■w,'f 
 
 
BOUGHT WITH THE INCOME 
 FROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 lienrg W, Sage 
 
 1891 
 
 ^■•z^SS't.g.l TiXi.i 
 
 1357 
 
Cornell University Library 
 QC 7.S39 
 
 The progress of physics during 33 years 
 
 3 1924 006 366 763 
 
Cornell University 
 Library 
 
 The original of tliis book is in 
 tine Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924006366763 
 
THE PROGRESS OF PHYSICS 
 
CAMBRIDGE UNIVERSITY PRESS 
 
 EonBDn: FETTER LANE, E.G. 
 C. F. CLAY, Manager 
 
 (Eljintursb : 'o°, PRINCES STREET 
 
 Bctlm: A. ASHER AND CO. 
 
 IcipjiB: F. A. BROCKHAUS 
 
 J3tro gock: G. P. PUTNAM'S SONS 
 
 JSombaB aria Calcutta: MACMILLAN AND CO., Ltd. 
 
 A// rights r^servec^ 
 
D ^l 
 
 u^ 
 
THE PROGRESS OF PHYSICS 
 
 DURING 33 YEARS (1875— 1908) 
 
 FOUR LECTURES DELIVERED TO THE UNIVERSITY 
 OF CALCUTTA DURING MARCH 1908 
 
 by 
 
 ARTHUR SCHUSTER, F.R.S. 
 
 Ph.D. (Heidelberg), Sc.D. (Cantab.), D.Sc. (Manchester and Calcutta), 
 D.te.Sc. (Geneva) 
 
 Cambridge : 
 at the University Press 
 
 191 1 
 
 -r 
 
INTRODUCTION 
 
 IN writing out these lectures which were dehvered 
 nearly three years ago, I have made a few additions, 
 in order to bring the subjects to their position at the 
 present time. Personal opinions have been influenced 
 in some cases by subsequent researches or by maturer 
 reflexion. All such expressions of opinion must there- 
 fore be considered as indicating my present attitude 
 rather than that taken up at the time of the lectures. 
 Their scope and aim has been sufficiently indicated in 
 the opening passages. 
 
 ARTHUR SCHUSTER. 
 
 December 1910. 
 
TABLE OF CONTENTS 
 
 LECTURE I 
 
 Scope of Lectures. State of Physics in 1875. Science of Energy. 
 Theory of Gases. Elastic Sohd Theory of Light. Maxwell's Theory of 
 Electricity. Training of Students. iVlaxweH's View. Accurate Measure- 
 ment and Discovery of Argon. German Methods. Kirchhoffs Laboratory. 
 Wilhelm Weber's Laboratory. The Two Laboratories of Berlin. Labora- 
 tory Instruction at Manchester. Position of Physics in Mathematical 
 Tripos at Cambridge. Todhunter's Views. The Cavendish Laboratory. 
 Spectrum Analysis. The Radiometer. Theory of Vortex Atom. 
 
 PP- 1—34 
 
 LECTURE II 
 
 Action at a Distance. Elastic Solid Theory of Light. Maxwell's 
 Theory of Electrical Action. Electro-magnetic Theory. Verification of 
 Electro-magnetic Theory by Hertz. Electro-magnetic Waves. Wireless 
 Telegraphy. First Suggestion of Molecular Structure of Electricity. 
 Early Experiments in the Electric Discharge through Gases. Kathode 
 Rays. Work of Goldstein and Crookes. Hittorf 's Investigations. Own 
 work on the Discharge through Gases. Ionization of Gases. Magnetic 
 Deflexion of Kathode Rays. J. J. Thomson's Experiments. Measure- 
 ment of Atomic Charge. ... ... ... pp. 35 — 72 
 
 LECIuRE III 
 
 Roentgen's Discovery. Theories of Roentgen Rays. Ionizing Power 
 of Roentgen Rays. Conduction of Electricity through Ionized Gases. 
 Discovery of Radio-Activity. Discovery of Radium. Magnetic Deflexion 
 of Rays emitted by radio-active Bodies. Discovery of Emanations. 
 Theory of radio-active Change. Decay of the Atom. Connexion between 
 Helium and the a Ray. Hehum produced by Radium. Strutt's Researches 
 on Helium accumulated in Rocks. Electric Inertia. Constitution of Atom. 
 J. J. Thomson's Theory of Roentgen Radiation. The Michelson-Morley 
 Experiment. Principle of Relativity. The Zeeman Effect. Other con- 
 sequences of Electron Theory. Contrast between old and modern School 
 of Physics. ... ... ... ... ... pp. 73—117 
 
