ILLUSTRATIONS 
 
 LONDON 
 
 WHITTAKER AND CO. 
 
 2, WHITE HART STREET, PATERNOSTER SQUARE, E.G. 
 NEW YORK : 66, FIFTH AVENUE 
 
 1903 
 
 All rights reserved 
 
Q c 7 
 
PREFACE 
 
 BEFORE leaving Cambridge, I had the good fortune to 
 meet the late Mr. Spottiswoode, who charmed me by his 
 beautiful experiments with vacuum tubes. Prior to that 
 time I took a great interest in the subject, partly from 
 the beauty of the phenomena, and partly because they 
 seemed to baffle all efforts to unravel them. 
 
 From that day to this the subject has engrossed 
 my attention, for I believed that a great deal could be 
 done with these tubes, if only the right methods were 
 devised. 
 
 At the time when Messrs. Gassiot, Warren De La Rue, 
 Spottiswoode, Moulton and others were at work, their 
 attempts to get at the bottom of the phenomena were 
 conducted by means of high E.M.F. and large currents. 
 In fact, the very methods which they pursued confused 
 the results, though interesting and instructive, and the 
 procedure they adopted can only be compared to the fable 
 of the " Frog and the Bull." 
 
 Coil after coil was made, each one larger than the pre- 
 ceding one, until the final one gave out (that at the 
 Polytechnic), and no useful results were obtained. 
 
 128145 
 
vi PREFACE 
 
 For a time this class of work was dropped, probably 
 because the idea obtained that nothing further could be 
 done, or that nothing serviceable could result. Then 
 Sir William Crookes entered the field with his new ex- 
 periments, termed by him " radiant matter " phenomena, 
 which so charmed the scientific world when the experi- 
 ments were first shown. 
 
 More recently, Professor J. J. Thompson in his brilliant 
 researches at Cambridge has been employing vacuum 
 tubes. 
 
 For spectrum analysis the vacuum tube has always been 
 of service. 
 
 The numerous experiments undertaken in the past were 
 made chiefly to ascertain the cause for the formation of 
 bands, and for the examination of the phenomena of electric 
 discharge in rarefied gases. 
 
 Latterly modified Crookes tubes have been largely 
 employed for X-ray work. 
 
 This little volume deals with a subject which appeals 
 to a very small portion of the Public, and even to a 
 limited number of those engaged in Science. 
 
 However, the results of the experiments about to be 
 described, may prove of some interest to those who are 
 working in the particular line with which this book 
 treats. 
 
 I, therefore, for the benefit of the few, deem it right 
 to publish the series of experiments which represent the 
 labour of many years. 
 
 A very large number of experiments made with ordinary 
 and with Crookes' vacuum tubes are not published here 
 
PREFACE vii 
 
 since the series is not yet complete, but I trust that, in 
 due course, a supplementary volume will appear describing 
 them. 
 
 It will be noticed in the following descriptions, that a 
 totally different method has been followed to that used by 
 those who have worked in the past, and the advantages to 
 be derived therefrom are evident from the results. 
 
 I have made no attempt to propound a theory, though 
 tempted to suggest possibilities in some instances. Such 
 suggestions must not be taken to imply that I necessarily 
 think no other reason or explanation as good or even 
 better might be advanced, and they are offered to connect 
 the chain of the subject in some intelligible manner. 
 
EXPERIMENTS WITH VACUUM TUBES 
 
 PART I 
 
 THE vacuum tube, so called, is, as we all know, not a 
 vacuum tube, but a glass tube containing rarefied gas or 
 mixed gases. Therefore when speaking of vacuum tubes I 
 do so in its common sense and not as a definition of the 
 apparatus. 
 
 The chief advantage gained from the uses of vacuum 
 tubes, is in the fact that it is possible to see what is going 
 on without requiring supplementary apparatus. 
 
 To make comparisons between currents passing through 
 such tubes and through other conductors, it must first be 
 shown that the manner in which the current travels through 
 rarefied gases is similar in some, if not in all, respects to 
 the manner in which it travels through other conductors. 
 
 It is now believed that the current travels through a gas 
 after the same fashion as it does through a liquid. It is 
 further supposed that the manner in which a current 
 travels through a solid conductor, is different to the way in 
 which it traverses a liquid. 
 
 It is quite possible that the current flows through a 
 solid in the same way as it does through a liquid, though, 
 of course, other phenomena may at the same time be 
 produced due to the substance being a solid, which mask 
 the result. 
 
2 EXPERIMENTS WITH VACUUM TUBES 
 
 In the year 1883 I conducted some exhaustive experi- 
 ments with a view to ascertain whether a metal conductor 
 was affected in any way by the passage of the current. 
 Large and small wires were employed with various currents. 
 The metals selected were copper, iron, German silver, 
 platinum and silver. In all cases the metals were 
 rendered crystalline and brittle. Authorities differ in 
 their views as to whether the passage of the electric cur- 
 rent renders a metal crystalline or not. I regard my 
 experiments as conclusive that such is the case, but of 
 course it is quite possible that the change may be pro- 
 duced by incidents due to the traverse of the current, and 
 not to the current itself. 
 
 It seems difficult to conceive a reason why the current 
 should pass from one point to another, through a gas, 
 liquid, or solid, in completely different ways. It would 
 appear more probable that the propagation in all cases is 
 alike, and the differences in the phenomena are more hidden 
 than real. 
 
 Whether such method of advance in all cases is the 
 same or not, it is impossible to deny that numerous pheno- 
 mena are common to the three classes named, and the 
 phenomena which are common, are those which may be 
 regarded as the most important. 
 
 We all know that, generally, bands of light are formed in 
 a vacuum tube, which are stationary, if the E.M.F. of the cur- 
 rent is absolutely steady, and these are present whether the 
 current be a direct, an intermittent-direct, or an alternate 
 one. We do not, however, know for certain whether the 
 analogous phenomenon is produced in solid conductors. 
 Professor Fleming has found that, if powdered coke and 
 magnesia or chalk are mixed, they will, on passing an alter- 
 native current through them, become separated into layers, 
 somewhat analogous to the bands produced in vacuum 
 
EXPERIMENTS WITH VACUUM TUBES 3 
 
 tubes ; but the cause may be different. lu some of the 
 Tesla experiments nodes are produced in the conductors, 
 which may have some kind of relationship to the bands. 
 
 These results may or may not be a fair analogy, but of 
 one thing we may be confident. It is, that the bright 
 bands are due to heated matter and are not visible 
 electricity. The very fact that these bands can be seen 
 from all points of view, is sufficiently conclusive that such 
 is the case. 
 
 That it is the passage of the current which causes these 
 bands to appear, is also certain. If a single discharge is 
 passed through a tube, the tube appears lit throughout for 
 the moment and there the phenomenon ends. This is due 
 to the persistence of vision. If the tube is examined with 
 a rotating mirror, it is found that a band of light advances 
 from end to end, at a speed approaching the rate at whicli 
 light travels ; yet when discharges are passed in rapid 
 succession through the tube, the bands become more or 
 less stationary, without any signs of advance. 
 
 A vast number of experiments have been made, to 
 ascertain the nature of these bands. It would appear to 
 me, that self-induction is probably the reason for their 
 appearance; and, if I may so express myself, it is the 
 various checks given to the advance of the successive dis- 
 charges, which cause the matter in the tube to become 
 heated to a high temperature ; and the current in the 
 darker portions travels freely without heating the matter 
 present, or only to a slight degree. The current from an 
 induction coil secondary is oscillatory as well as intermit- 
 tent, which fact should not be lost sight of. Resonance 
 would also be an explanation for the appearance of bands. 
 
 Rightly or wrongly, it appears to me that the current 
 passing through the vacuum tube, when bands are pro- 
 duced, is analogous to a lightning discharge through a very 
 
4 EXPERIMENTS WITH VACUUM TUBES 
 
 good conductor, in which case self-induction retards the 
 passage of the current. 
 
 Professor Hughes some years ago, was the first to point 
 out a number of new phenomena in connection with 
 unsteady currents, and to suggest that there would pro- 
 bably be an advantage in using iron, instead of copper, for 
 lightning rods ; and the vacuum tube experiments appear 
 to me one of the means for confirming this view. 
 
 By varying the exhaustion in a vacuum tube, and con- 
 sequently its resistance, many of these effects can be shown. 
 I am thus led to the conclusion that it is quite possible, 
 we see in the vacuum tube at least some of the phenomena 
 which really do occur in other conductors. To put this 
 more plainly in the instance given, if a tube is very little 
 exhausted, the discharge has no band. When the resist- 
 ance of the tube is greatly diminished, by further ex- 
 haustion, the bands appear, provided a suitable tube be 
 chosen. Although I do not assume that the method of 
 current propagation in rarefied gases and metal conductors 
 is identical, I do think that there is much in common, 
 especially if the modern view be accepted that the 
 current travels by means of the dielectric and sinks into 
 the conductor. 
 