Table of Contents 
 
 LECTURE IV 
 
 Observational Sciences. Judgment affected by Scale. Terrestrial 
 Magnetism. E.xistence of Potential. Separation of internal and external 
 Causes. Diurnal Variation. Magnetic Storms. Their causes. Solar 
 Influence. Theories of Secular Variation. Atmospheric Electricity. 
 Negative Charge of Earth. Ionization of Air. Origin of Atmospheric 
 Electricity. Electric Charge of Rain. Ebert's Theory. Cause of 
 Thunderstorms. The Age of the Earth. Rigidity of Earth. Displace- 
 ment of Axis. Gravitation. Identity of Molecules of the same kind. 
 
 pp. 1 18—159 
 
 Index • pp. 161—164 
 
 The author is indebted to the proprietors of Nature for permission 
 to insert, as frontispiece, a copy of the portrait of J. Clerk Maxwell, 
 which appeared in that journal in the series of "Science Worthies." 
 
LECTURE I 
 
 It is more than thirty years ago, that returning from 
 a scientific expedition I hurriedly passed through India, 
 and fascinated by its sunshine, the beauties of its moun- 
 tain scenery and the mysteries of its national life, 
 carried home with me the intense desire of a more 
 leisurely visit. I then stood at the beginning of my 
 scientific career and when now, nearing its end, my 
 wish is fulfilled by being called upon to fill the honour- 
 able office of " Reader" to the University of Calcutta, 
 I thought it might be interesting to you — as it is to 
 me — if I discharge my duties by reviewing the progress 
 of scientific thought in the interval. 
 
 It is my aim to trace the changes in our point of 
 view, rather than to give an historical account of the 
 sequence of the discoveries which make this period 
 memorable. In taking this course I am aware of the 
 risk incurred in an attempt, which to a great extent 
 must be guided by subjective impressions ; for though 
 a man may see clearly the changes that have taken 
 place in his own mind, he may be wrong in estimating 
 the rise of the average level of scientific thought by 
 the standard of his individual progress. According as 
 
 S. I 
 
2 Lecture I 
 
 he has been taught by advanced or antiquated teachers, 
 according as he ends his work ahead of, or behind 
 his time, will he give a very different version of the 
 progress which has been achieved. But rightly or 
 wrongly I have chosen my course : I prefer to be 
 frankly subjective, and warn you beforehand that my 
 account will be fragmentary, and to a great extent 
 reminiscent of those aspects which have come under 
 my own personal view. 
 
 Let us briefly survey the general position of physical 
 science at the time our story begins. The work of 
 the previous twenty years was dominated by the 
 gradual recognition of the Science of Energy founded 
 experimentally on the work of Joule, and theoretically 
 on that of Kelvin, Clausius and Helmholtz. Destined 
 from the beginning to serve as a bond between all 
 branches of physical knowledge, it found its first 
 triumph in establishing the connexion between the 
 science of heat and the theorems of dynamics. The first 
 law of thermodynamics, which teaches us that heat is 
 generated or destroyed in exact proportion to the 
 amount of mechanical work lost or gained, was not 
 only recognised as unassailable, but had truly saturated 
 the soul of the scientific body. Such was not entirely 
 the case (is it now .^) with the second law, which regu- 
 lates the conditions, under which heat can be chang-ed 
 
 o 
 
 into mechanical work. It is true that every one whose 
 opinion counted, accepted the law, and no student was 
 allowed to remain ignorant of Carnot's cycle, and of the 
 mathematical equations which can be derived from it ; 
 but the law was not a living and fertilising part of 
 physical science ; except in its special applications to 
 
Historical Review 3 
 
 heat engines. In its more refined consequences dealing 
 with entropy and available energy it had not yet be- 
 come common property, and in its application to the 
 decay of universal energy it was not sufficiently under- 
 stood and appreciated. 
 
 The tendency to hide ignorance under the cover 
 of a mathematical formula had already appeared but 
 was not openly advocated ; hence students were still 
 taught to form definite ideas of the processes of nature. 
 In the kinetic theory of gases ample material was found 
 for a mental picture of the state of matter in its simplest 
 form. The calculation of the average molecular velo- 
 cities, the application of the theory to the jorocesses 
 of diffusion, conduction and viscosity had been ac- 
 complished and was taught at some, at any rate, of the 
 Universities. I became acquainted with the theory of 
 gases in lectures given by L. Soret at Geneva in 
 1869, and well remember how the possibility of de- 
 termining the size of molecules and their number in a 
 given volume seemed to open out before me, but though 
 I attacked the problem I could make nothing of it, 
 and it was not till some time later that I found it 
 already solved by Johnstone Stoney and Loschmidt. 
 Those who could follow the more intricate mathematical 
 developments of the subject found in Maxwell's in- 
 vestigation on the law of distribution of molecular 
 velocities a new revelation, which introduced for the 
 first time the theory of probability into problems relating 
 to the physical state of bodies. In the succession of 
 new ideas that have influenced the progress of Physics 
 this great step must always hold a pre-eminent position. 
 