 To investigate this subject, with practically no given 
 data, the simplest course to pursue, was to obtain a large 
 number of tubes, and sort them according to some pheno- 
 menon which appeared common to each set. It then 
 occurred to me that by the adoption of the method, 
 which has proved so successful in mathematics, viz. that 
 of diminishing the quantity under examination to near 
 the limit of its existence, might give better results than 
 the methods employed by other investigators who tried 
 rather to make the values infinite, were it possible. 
 
 By employing very small E.M.F. and very small 
 
EXPERIMENTS WITH VACUUM TUBES 5 
 
 current, the difficulties which beset earlier investigators 
 disappeared, and the phenomena were seen in their 
 purity. Then on raising the E.M.F. successive super- 
 added phenomena appeared, which would have been ex- 
 ceedingly puzzling, had the start not been made in the 
 manner mentioned. 
 
 Briefly the following facts came to light : 
 1. That the band produced, when the current passes 
 over any small solid body, such as a disc glass, is double 
 and not single. 
 
 A = Bright Bands. B = Dark Spaces. 
 
 Bright bands A expanded overlap at dark spaces B, which now appear 
 twice as bright as at A, and the spaces A appear dark by contrast. 
 
 FIG. 1. 
 
 2. That this double band may be considerably separated 
 by elongating the disc, say, to let the disc become a 
 short or a long rod. A long electrode also produces 
 the effect. 
 
 3. That when the E.M.F. is not exceedingly small, the 
 bright bands in the tube are spurious and are really 
 situated at the dark spaces ; which is due to the fact that 
 the bright bands have expanded and overlapped, and that 
 the dark spaces, which are the expanded bright bands, are 
 only dark by contrast with the overlaps. (See Fig. 1.) 
 
 4. That this overlapping may be repeated again and 
 
6 EXPERIMENTS WITH VACUUM TUBES 
 
 again as the E.M.F. is raised, the limit being the destruc- 
 tion of the tube. 
 
 5. That, if there is inserted at one end of the tube a 
 rod, disc, contraction or diaphragm, or any other equivalent 
 arrangement, bright bands are formed throughout the tube, 
 at a distance apart ; which is governed by the nature of 
 the arrangement placed in the end of the tube. 
 
 6. That the glass of the tube appears to act, in some 
 way, in forming the bands. 
 
 7. In a free tube the number of alternations may vary 
 the number of the bands. 
 
 The knowledge of these various effects is a great assist- 
 ance in pursuing this subject. It becomes possible to make 
 
 X 
 
 7- 
 
 FIG. 2. 
 
 two or more tubes to give precisely the same phenomena. 
 It is possible to make tubes giving narrow or wide bands 
 at pleasure, and many of the experiments which will be 
 spoken of later could not be performed with certainty 
 unless such tubes could be made. Formerly it might 
 have become necessary to manufacture 100 tubes before 
 finding one that would answer the purpose ; and then to 
 produce a second one, the same process would have to be 
 gone through. 
 
 I do not wish to say more on the method of making the 
 tubes beyond giving a few examples, as I have already 
 published Papers on the subject, which are to be found in 
 the Proceedings of the Royal Society, vol. 5G, and elsewhere, 
 and there all details are given. 
 
 Figs. 2, 3, and 4 represent vacuum tubes, of various 
 
EXPERIMENTS WITH VACUUM TUBES 1 
 
 types, in each of which a little glass apparatus is inserted 
 at one end. The bands will be produced throughout the 
 
 tao n / 
 FIG. 3. 
 
 FIG. 4. 
 
 length of the free tube, at a distance apart equal to the 
 distance between the glass discs. 
 
 Fig. 5 represents a special form of glass tube, wherein 
 
 the bands will be produced at a distance apart, approxi- 
 mately equal to the length of the little rod seen at the 
 end of the tube. 
 
 Fig. G shows a tube with a contraction. The bands 
 will appear apart, at a distance equal to the length of the 
 contraction. 
 
8 EXPERIMENTS WITH VACUUM TUBES 
 
 Fig. 7 illustrates the tube with a diaphragm for produc- 
 ing bands at given distances, which in this case will be 
 
 300 "* , 
 
 330 " 
 
 FIG. 6. 
 
 approximately equal to the distance beween the diaphragm 
 and the electrode nearest to it. 
 
 Fig. 8 shows a tube lit by induction, there being no 
 connection between the electrodes and the free portion of 
 
 430 m/ 
 
 FIG. 8. 
 
 the tube, and the bands are formed at a distance apart 
 which is regulated by the size of the inductive glass spheres. 
 Fig. 9 is a tube in which a little glass disc can be made 
 to stand up, or lie down, at pleasure. The method of 
 regulating the bands is perfectly shown in such a tube, 
 
VN1 
 
 EXPERIMENTS WITH VACUUM TUBES 9 
 
 using it first as an open tube, and then as one having a 
 special arrangement for making the bands appear at a 
 given distance apart. In this case the distance will be 
 equal to the distance between the glass disc and its nearest 
 electrode. 
 
 FIG. 9. 
 
 Professor Fleming, after seeing the tubes with fixed 
 discs, etc., suggested that a tube might be constructed to 
 show the action of what I term the "deflector" in one 
 tube, and in consequence the form Fig. 9 was made. 
 
 If in a tube in which bands are formed, bulbs are blown 
 along its length, the bands refuse to appear in the bulbs, 
 which may be taken as an indication that the glass plays 
 
 Fro. 10. 
 
 some part in their formation. If the tube consists of a 
 large glass bulb in the course of a tube, it will be observed 
 that little bands appear at an entrance to the bulb. They 
 stick to the entrance, and are bulged into the bulb, and it 
 requires some coaxing (increase of E.M.F.) to induce these 
 bands to form in the bulb. 
 
 Fig. 10 is a tube with bulbs blown along its course. 
 The bands will not form in the bulbs. 
 
10 
 
 EXPERIMENTS WITH VACUUM TUBES 
 
 Fig. 11 exhibits a tube lit inductively, having a large 
 sphere in the centre of its course, 
 
 FIG. 11. 
 
 Fig. 12 shows the position of the bands with a small 
 current passed in tube Fig. 11. 
 
 Fig. 13 gives the appearance of the bands when the 
 current is greatly increased in same tube. This experi- 
 
 First Stage. 
 
 FIG. 12. 
 
 Second Stage. 
 (More current passed.) 
 
 FIG. 13. 
 
 ment well illustrates the tendency of the bands to stick to 
 the glass, and bears out my contention that the dielectric 
 has a considerable influence in their formation. 
 
 Then with the " electric egg," which is so well known, 
 on being exhausted to a considerable degree, the current 
 passes in the centre as a column of light, and if matter is 
 inserted in any quantity, which is usually done by intro- 
 ducing the vapour of alcohol, of naphtha, or of any other 
 volatile fluid, bands will appear. 
 
EXPERIMENTS WITH VACUUM TUBES 
 
 11 
 
 This would indicate that the matter present enables 
 the current to reach the glass. Although frequently the 
 bands may seem not to reach the glass, it can generally be 
 shown that this is the case when they are present, but 
 the fact of the eye not always seeing the effect does not 
 necessarily prove that it is non-existent. 
 
 It might be imagined, from what I have said, that it is 
 not uncommon for bands to be produced in a globe or tube 
 
 FIG. 14. 
 
 without reaching the sides of the glass, but it will be found 
 no easy matter to produce the effect. 
 
 It must be remembered that in the electric egg the 
 glass sides are very far removed from the electrodes, 
 whereas in the case of the tube mentioned this is not so. 
 But if, however, in a tube, say 300 mm. and 40 mm. diameter, 
 
 FIG. 15. 
 
 a little metal plate is placed across the tube, at about the 
 centre of its length, this acts as a bridge, and the current 
 discharge will now be very similar to, yet not identical 
 with, that which takes place in the electric egg. 
 
 The following experiments illustrate the foregoing 
 points. Fig. 14 contains rarefied air, and on passing a 
 current a faint, unstriated light makes its appearance. 
 Fig. 15 is a similar tube, but some alcohol vapour is 
 
12 EXPERIMENTS WITH VACUUM TUBES 
 
 introduced. In this case closely-packed bands are 
 produced. Fig. 16 is, again, a similar tube with a 
 platinum plate across the tube. When the current is 
 passed, the discharge is almost similar to that in the 
 electric egg. 
 
 The experiment of producing a radial discharge in 
 vacuo, to give an appearance somewhat similar to that of 
 " Newton's Kings," is a difficult one to produce, on account 
 of the great care necessary in the construction of the 
 
 FIG. 16. 
 
 apparatus. After many failures I succeeded in producing 
 a tube which gives the desired result. 
 
 Two hemispheres of glass, the edges being carefully 
 ground, are placed together with suitable grease, and they 
 are so arranged that what becomes a glass sphere can be 
 placed on an air pump to be exhausted. 
 
 Within this sphere and from below there is supported 
 by a metal rod, surrounded by ebonite, a flat gun-metal 
 saucer, in the bottom of which is placed a sheet of glass. 
 The exposed inside of the saucer rim is employed as an 
 electrode. 
 