 As regards light, the elastic solid theory held the 
 
4 Lecture I 
 
 field, though serious difficulties, passed over too lightly 
 in the period of its triumph, already foreshadowed its 
 ultimate defeat. 
 
 Stokes by his brilliant investigations had placed the 
 laws of transmission of vibrations through elastic solids 
 on a firm basis. He had shewn how a complete theon,- 
 must take account of two waves, the condensational 
 wave and the distortional one. The first of these is 
 important in the theory of sound, but no optical phe- 
 nomenon indicates its existence and hence arises a 
 certain difficulcv. When a sing^le homogeneous medium 
 is considered, this difficulty is most easily overcome by 
 assuming the condensational wa^'e to be transmitted 
 with infinite velocity, but Lord Kelvin shewed that 
 consistent results are obtained equally well, by intro- 
 ducing the condition that these waves are propagated 
 with a velocity which is infiniteh' small. This howe^'er 
 was at a date considerably later than the one to which 
 I am at present referring. Condensational waves were 
 then generally ignored, but another stumbling block 
 worried our minds. The laws of transmission of light 
 through ditterent bodies could, up to a certain point, 
 be equally well explained by assuming the elasticity of 
 the arther to be the same ever)-where, and its density 
 to be ditterent in ditterent bodies, or bv specifying that 
 the densitv is constant, while the elasticitv changes 
 An increased velocity inside a body could accordingly 
 either be due to an increased elasticity or to a diminished 
 density. The relationship between the direction of 
 vibration in a beam of light and what is called its plane 
 of polarisation depends on which of the two alternatives 
 we adopt. If the density be variable, the direction of 
 
Elastic Solid Theory of Light 5 
 
 vibration must be at right angles to the plane of polar- 
 isation, while parallelism between the direction of 
 vibration and the plane of polarisation would signify 
 that the elasticity differs. The more refined conse- 
 quences of the theories gave rise to the hope, that 
 experiments might be devised to decide between the 
 two alternatives, but unfortunately the verdict of ex- 
 periment was not uniformly on the same side. When 
 Fresnel's formula for the intensity of a ray reflected 
 from a transparent body, which had been verified 
 experimentally by Brewster, was closely criticised by 
 Lord Rayleigh, its theoretical foundations were only 
 found to be sound, on the supposition of variable 
 density. This meant that the vibration was at right 
 angles to the plane of polarisation, a result which was 
 also supported by the polarisation observed in a wave 
 scattered by small particles. 
 
 But the law of double refraction in crystalline media 
 led to the opposite conclusion. Here again Fresnel 
 was the pioneer, who gave us the equation of the wave 
 surface by specifying that elasticity in a crystal depends 
 on the direction of displacement. Physicists were now 
 placed in the dilemma, that in different parts of optics 
 two mutually destructive theories on the properties of 
 the aither were supported by experiment. For though 
 Rayleigh shewed that we may obtain a surface not 
 differing very much from Fresnel's, by assuming a 
 variable inertia according to the direction of vibration, 
 the difference was sufficient to admit of the question 
 being decided by experiment, which, performed by 
 Glazebrook gave its verdict emphatically in favour of 
 Fresnel. These difficulties and discrepancies, though 
 
6 Lecture I 
 
 not considered fatal, prepared men of science for the 
 great revolution, which was soon to come and sweep 
 away the whole elastic solid theory. 
 
 The state of electrical science in 1S70 must present 
 itself somewhat differently to the scrutiny of a historian 
 who derives his information from the records of 
 published papers, and to the recollection of a student, 
 who received his impressions from the teachers at the 
 time. Maxwell's great paper "A Dynamical Theory 
 of the Electromagnetic Field" appeared in 1S64, but 
 I doubt whether the younger generation of physicists 
 had their attention drawn to, or seriously arrested by 
 it, before the publication in 1S72 ot the two volumes 
 Electricity and Magnetism. I believe that the first 
 systematic course of lectures based on Maxwell's theory 
 was given by myself at the Owens College during 
 the session iS 75-76. (Sir Joseph Thomson was one 
 of the three students attending the course.) At the 
 time jMaxwell's volume appeared, the teaching of 
 electricity centred round the calculation of coefficients 
 of induction and iutile discussions on the laws of action 
 of so-called current elements. A wider and more 
 philosophic aspect was now brought before us, and 
 though [Maxwell's treatise was essentially mathematical, 
 it put an end to the university tradition of looking upon 
 electricity as part of applied mathematics to the neglect 
 of its physical aspect. Hence INIaxwell's work changed 
 the whole point of view of the study of electrical science. 
 