 There is another rod in the upper portion of the 
 sphere, also protected by ebonite, the lower end of the rod 
 passing through a hole in a circular piece of glass and has 
 a small cylindrical nut screwed on it to keep the glass on 
 the rod, and this circular piece of glass fits exactly in the 
 gun-metal saucer, and thus the small nut on the end of 
 the rod is truly central with the inside rim of the saucer. 
 
EXPERIMENTS WITH VACUUM TUBES 13 
 
 Consequently there is a central electrode, viz. the nut 
 at the end of the rod, and a circular electrode being the 
 inside of the rim of the saucer so much as is exposed 
 between the two glass plates. 
 
 FIG. 17. 
 
 This space between the glass plates is equivalent to a 
 vacuum tube, and suitable connections exist outside the 
 glass sphere in order to make the connections with the 
 secondary of the induction coil. Fig. 17 shows the 
 
 arrangement. 
 
14 EXPERIMENTS WITH VACUUM TUBES 
 
 A very small E.M.F. has to be employed in this 
 experiment, the spark in air not exceeding 0'15 mm. 
 
 If the room is quite dark the whole of what appears 
 to be the inside of the saucer becomes illuminated. To 
 better see the effect, the inside of the saucer is blackened. 
 Every now and then, if the phenomenon is carefully 
 watched, distinct striation is visible, something in appear- 
 ance to the well-known " Newton's Rings." 
 
 If a vacuum tube is so constructed as to have another 
 tube within it (see Fig. 18), also exhausted, with two sets 
 of electrodes, one set for the outer tube and the other set 
 for the inner one, no matter whether the outer tube is more 
 exhausted than the inner one, or the inverse, provided a 
 
 cfe 
 
 FIG. 18. 
 
 spark in air of a given length would pass through either 
 tube, it will be found, on lighting up each tube successively 
 that bands form in each tube, which can be resolved when 
 placed in the magnetic field (I shall refer to this subject 
 later), and this procedure renders it more clear that there 
 are bands in the inner tube. 
 
 If the experiment is made by joining together the pairs 
 of electrodes at each end, then, according to the general 
 law, particularly if the resistance of the outer tube is made 
 less than that of the inner one, the current should travel 
 through the outside tube, very little through the inner 
 tube, but this is not the case. 
 
 This experiment does not contradict the general or 
 Faraday's law, but offers what I believe to be a further 
 proof that the glass, i.e. dielectric of the vacuum tube, is 
 
EXPERIMENTS WITH VACUUM TUBES 15 
 
 necessary to produce the bands ; also a further proof of the 
 present view that the current travels through the di- 
 electric, and sinks into the conductor. The current in this 
 experiment practically enters both tubes at the same 
 point, and refuses to leave the dielectrics. 
 
 Now, had the inner tube been constructed of metal, would 
 the current appear in the outer tube ? 
 
 To settle this point such a tube was made, notwith- 
 standing the structural difficulties which presented them- 
 selves. A current of very high E.M.F. (6 inch spark in 
 air) was passed through the inner tube. Naturally the 
 current sparked from the electrodes to the metal tube, and 
 this latter conducted the current. At the same time the 
 outer tube glowed, instead of remaining dark. This 
 experiment may be described as a modified form of the 
 " hemisphere " experiment, employed in static electricity, 
 applied to dynamic electricity. 
 
 I think that I have so far said enough to show that the 
 glass, for some reason or another, is an active agency in 
 producing the bands. If I may be permitted to express 
 an opinion which may on further experiment be modified, 
 it appears to me that the vacuum tube has a strong 
 analogy to the ordinary metal conductor. The glass takes 
 the place of the dielectric, and the gas within represents 
 the metal. The current travels along the dielectric and 
 sinks into the gas ; that is, the conductor. When the 
 dielectric is sufficiently distant, the discharge is of a 
 different nature and is disruptive. This latter phenomenon 
 may be absent in solid conductors, but in any case it may 
 not be possible to produce it, on account of structural 
 difficulties. 
 
 Formerly, were it required to produce bands at given 
 distances, it would have been necessary to fix, in some 
 suitable manner, solid bodies along a tube at stated places, 
 
16 EXPERIMENTS WITH VACUUM TUBES 
 
 bands being formed where the solid bodies were situated. 
 I am not aware whether any such tubes have ever been 
 made, for they would have been inconvenient for experi- 
 ment, but in order to try the effect, I constructed two 
 tubes in the following manner : Each tube had a slight 
 glass rod placed in its axis and was supported from the 
 ends of the tube. In one tube little discs of glass were 
 placed at equal distances along the rod, and in the other 
 tube the discs were replaced by small glass spheres. 
 
 FIG. 19. 
 
 Fir,. 20. 
 
 These tubes were lit up inductively, otherwise the effects 
 could not be produced with certainty. 
 
 These two tubes are shown in Figs. 10 and 20. It will 
 be observed that a bright band appears on each side of 
 every glass disc, but in the case of the glass spheres, it will 
 be found that four bands are produced on each sphere : 
 one on either side of the equator of the sphere, and one at 
 each pole. This is worth notice, because it shows that the 
 sphere acts as a disc in the plane of its equator, and as a 
 rod the length of its polar axis. 
 
 I have no doubt that some pretty deductions could 
 be made from this phenomenon, which must be clearly 
 due to certain properties possessed by a sphere. 
 
 I might mention that, in the case of the sphere, under 
 
EXPERIMENTS WITH VACUUM TUBES 17 
 
 no condition can bands be produced differently ; provided 
 the volume of the sphere bears a comparatively small 
 relation to the volume of the tube. Other methods for 
 producing bands at given distances have been employed, 
 but they are cumbrous and unpractical. 
 
 At one time my electric energy was drawn from an 
 alternate current source, as well as from an induction coil ; 
 but I found it possible to produce with the induction coil 
 a current identical in nature with that produced from the 
 alternate current source. When the induction coil is 
 employed in the customary manner, the chief current is 
 produced on the "break." Consequently, the tube is lit 
 up by a current of one kind which is intermittent, the 
 opposite current being so faint as to be barely visible, if at 
 all ; but the coil can be so arranged that the " make " cur- 
 rent may be rendered visible. The " break" current, how- 
 ever, is always the most important. If the source of current 
 supply for an induction coil has a high E.M.F., higher than 
 that for which the coil was constructed, a resistance being 
 inserted to prevent too much current from passing through 
 the primary, then a different effect will be produced. 
 
 In my own case I employ a 100 volt current on a 
 primary constructed for 10 volts. The " break spark " on 
 the " contact maker," is then much larger, also a spark is 
 created on the " make," and the final result is that the 
 effects, produced in the vacuum tube, are identical with 
 those when the alternate current was used, i.e. the " break " 
 current is diminished and the " make " current shows it- 
 self. This is not a condition of the coil which is unstable, 
 but is a created condition which is a stable one and can be 
 reproduced at pleasure in any coil. 
 
 A separate " make-and-break " apparatus is better for 
 this class of work than the common form of vibrator. The 
 currents used had an E.M.F. as low as 500 volts ; and frorn 
 
18 EXPERIMENTS WITH VACUUM TUBES 
 
 this pressure to 1000 volts every phenomenon can be pro- 
 duced in its purity. 
 
 It must not be forgotten that, when the two currents 
 are seen in the tube at the same time, they really follow 
 one another. The rotating mirror will prove this to any 
 one who does not realize the fact. 
 
 It might also be mentioned that, when very small cur- 
 rents are used, the discharges are in what has been termed 
 the "sensitive state." 
 
 The exhaustion for these low pressures must not be very 
 high, and there is no actual need that it should be ; nor is 
 it material as to which gas is used. Ordinary air is as good 
 as anything else. In saying this it is not meant that 
 for certain classes of experiment high vacua and special 
 gases, particularly the simple ones, are not desirable ; but 
 we are not dealing now with these. 
 
 Up to this point I have endeavoured only to show cer- 
 tain analogies between the current passing through the 
 vacuum tube, and the current passing through an ordinary 
 conductor such as a cable. I have not attempted to deal 
 with the considerations which arise from the mode of pro- 
 pagation, or in bringing the phenomena into line with the 
 ether theory. The latter can be done by any one having 
 a knowledge of the subject, and I would therefore sooner 
 confine myself to experiments and their deductions, or 
 probable deductions. 
 
 I must remind you that the ether, or any other theory, 
 bears a relation to Science in the same way as the fingers 
 do to a child, when they are employed in learning to 
 count. A theory is merely a mechanical idea to enable 
 our imperfect minds to understand that which would other- 
 wise be incomprehensible. 
 
 If we have at the present time, say, a million phenomena 
 which fit in with the undulatory theory, it is quite certain 
 
EXPERIMENTS WITH VACUUM TUBES 19 
 
 that there are millions of other phenomena unknown to 
 us, numbers of which may not agree with that theory. It 
 is, therefore, quite conceivable that the time may arrive 
 when some other theory will be required in the place of 
 the one which we believe to be correct at the present 
 time ; but to take up the position that all theories are 
 false, and to work without one at all, is not possible. You 
 might as well suggest a problem to be solved without going 
 through a single mathematical operation, which is quite 
 conceivable if the mind of man were perfect ; but no one 
 would suggest the abolition of mathematics on the ground 
 that the perfect mind would not need it. For this reason 
 any school which rejects all theories is doing an injury to 
 the advance of Science. 
 
 NOTE. Since the experiments described were performed, many 
 new kinds of interrupters liave appeared, such as the " mercury jet," 
 " electrolytic," and others. Their use simplifies in some degree the 
 arrangements for making the induction coil produce given desired 
 results 
 
PART II 
 
 I NOW propose to turn to the second part of the subject, 
 which to many workers may be more attractive, namely, 
 the behaviour of electric currents in the magnetic field. 
 
 In order to carry out these experiments, Messrs. Nalder 
 constructed a large magnet from my designs, and I have 
 to congratulate this firm on the admirable manner in 
 which all the details have been carried out. An illus- 
 tration of this apparatus is given in Fig. 21. The weight 
 of the magnet is about 13 cwt., and the arrangements are 
 thoroughly practical. In every detail it is a departure from 
 the old Faraday type. 
 
 The great magnet which Faraday employed was but a 
 rough instrument which he would probably have modified 
 when building another. I found it very difficult to work 
 with this type of apparatus, and therefore decided to build 
 one on engineering lines. In this magnet the poles can 
 be approached and receded by screws, and small pole- 
 pieces of various types can be fixed to the poles in the 
 same manner as is done in the case of chucks on a watch- 
 maker's lathe. Hollow poles also exist with microscope 
 and polariscope fitted. The suspension device is very con- 
 venient, and removable at pleasure. Switches for making 
 both poles alike, and for commutating the current are added, 
 in addition to the usual switch for making and breaking 
 the current. Before breaking the circuit a resistance is 
 
 20 
 
EXPERIMENTS WITH VACUUM TUBES 21 
 placed in parallel with the coils, and this is done by the 
 
 movement of the main switch, and not by a supplementary 
 
 one. 
 
 The illustration (Fig. 21) shows the magnet with the 
 
22 EXPERIMENTS WITH VACUUM TUBES 
 
 glass case removed. There are many pieces of supple- 
 mentary apparatus, one of which converts the magnet into 
 a huge D'Arsonval galvanometer. The magnet is excited 
 with three amperes from 100 volt circuit. 
 
 To this date very little appears to have been published 
 on the effects produced upon electric discharges in rarefied 
 gases when placed in the magnetic field. That certain 
 effects are produced under such conditions is known, but 
 not much more has been published, as far as I can 
 ascertain. The rotation of the discharge about the pole of 
 a magnet is also known. 
 
 Experiments have likewise been made to show that there 
 is a slight displacement of current in a conductor, when 
 placed in the magnetic field. I believe this was shown 
 many years after the effect had been seen in my laboratory 
 on a much larger scale than that which was published. 
 
 I will now describe certain phenomena shown by 
 vacuum tubes in the magnetic field. 
 
 The tubes as generally made are of little service for 
 this purpose. Those which have already been described 
 are better, also tubes about 300 mm. long and 16 to 25 mm. 
 in diameter, with suitable brush electrodes, the object 
 being to obtain distinct bands at some given convenient 
 distance apart. The E.M.F. employed in connection with 
 these tubes is very low, rarely exceeding 1000 volts, and 
 the induction coil is arranged to give effects in the tubes 
 the same as if an alternate current had been employed 
 from a transformer. 
 
 If such a tube as described (see Fig. 22) is placed between 
 the poles of the magnet when excited, one pole being 
 north and the other south (Fig. 23), the two currents 
 separate in the tube equatorially. 1 The appearance is that 
 of an unlit tube with two lines of bands clinging to the 
 
 1 See Appendix for meaning of " equatorially " and " axially." 
 
EXPERIMENTS WITH VACUUM TUBES 23 
 
 tube. (See Fig. 24, the dotted lines indicate the position 
 of the bands.) 
 
 FIG. 22 
 
 FIG. 23. 
 
 I will describe the appearance more minutely; for 
 convenience let it be assumed that the two currents are 
 
24 EXPERIMENTS WITH VACUUM 
 
 visible in the tube. There will then be seen the two lines 
 of bands placed equatorially. On examination it will be 
 observed that the only difference between the column of 
 bands on one side and those on the other is in the direction 
 of the convex sides of the bands. 
 
 FIG. 24. 
 
 Consider only one side : it will be noticed that the 
 bands become smaller and smaller as they approach the 
 strongest portion of the field. At the same time they 
 approach one another until the bands appear to be a mere 
 line of light, yet with a magnifying glass the bands can 
 be still observed. 
 
 Fig. 25 gives a rough idea of what takes place. Only 
 one line of bands is shown. 
 
 It is therefore evident that the two currents tend to 
 
 FIG. 25. 
 
 pass out of the magnetic field, and each current takes a 
 definite position in regard to the field. If the current in 
 the tube is reversed the appearance in the tube will be 
 the same, that is, similar currents occupy the same 
 positions, but if the poles of the magnet are reversed, 
 the convexities of the bands reverse also, i. e. the currents 
 change sides. 
 
 That the currents are separated is certain, but the 
 absence of light in the centre of the tube, i. e. in the 
 strongest part of the field, may be due to three causes, if 
 
EXPERIMENTS WITH VACUUM TUBES 25 
 
 not more. It may be that the resistance of the gas has 
 been increased, and the current therefore travels at places 
 of less resistance, or the resistance of the gas may 
 remain unaltered, and it is simply that the two currents 
 desire to take up a particular position in regard to the 
 field, or both conditions may exist. 
 
 A tube under these conditions will show a higher resist- 
 ance when the magnet is excited than when unexcited, 
 but on reflection this is not necessarily a proof that the 
 resistance of the gas is actually increased. It may be a 
 spurious increase in the resistance. 
 
 It must not be forgotten that a magnet is equivalent to 
 a current travelling in a circular conductor, and that the 
 current travelling through the tube is one in a straight 
 and movable conductor. 
 
 In other words, this experiment is a pretty and convinc- 
 ing proof of Ampere's theory as shown with movable 
 conductors, which has hitherto been made in a different 
 and more clumsy manner. 
 
 It appears to me that the true explanation why the two 
 currents move out of the field, is due rather to this 
 Amperian explanation than to the actual increase in the 
 resistance of the gas. 
 
 If the induction coil is now supplied with another 
 source of current, having a lower E.M.F., the usual state 
 of things is produced in the tube, which can be so 
 arranged that only the break current becomes visible. 
 On placing the tube in the field, only one column of 
 bands will appear and not the two. Whether the arrange- 
 ments are such that one or both currents are visible in the 
 tube, on reducing the current considerably the time is 
 reached, when although the tube may be lit throughout 
 when held outside the magnetic field, yet on placing the 
 tube between the poles, the passage of the current ceases. 
 
26 EXPERIMENTS WITH VACUUM TUBES 
 
 This would indicate at first sight that the resistance of 
 the gas is increased. Consequently, in the first experi- 
 ment described, not only did the currents tend to take up 
 definite position in respect to the field, but an increased 
 resistance of the gas assisted the phenomenon. Another 
 consideration may also enter, and it will be referred to 
 later. 
 
 I am unable to say positively whether this is the case 
 or not, but my various experiments lead me to think that 
 when the current is exceedingly small, the reason why it 
 does not pass through the tube when under magnetic 
 influence is due to the current being, so to speak, "crowded 
 out," i. e. driven as it were into the glass of the tube, which, 
 having such a high resistance, stops it passing, as far as can 
 be observed, on account of the low E.M.F. of the current. 
 
 There are a large number of other interesting experi- 
 ments which can be made by holding this tube in various 
 positions, below, near, and around the poles of the magnet, 
 but as they can all be reduced to the principles I have 
 mentioned, I will not occupy more time in describing 
 them, beyond giving a few diagrams to show what occurs. 
 
 I have already described the appearance of the tube 
 when placed between the poles N. and S., when the 
 magnet is excited. 
 
 If, now, the current is turned off the magnet, the 
 magnetism falls gradually, and the -successive stages of 
 the phenomena will manifest themselves in the tube from 
 the strongest field to zero. This is the best method of 
 making the observations, on account of the process being 
 gradual. 
 
 Figs. 26, 27, 28 and 29 show the various phenomena 
 when the tube is between the poles, the bands seen in 
 Fig. 29 are their position when the magnet has lost its 
 field, and Figs. 30, 31 and 32 illustrate the appearance 
 
EXPERIMENTS WITH VACUUM TUBES 27 
 
 FIG. 26. 
 
 FIG. 27. 
 
 FIG. 28. 
 
 FIG. 29. 
 
 FIG. 30. 
 
 of the tube in other positions near the magnet when 
 excited. 
 
28 
 
 EXPERIMENTS WITH VACUUM TUBES 
 
 One important use is open for vacuum tubes. It is that 
 of exploring the magnetic field of dynamos, alternators, 
 
 + 
 
 FIG. 31. 
 
 or for any other class of practical work of this nature as 
 well as for theoretical investigation. A very small coil 
 and tube will suffice for such testing. 
 
 FIG. 32. 
 
EXPERIMENTS WITH VACUUM TUBES 29 
 
 It is desirable to point out, that when there is a free N. 
 and S. pole, anywhere in the proximity of the tube there 
 are banded lines of light on each side of the tube when 
 the two currents are present, and a single band of light 
 when only one current is present, and the position taken up 
 will always be situated equatorially to the lines of force. 
 Consequently the position of the light in the tube is the 
 test for lines of force. 
 
 The smallness of the bands, and their closeness, gives 
 the strength of the field. 
 
 Considering for a moment one line of bands, in all cases 
 this line will be straight when the field is symmetrical. 
 When unsymmetrical the line will be bent into various 
 curves, generally spiral. 
 
 ( ( ( 
 
 ' ' 
 
 FIG. 33. 
 
 By this method of testing, the field can be mentally 
 mapped out in a few seconds, without the necessity of 
 employing the various apparatus as at present used, where 
 the results have to be set down and plotted before any 
 idea can be formed of what is going on. 
 
 Fig. 33 roughly illustrates the appearance of the tube 
 when held in an unsymmetrical field. 
 
 We will now examine what phenomena occur when the 
 tube is placed between the poles of the magnet and the 
 current is increased by raising the E.M.F. 
 
 If the current traversing the tube is gradually increased, 
 although the two columns of bands will still be there, the 
 " Waist," so to speak, becomes widened, and a faint light 
 shows itself throughout the remainder of the tube. 
 
 Continue to increase the current, and if all the conditions 
 
30 EXPERIMENTS WITH VACUUM TUBES 
 
 have been suitably arranged, i. e. the magnetic field 
 sufficiently powerful, the amount of current nicely adjusted, 
 and the tube is of suitable diameter, then there will 
 suddenly start out between the poles, and within the tube, 
 bright lines which can only be described as apparently the 
 lines of force becoming visible. If this experiment is 
 carefully made, the lines of force are very brilliant and 
 unmistakable. Indeed, the effect is better shown than in 
 the old manner with iron filings, because it is exhibited 
 
 FIG. 34. 
 
 FIG. 35. 
 
 in a three dimensional manner. Fig. 34 is a general 
 diagram, and Fig. 35 illustrates the actual appearance of 
 the tube, i. e. the apparent lines of force are shown in the 
 centre of the tube, also the bands. 
 
 It is well known that the current passes with greater 
 freedom with the lines of force than across them. 
 
 The experiment showing the lines of force appears to me 
 a further proof of this phenomenon. I also regard it as a 
 strong evidence that my view as to the formation of the 
 bright bands is correct, namely, self-induction. 
 
' vm\ v 
 
 EXPERIMENTS WITH VACUUM TUBES 31 
 
 When the current has to cross the lines of force in the 
 stronger portion of the field, it meets with more resistance. 
 You have therefore a tube containing a conducting material 
 with varying resistance. It is quite conceivable that heat 
 may be developed in sufficient quantity, when crossing the 
 stronger portion of the field, to render any finely-divided 
 matter present sufficiently heated to become visible, and 
 that the reason why the imaginary lines of force appear 
 actually as lines, may be due to self-inductive effects, 
 similar in character to those which formed the bands under 
 normal conditions, that the lines are really bands tak- 
 ing upon themselves the form of the lines of force. It, 
 would then appear that it is possible, even probable, that 
 the apparent increase in the resistance of the gas in the 
 tube is due to two causes ; firstly, the current or currents, 
 if two are present, tending to take up a position outside 
 the field; and, secondly, that the current or currents prefer 
 to evade crossing lines of force, in order to find a path of 
 less resistance. Consequently it does not follow that the 
 actual resistance of the gas has become increased. It is 
 still a spurious effect. Analyzing the conditions more 
 closely, if the current is driven out of the field equatorially, 
 then the increased resistance which would be offered to the 
 current when crossing the lines of force is greatly dimin- 
 ished, because the current in this case is in a weaker 
 field. This is true probably for small currents, but for 
 larger ones the field must be made much stronger than can 
 be done in practice to produce the effect. 
 
 For a long time past it has been known that bismuth 
 shows a higher resistance, when placed in a strong mag^ 
 netic field ; I believe that Faraday was the first to show 
 this phenomenon for the metal named, and Professor 
 Sylvanus Thomson to publish the fact for other metals. 
 
 It will be evident, after describing the vacuum tube 
 
32 EXPERIMENTS WITH VACUUM TUBES 
 
 experiments, that another explanation is possible for this 
 apparent increase in the resistance of metals when placed 
 in the magnetic field. It is clear that the methods which 
 can be employed to prove either of these cases run entirely 
 in parallel, and that the question would still remain an 
 open one as to which is the true explanation, were it not 
 that in the vacuum tube, the experimenter can actually 
 see what is going on in a rarefied gas, and to some extent 
 the same phenomenon occurs in metals, as shown by the 
 experiments I have made, and I am strongly of opinion 
 that the resistance of the metals is not really increased, 
 but it is simply the current or currents which are pushed 
 out of the field, and tend to take up a position of equili- 
 brium, combined with the traverse of the lines of force, 
 which gives the apparent result of increased resistance, 
 which, in fact, may be due to the current traversing a 
 smaller section of the area. 
 
 If this be true, a curious circumstance must be noticed. 
 Generally speaking I do not know whether it is true in 
 all cases the bad conducting metals appear to increase 
 their resistance more in the magnetic field than is the 
 case with those which are better conducting. I have found 
 this result with many metals. It might therefore be 
 assumed that the current when crossing the magnetic 
 lines of force in bad conducting metals, can move with 
 greater freedom than is the case in better conducting 
 metals. 
 
 I think it more than probable, that if a careful series of 
 experiments were made, on various metals placed in the 
 magnetic field, it would be found that there is no simple 
 relation between the resistance of the metal and the 
 strength of the magnetic field, except for a given current. 
 This, indeed, would be one of the best proofs to uphold 
 the view which I believe to be correct, but I have so far 
 
EXPERIMENTS WITH VACUUM TUBES 33 
 
 not had the opportunity of pursuing this part of the subject 
 further. 
 
 To turn to one or two practical deductions, which may 
 almost be termed commercial, depending on this pheno- 
 menon, but having no bearing as to which theory is correct, 
 I may give two examples which will suggest many others. 
 
 Take the case of a coil in an alternator or a direct current 
 machine, which is continually passing through powerful 
 magnetic fields. It must be clear that at one moment 
 the current is being driven, as it were, out of the copper, 
 and the next moment, when passing through a weaker 
 field, the current density at every portion of the section of 
 the copper is more or less the same. In other words, there 
 is what may be termed a pulsating of the current in the 
 copper from without to within and the inverse, during the 
 time that the coil or the magnets are in motion, as the 
 case may be. 
 
 It would therefore appear, that apart from any other 
 conditions which may exist, there must be some slight 
 pulsation or period in the direct current as well as in the 
 alternative current, due to the cause mentioned. 
 
 The copper in all armatures, therefore, should be larger 
 in section than needed by theory to meet this variation, 
 Avhich, however, is exceedingly small, and the margin 
 of safety at present allowed more than suffices for the 
 purpose. 
 
 To give another example, it seems obvious that it is 
 undesirable in the D'Arsonval galvanometer to make the 
 field too strong. A point must be reached when the 
 amount of current is so greatly reduced in the swinging 
 coil, that the instrument becomes less sensitive. This is 
 admirably shown on the large magnet, when arranged as a 
 D'Arsonval galvanometer. The magnet is excited, and the 
 position of the dot of light observed when a very small 
 
34 EXPERIMENTS WITH VACUUM TUBES 
 
 current is passed. On breaking the circuit of the magnet, 
 the magnetism falls, and the dot of light indicates a much 
 larger deflection, the E.M.F. being kept constant on the 
 galvanometer coil terminals. This experiment can be 
 made in another way by leaving the magnet excited 
 and placing a resistance so as to reduce the current 
 flowing in the field circuit, when a similar result is 
 obtained. There are other reasons why the field should 
 not be made too strong, yet my experiments lead me 
 to the conclusion that those referred to are the main 
 causes for loss of sensibility under such conditions. 
 
 If the magnet is so arranged that both poles are made 
 N. or S., and then a thick plate of iron is laid across them, 
 and upon this plate a pole is stood upright, you obtain the 
 equivalent of a bar magnet ; one pole pointing upwards, 
 the other pole being the base of the instrument. It will 
 be found that if the vacuum tube, with which most of the 
 previous experiments were made, is placed over this pole, 
 the effects are the same as when placed between the poles 
 N. and S., but the phenomena are not quite so pronounced. 
 This result could easily have been anticipated, because the 
 lines of force close to the pole travel much in the same 
 way as when the poles were opposite one another with the 
 tube between. It is, however, instructive -to trace the 
 lines of force with a vacuum tube, when the magnet is 
 arranged as mentioned. 
 
 The phenomenon is considerably altered, if the magnet 
 is now arranged so that the poles are N.N. or S.S., and the 
 tube is placed between them. Currents within the tube 
 take up their positions separately as in the case when the 
 poles were N. and S., but now the lines of bands will be 
 placed axially, and not equatorially, as before. Moreover, 
 the lines of bands cross one another in the strongest portion 
 of the field. This is shown in Fig. 36, where to avoid con- 
 
EXPERIMENTS WITH VACUUM TUBES 35 
 
 fusion, the lines of bands are simply represented as dotted 
 lines. If the tube is placed below the poles the bands will 
 appear as in Fig. 37 ; if above, as in Fig. 38, the difference 
 
 FIG. 36. 
 
 being that they face the reverse way. If only one current 
 is employed in the tube, then the appearance of the bands, 
 when placed between N.N. or S.S., is very similar to that 
 of an unsymmetrical field, which it really is. This is an 
 
 ^ ^ ^ 
 
 \ \ X 
 
 FIG. 37. 
 
 instructive experiment in many ways, and particularly if 
 the tube is used to trace the field. 
 
 A very pretty experiment is that of taking a vacuum 
 tube about 16 mm. diameter and 500 mm. long, and 
 bending it almost into a circle, then holding it over the 
 
 FIG. 38. 
 
 poles of the magnet when N. and S. The appearance of 
 the bands will then be as shown in Fig. 39. Observe that 
 the currents cross at the strongest portion of the field. 
 Although the N. and S. poles are employed in this experi- 
 ment, the phenomenon in the circular tube is very similar 
 to that in the straight tube, when placed between N. and N. 
 
36 EXPERIMENTS WITH VACUUM TUBES 
 
 It may also be noted that if the tube containing a 
 metal tube, which was mentioned earlier, has one elec- 
 trode placed in the magnetic field, the spark from each 
 electrode to the metal tube is divided into two, equatorially, 
 as if the currents traversed the metal tube separated, which 
 appears thus to be proved. 
 
 FIG. 39. 
 
 From many of the foregoing experiments, it appeared 
 to me possible to take an alternate current and separate it 
 by magnetic means into its two constituents, i.e. two 
 pulsating direct currents. That such a current can be 
 split up by other methods is well known, but they are not 
 simple, nor can any very instructive deductions be made 
 from them. The separation by the magnetic method is 
 very interesting, for quite apart from the analysis of the 
 alternate current, a neat illustration is given of many 
 well-known laws. 
 
 We have seen how the two currents become separated 
 in the magnetic field. It appeared to me therefore prob- 
 able that if a glass septum were placed in the tube with 
 which the experiments referred to were made, such a 
 division not reaching to the electrodes, the currents would 
 
EXPERIMENTS WITH VACUUM TUBES 
 
 37 
 
 be completely separated in their course. The equivalent 
 to this construction would be a double tube. I therefore 
 constructed tubes of the shapes given in Figs. 40, 41, 42. 
 
 FIG. 40. 
 
 Then, in the case of any of these tubes, if one of the 
 bulbs containing an electrode is placed in a strong 
 magnetic field, the plane of the tube being placed 
 equatorially, complete separation is evident from the fact 
 
 FIG. 41. 
 
 that there is but one line of bands in any portion of the 
 tube. 
 
 It is interesting to note that if one of these tubes, say 
 Fig. 41, which is about 150 mm. in length, is placed 
 
38 
 
 EXPERIMENTS WITH VACUUM TUBES 
 
 equatorially, and the current so adjusted that when the 
 field is excited no current passes, on the tube being 
 turned out of the equatorial plane, the current makes its 
 appearance. With the tubes mentioned, if the excitation 
 of the field is permitted gradually to drop, the various 
 stages of the phenomena are very distinct, i. e. from the 
 
 FIG. 42. 
 
 alternate current traversing throughout the tubes to the 
 moment when the current becomes separated into two 
 pulsating direct ones. There is also a stage when the 
 
 C 
 
 FIG. 43. 
 
 discharge appears to be spiral, but I think this is probably 
 an optical illusion. 
 
 In the bent tube, as shown in Fig. 43, if the curled portion 
 
EXPERIMENTS WITH VACUUM TUBES 39 
 
 is placed in the magnetic field, the spiral discharge is very 
 well shown. 
 
 The experiment was varied by employing two tubes as 
 shown in Figs. 44, 45. The end electrodes were joined 
 
 FIG. 44. 
 
 together in each case. The middle electrode in Fig. 44 
 is centrally situated ; in Fig. 45 it is nearer to one end. 
 If no separation of the current took place, it is clear that 
 in tube Fig. 45 most of the current would take the shortest 
 course, but when the electrode in the body of the tube in 
 each case is placed in a strong magnetic field, the current 
 
 FIG. 45. 
 
 travels to the two ends in single lines of bands. The dis- 
 charge is very pretty, for the reason that there is a glow of 
 light very much like a fountain, in addition to the lines of 
 bands. 
 
 Carrying this method of experimenting a stage further, 
 I employed a tube U-shaped, as shown in Fig. 46. The 
 two electrodes at the free ends of the tube are joined 
 together, and a third bulb containing the other electrode 
 is placed in a strong magnetic field. If the current is 
 suitably adjusted, only one line of bands (each due to a 
 current in one direction) appears in each arm of the tube. 
 
 Going a step further, another similar tube is taken, and 
 
40 EXPERIMENTS WITH VACUUM TUBES 
 
 the free ends joined by two wires, as shown in Fig. 47. 
 On passing the current it becomes visible in the second 
 tube, and this current can be analyzed in another strong 
 magnetic field. In each case only a single line of bands 
 will appear in each arm. It is quite immaterial how long 
 
 FIG. 46. 
 
 the wires joining the tubes may be. This experiment is 
 not easy to perform on account of the care required in 
 adjusting the current, but when this condition has been 
 complied with, the phenomenon is unmistakable. 
 
 This experiment may be a forerunner of a new system 
 of electric lighting by the employment of alternators with 
 
 3 C 
 
 c 
 
 FIG. 47. 
 
 a service system of a direct pulsating current, without the 
 employment of moving machinery for re-dressing. At 
 present I am only able to deal with very small currents. 
 
EXPEEIMENTS WITH VACUUM TUBES. 41 
 
 In order to be sure of the ground, these experiments 
 were conducted not only with an induction coil arranged 
 to show the two currents in a tube, but also with an 
 alternate current from an alternator, with similar results 
 in each case. 
 
 As it is now generally recognized that a current 
 advances through a liquid and through a gas in a similar 
 manner, I therefore thought it desirable to repeat the 
 experiments which were conducted with tubes of the 
 shape of Fig. 46 with a water tube, and examine the 
 result. Fig. 48 illustrates the tube used, the portion at 
 
 FIG. 48. 
 
 the bend of the U being open. This end of the tube was 
 placed between the magnet poles, and the tube was joined 
 up by means of wires to an analyzing tube, as shown in 
 Fig. 47, which was placed between the poles of another 
 powerful magnet. The difference observable in the nature 
 of the discharge between the excited and unexcited con- 
 dition of the magnet which influenced the water tube, 
 was barely noticeable. There was a slight change in the 
 appearance of the discharge in the analyzing tube, but no 
 more can be said. This experiment has no direct bearing 
 upon the manner in which the current is propagated, but 
 I regard it conclusive as showing there is not that same 
 freedom for a current to take up a new position when 
 
42 
 
 EXPERIMENTS WITH VACUUM TUBES 
 
 traversing a liquid, as is the case with a gas, that is, fluid 
 is not a true movable conductor. 
 
 If a current is made to traverse an exhausted globe of 
 glass, say 50 mm. in diameter, with an electrode at each 
 pole, such as shown in Fig. 49, the discharge has a similar 
 
 appearance to that in the electric egg. When this globe 
 is placed in the magnetic field it presents a beautiful 
 appearance. A purple disc is seen near each electrode at 
 right angles to the polar axis, the planes coinciding with 
 the lines of force at these places. The current travels 
 between the two electrodes as two separated lines of light 
 joined at the electrodes and considerably parted midway 
 between them. 
 
 The current can be so adjusted that these two lines of light 
 
EXPERIMENTS WITH VACUUM TUBES 43 
 
 are so much separated that they travel nearly round the 
 inside of the glass of the sphere. The position of the lines 
 of force also becomes visible at all parts within this tube. 
 In fact, this spherical tube is, in a way, a summation of the 
 various phenomena shown in a number of the other tubes. 
 
 We are all familiar with the fact that the electric spark 
 in air is distinctly zig-zag in form. This is not clearly 
 visible, until the spark is made a sufficient length, and is 
 so well shown by an induction coil. 
 
 Most workers have also tried that well-known experi- 
 ment of discharging the spark from the induction coil 
 through a very fine wetted filament, or a very fine festoon 
 of wire, when it is observed how the filament or wire 
 crumples into the zig-zag form, during the time the dis- 
 charge is passing. It would therefore appear as if the 
 zig-zag form of discharge is not dependent on the fact 
 that the current passes through the air, but rather that 
 there is a tendency for the discharge to be zig-zag when 
 disruptive, and can be shown when the conductor is free 
 to move. 
 
 If the experiment of showing the magnetic lines of 
 force in a vacuum tube is performed with the E.M.F. 
 slightly raised, there will appear throughout the tube a 
 number of line discharges zig-zag in form very similar in 
 appearance to the ordinary spark in air. 
 
 In this instance there can be little doubt that the 
 magnetic field, as well as self-induction, plays a part in 
 the phenomenon. I would venture to suggest that in all 
 cases the zig-zag advance of the current may be due to its 
 own magnetic influence, combined with self-induction, 
 since it is unlikely that the magnetism of the earth is 
 sufficiently great to produce any visible result in this 
 direction. 
 
 To examine this phenomenon further : 
 
 D 
 
44 EXPERIMENTS WITH VACUUM TUBES 
 
 The air in a vacuum tube is only rarefied in comparison 
 with air at normal atmospheric pressure; thus, air in the 
 latter condition may also be regarded as rarefied when 
 compared with air under greater pressures. 
 
 Consequently it is not improbable that striation may be 
 produced in gases at all pressures, provided a sufficiently 
 high E.M.F. is employed, and the conditions made favour- 
 able for the formation of bands. The zig-zag discharge 
 
 FIG. 50. 
 
 may therefore be the tendency for the current to form 
 striation, and a compromise is brought about, resulting in 
 the uneven line of light ; the self-induction of the current 
 assisted by its own magnetic field being the primary 
 cause. 
 
 If an electric spark between two points of, say, 10 or 12 
 mm. apart (see Fig. 50) is placed between the poles N. 
 
 
 
 X 
 
 / 
 
 1 
 
 ' 
 
 
 ^>V 
 
 
 \ 
 \ 
 
 > 
 
 / 
 
 FIG. 51. 
 
 and S. in a powerful field, one of the poles being hollow so 
 that, the effect can be easily observed, then the following 
 phenomena result : If the coil is arranged to give the two 
 currents in almost equal proportions, the appearance will 
 be that of Fig. 51. The spark will appear bluish- white, 
 
EXPERIMENTS WITH VACUUM TUBES 45 
 
 and two reddish sparks will appear, one on either side, 
 giving the appearance of a letter D, on one side, and a 
 reversed D on the other ; the semi-circular portions do not 
 
 X 
 
 \ 
 
 FIG. 52. 
 
 form a circle, being somewhat displaced. If the current is 
 reversed, the same phenomenon is produced, but the dis- 
 placement is inverted. See Fig. 52. 
 
 FIG. 53. 
 
 If the coil is arranged to give practically only the break 
 current, the appearance will be as in Fig. 53, and when 
 reversed as in Fig. 54. 
 
 r 
 
 v / 
 
 FIG. 54. 
 
 It is evident that since the supplementary arcs follow 
 the variations mentioned, they must be electric discharges 
 or currents and not chance flames. 
 
46 EXPERIMENTS WITH VACUUM TUBES 
 
 In connection with this subject, there are a large number 
 of other very interesting experiments, but they will be 
 dealt with at some future time. 
 
 I started with the belief that vacuum tubes would prove 
 a useful aid for investigating electric phenomena, and I 
 also felt assured that for many purposes they could be 
 turned to practical account. I already employ them to a 
 considerable extent for the rapid examination of magnetic 
 fields for adjusting galvanometers in which the magnetic 
 field forms a portion of the apparatus, and for a variety of 
 other purposes. 
 
 I hope that those who have had but little faith in 
 vacuum tubes on account of their distant acquaintance- 
 ship with them, may be tempted to modify their views, 
 after what I have written, and give them a trial. 
 
 I have still in hand several series of experiments which, 
 I think, will probably be found of interest when completed, 
 if the word completed can be applied to any subject. I 
 simply mean a set of experiments which prove some 
 definite point, but nothing more. 
 
 It would be interesting to examine the results described, 
 and many others, with a steady current such as would be 
 produced from a large accumulator. The late Mr. Warren 
 De La Rue carried out some useful work in this direction, 
 but though the E.M.F. which he employed was high, the 
 currents were very small. I am at the present time con- 
 structing an accumulator of 1000 cells, capable of giving 
 two amperes, without injury to the cells. This E.M.F. is 
 more than is required for almost every experiment described. 
 I therefore hope to obtain some new and interesting results. 
 I have tried here to express myself freely without 
 any regard to the orthodox opinions which prevail at 
 present. It is most likely, therefore, that I have "run 
 off the rails " at many points. It is sometimes con- 
 
EXPERIMENTS WITH VACUUM TUBES 47 
 
 venient to do this, in order to make explanations clearer, 
 and avoid in this manner long descriptions, studded with 
 still longer parentheses. The advantage also exists that it 
 helps the worker to look at things from different points of 
 view, which may be described as the Romance of Science. 
 
APPENDIX 
 
 FOR the convenience of experimenters who are interested 
 in the subject of phenomena in vacuum tubes, I add this 
 Appendix, to give some information on the apparatus I 
 employ. 
 
 The high speed contact apparatus consists of a star 
 wheel with twenty spokes, the end of each spoke being tipped 
 with a platinum stud. 
 
 A platinum-tipped spring is touched as these spokes 
 pass, and the star wheel can be rotated by means of 
 an electro-motor at the rate of 2500 times per minute or 
 more without failure of contact. I do not in practice employ 
 more than 1500 breaks per minute. 
 
 The apparatus for slow makes and breaks consists of a 
 rod with platinum-tipped ends, working in guides, this rod 
 having a sliding action given to it by means of a revolving 
 eccentric which is moved by an electro-motor. 
 
 Two posts, platinum tipped, are placed on the stand 
 which carries the apparatus, one at each end, and the ends 
 of the rod strike these posts alternately, and thus the 
 makes and breaks are produced. 
 
 To make sure of contact, and to avoid vibration, the 
 platinum blocks at the ends of the rod are mounted on 
 coach-springs, which in their turn are supported by spiral 
 springs. The speed of the makes and breaks with this 
 
EXPERIMENTS WITH VACUUM TUBES 49 
 
 apparatus is at the rate of 800 per minute, which can be 
 lowered to 400 at pleasure. 
 
 Condensers are used in connection with both these in- 
 struments, so that they are very free from sparking. 
 
 The current passes through both these instruments and 
 thence to the primary of a suitable, and carefully made, 
 Apps Induction coil. A switch permits either the rapid or 
 slow " contact breaker " to come into play at pleasure. The 
 current used for the primary circuit is 100 volt direct 
 current. 
 
 A variable resistance is placed in connection with the 
 coil primary for adjusting the current. 
 
 A variable resistance is also placed in the secondary 
 circuit. 
 
 The length of spark which I employ in most of the 
 experiments is only 0*17 mm. The E.M.F. measured 
 upon a Kelvin electro-static-volt-meter is approximately 
 700 volts with the rapid breaker, but by altering the 
 pressure of the spring on this contact breaker, thereby 
 making the break more or less sudden, this voltage could 
 be brought down to 600 or raised to 800 at pleasure. 
 
 If the speed is increased, the E.M.F. rises in the second- 
 ary circuit to the limit of 1200 volts in these experiments. 
 With the slow breaker the E.M.F. varied from 300 to 400 
 volts. 
 
 In describing the behaviour of vacuum tubes in the 
 magnetic field, the words "equatoriaJly " and "axially" 
 occur. To prevent misunderstanding, by " equatorially " 
 is meant a direction at right angles with the axis of the 
 poles ; and " axially," in the direction of the axis of the 
 poles. If the lines of force be considered as an imaginary 
 approximate sphere, then the meanings of the words 
 become obvious. 
 
 UNI 
 
RICHARD CLAY <fc SONS, LIMITED, 
 LONDON <fe BUNGAV. 
 
( 4 ) 
 
 Inventions, How to Patent, 2s. 6d. 
 Iron Analysis, Arnold, IDS. 6d. 
 
 Analysis, Blair, iSs. net. 
 
 and Steel Structures, Twelve- 
 trees, 6s. net. 
 Ironfounding, Homer, 4-r. 
 
 Jack's Cooking, 2s. 
 
 Laundry Work, 2s. 
 
 Jacobi's Printer's Handbook, $s. 
 
 Vocabulary, 35. 6d. 
 
 Jukes-Browne's Geology, 2s. 6d. 
 
 Kapp's Electric Transmission of 
 
 Energy, los. 6d. 
 
 Transformers, 6s. 
 
 Kirkman's Science of Railways, 
 
 12 vols. 61. 6s. net. 
 
 Lamps, Electric, Maycock, 6*. 
 
 Land Measuring Tables, Cullyer, 
 2s. 6d. 
 
 Surveying, Middleton, $s. 
 
 Surveying and Levelling, 
 Walmisley, 6s. net. 
 
 Landolt's Optical Activity and Chemi- 
 cal Composition, qs. 6d. 
 
 Lathes, Homer, 21 s. 
 
 Laundry Work, Jack, 2s. 
 
 Leather Work, Leland, $s. 
 
 Leland's Wood-carving, $s. 
 
 Metal Work, $s. 
 Leather Work, $s. 
 
 Drawing and Designing, 2s. 
 
 Practical Education, 6s. 
 
 Lens Work, Orford, $s. 
 Lenses, Photographic, Taylor, 3.$-. 6d. 
 Light, Sir H. T. Wood, 2s. 6d. 
 Lightning Conductors, Lodge, i$s. 
 Locomotive Appliances, Kirkman, 
 IQS. 6d. net. 
 
 Engineer, McDonnell, is. 
 
 Locomotives, Cooke, 7*. 6d. 
 
 Development, Cooke, 2s. 6d. 
 Lodge's Lightning Conductors, i$s. 
 Loppe and Bouquet's Alternate 
 
 Currents, ios. 6d. 
 Lubrication, Wheeler. 
 Lunge and Hurter's Alkali Makers' 
 
 Handbook, los. 6d. 
 
 McDonnell, How to become a Loco. 
 
 Engineer, is. 
 Maginnis' Atlantic Ferry, "js. 6d. 
 
 Bricklaying, "js. 6d. 
 House Framing, $s. 
 
 Maginnis' Roof Framing, 5^. \2s. 6<t. 
 
 Magnetism and Electricity, Ashworth, 
 Bottone, 2s. 6d. 
 
 Houston, 3^. 6d. 
 
 Maycock, 2s. 6d. 
 
 Manual Instruction and Training. 
 
 Manures, Griffiths, 7s. 6d. [2s. 6d. 
 
 Marine Engineers, advice to, Roberts, 
 
 Drawing and Designing, 6s. 
 
 Electric Lighting, Walker, 5*. 
 
 Questions and Answers, Wan- 
 nan and Sothern, $s. 6d. net. 
 Verbal Notes, Sothern, 45. 6d. 
 
 Marshall's Cakes, is. [net. 
 
 Cookery, is. 
 
 Martin's Structures, 4_r. 
 
 Massee's The Plant World, 2s. 6d. 
 
 Mathematical Tables, Hutton, 12s. 
 
 Mathematics, Elementary, Hatton, 
 2s. 6d. 
 
 May's Ballooning, 2s. 6d. [2s. 6d. 
 
 Maycock's Electricity and Magnetism, 
 
 Electric Lighting, 6s. [2*. 6d. 
 
 Alternating Current Circuit, 
 
 Electric Wiring, Fittings, 
 
 Switches, and Lamps, 6s. 
 
 Details Forms, 2s. 6d. net. 
 
 - Electric Wiring Tables, 3*. 6d. 
 Measurements, Electrical, Crapper, 
 
 Houston, 5*. [2J-. 6d. 
 
 Mechanical Tables, Foden, is. 6d. 
 
 Refrigeration, Williams. [55-. 
 
 Mechanical Engineer's Pocket -Book, 
 Medical Terms, Hoblyn, IDJ. 6d. 
 Merrill's Electric Lighting Speci- 
 fications, "js. 6d. 
 
 Tables for Street Railway 
 
 Engineers, 5*. 
 
 Metal Turning, Horner, 4^. 
 
 Work, Leland, 5*. 
 
 Miller. 
 
 Metric System, Wagstafi", is. 6d. 
 
 Measures, Bom, 35. 
 
 Middleton's Surveying, 5*. 
 
 Mill Work, Sutcliffe, 21 s. [151. 
 
 Miller's American Telephone Practice, 
 
 Metal Work. 
 
 Millinery, Ortner, 2s. 6d. 
 Mine, Coal, Boyd, "js. 6d. 
 Mineralogy, Hatch, 2s. 6d. 
 Model Engine Making, los. 6d. 
 Modern Safety Bicycle, Garratt, 31. 
 Moffett's Alternate Currents, ios. 6d. 
 Motor Cars, Farman, 5^. 
 Motors, Electro, Bottone, 31. 
 Houston, 5^. 
 
THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 Munkip; AN INITIAL FINE OF 25 CENTS 
 
 bings, WILL BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 Nadieinc WILL INCREASE TO SO CENTS ON THE FOURTH 
 
 Neuman DAY AND TO $1.00 ON THE SEVENTH DAY 
 
 OVERDUE. 
 
 Optical 
 
 \s. 6d. 
 rs, 6s. 
 'nirin- 
 
 positu 
 Optical MAR *RA 
 
 optics, nr( Jty i 
 
 \Af\ 
 
 n and 
 
 .. O f 
 Orford's MAR 1 4 19 
 
 l *ru 
 
 Oj 
 
 Organic 
 
 
 ewi 
 Ortner's AAAM 
 Ow r en'sik|/\\f "I O nOiJl*'* 
 
 [tion : 
 astalla- 
 
 Parham 
 
 
 )s and 
 French 
 
 ,2s.6<i. 
 :chism, 
 . [ios. 
 
 L net. 
 cetches, 
 
 irs for 
 et. 
 
 tionary, 
 
 12s. e 
 Parkhu 
 
 
 Patentii 
 Pattern 
 
 Prrinrli 
 
 
 Evol 
 Pptrole 
 
 
 Photog 
 Physics 
 
 
 Pickwc 
 
 each. 
 
 
 Pipes i 
 Plant ^ 
 
 
 Plante' 
 Ponce 
 
 
 logic 
 net. 
 
 
 Poole s 
 Practic 
 
 Prntt 
 
 
 Preece 
 Price's 
 
 
 Ten 
 Printe: 
 
 d. 
 . 6d. 
 
 Quant 
 
 .rk, Sut- 
 
 Radio 
 
 
 Twelve- 
 nt Phe- 
 
 125. 6d. 
 
 Railw 
 Kir 
 
 
 =' 
 
 LD 21-100m-7,'39(402s) 
 
Stevenson's Trees of Commerce, 
 
 3*. 6d. 
 
 Still's Alternating Currents, 55. 
 Storage Battery, Tread well, Is. 6d. net. 
 Street Ps.ailroad, Pratt and Alden, los. 
 Stresses and Strains, Farrow, 5^. net. 
 Structural Iron and Steel, Twelve- 
 Structures, Martin, 4^. [trees, 6s. net. 
 Stubbs' Telephony, i$s. 
 Surveying, Middleton, $s, 
 
 Walmisley, TS. 6d. [Work, 2is. 
 
 Sutcliffe's Steam Power and Mill 
 
 Sanitary Fittings, $s. 
 
 Switches, Electric, Maycock, 6s. 
 
 Tables, Electric Wiring, Maycock, 
 
 YB 09642 
 
 
 UNIVERSITY OF CALIFORNIA LIBRARY 
 
 Still, $s. [net. 
 
 Treadwell's Storage Battery, 7s. 6d. 
 Trigonometry, Adams, 2s. 6d. net. 
 Tubes and Pipes, Bjorling, 3.5-. 6d. 
 
 Vacuum, Salomons. 
 
 Turbines and Hydraulic Motors, 
 
 Bodmer, 155. 
 
 Turning Lathes, Homer, 2is. 
 Turning, Metal, Horner, 4$. 
 Twelvetrees' Structural Iron and 
 
 Steel, 6s. net. 
 Typography Questions, 6d. 
 
 Volumetric Analysis, Coppock, 2s. 
 
 Wagstaffs Metric System, is. 6d. 
 
 Walker's Colliery Lighting, 2s. 6d. 
 
 Electricity in Homes, 6s. 
 
 Electric Lighting for Marine 
 
 Engineers, $s. 
 
 Wallis's Dynamo, los. 6d. 
 
 Wallis-Tayler's Bearings and Lubri- 
 cations, 3-r. 6d. 
 
 Sugar Machinery, $s. 
 
 Walmisley's Held Work and Instru- 
 
 ^evelling, 
 
 itions and 
 Engineers, 
 
 lonary, 4 
 French, 
 
 T, 3J. 6d. 
 ction, etc. 
 Engineer's 
 
 Pocket 
 
 J, iS/ 
 
 Deration. 
 
 ic, 3J. 
 
 6s. 
 
 : 6d. net. 
 
 , 5. 6d. 
 
 _ 
 
 Wood Carving, Lelariff, $s, 
 Wood's Discount Tables, is. 
 
 - Light, 2s. 6d. 
 Woodwork, Barter, 6s. 
 
 Drawing, 35. 6d. 
 
 - Joints, Adams, is. 
 Woven Design, Beaumont, 2is. 
 Wyatt's Differential and Integral 
 
 Calculus, 3-r. 6d. 
 
 X Rays, Bottone, 31. 
 
 Yeaman's Electricity Supply, los. 6d. 
 
 FULL CATALOGUE of Books for Civil, Electrical and 
 Mechanical Engineers, Post Free, 
 
 WHITTAKER & CO. 
 2 WHITE HART STREET, PATERNOSTER SQUARE, LONDON, E.G.