 While the above indicates in brief outline the 
 general condition of the main branches of Physics at 
 the period of which we are speaking, we must now 
 turn our attention to the important question as to how 
 
Training of Students 7 
 
 students were prepared by teaching and example to 
 take an active part in advancing their subject. 
 
 I think I interpret correctly the recollection of those 
 who passed through their scientific education at the time, 
 when I say that the general impression they received 
 was that, apart from theoretical work, a reputation 
 could only be secured by improved methods of 
 measurement which would extend the numerical ac- 
 curacy of the determination of physical constants. 
 In many cases the student was led to believe that 
 the main facts of nature were all known, that the 
 chances of any great discovery being made by ex- 
 periment were vanishingly small, and that therefore 
 the experimentalists work consisted in deciding between 
 rival theories, or in finding some small residual effect, 
 which might add a more or less important detail to 
 the theory. There were no doubt great differences of 
 opinion depending on the temperament of the teacher, 
 as to how far increased accuracy of measurement was 
 an object desirable in itself or only a means to 
 an end, and in this connexion a passage taken from 
 Clerk Maxwell's Introductory Lecture in Experimental 
 Physics deserves to be quoted' : 
 
 "This characteristic of modern experiments — that 
 they consist principally of measurements — is so promi- 
 nent, that the opinion seems to have got abroad, that 
 in a few years all the great physical constants will 
 have been approximately estimated, and that the only 
 occupation which will then be left to men of science 
 
 ^ Collected Works, n. p. 241. No date is attached to this 
 lecture in the published volume, but according to the information 
 given in Maxwell's Life it was delivered in October, 187 1. 
 
8 Lecture I 
 
 will be to carry on these measurements to another 
 place of decimals. 
 
 •• If this is really the state of things to which we 
 are approaching, our Laboratory may perhaps become 
 celebrated as a place of conscientious labour and con- 
 summate skill, but it will be out of place in the 
 University, and ought rather to be classed with the 
 other great workshops of our country, where equal 
 abilit}' is directed to more useful ends. 
 
 " But we have no right to think thus of the un- 
 searchable riches of creation, or of the untried fertility 
 of those fresh minds into which these riches will 
 continue to be poured. It may possibly be true that, 
 in some of those fields of discovery which lie open to 
 such rough observations as can be made without 
 artificial methods, the great explorers of former times 
 have appropriated most of what is valuable, and that 
 the gleanings which remain are sought after, rather 
 for their abstruseness, than for their intrinsic worth. 
 But the histon,- of science shews that even durina: the 
 phase of her progress in which she devotes herself to 
 improving the accuracy of the numerical measurement 
 of quantities with which she has long been familiar, 
 she is preparing the materials for the subjugation of 
 the new regions, which would have remained unknown 
 if she had been contented with the rough methods of 
 her early pioneers. I might bring forward instances 
 gathered from every branch of science, shewing how 
 the labour of careful measurement has been rewarded 
 by the discovery of new fields of research, and bv the 
 development of new scientific ideas. But the history 
 ot the science of terrestrial magnetism affords us a 
 
Maxwell on Laboratory Instruction 9 
 
 sufficient example of what may be done by experi- 
 ments in concert, such as we hope some day to 
 perform in our Laboratory." 
 
 Another interesting passage throws Hght on Max- 
 well's idea of the proper function of a laboratory. 
 
 " Our principal work, however, in the laboratory 
 must be to acquaint ourselves with all kinds of scientific 
 methods, to compare them, and to estimate their value. 
 It will, I think, be a result worthy of our University, 
 and more likely to be accomplished here than in any 
 private laboratory, if by the free and full discussion 
 of the relative value of different scientific procedures, 
 we succeed in forming a school of scientific criticism, 
 and in assisting the development of the doctrine of 
 method." 
 
 Maxwell's view was that of the most enlightened 
 and progressive philosopher then living, but his 
 optimism was not shared by others who at the time 
 enjoyed an equal or greater reputation. Gustav 
 Kirchhoff, for instance, was by no means a man who 
 despised experimental enquiry ; I have heard him 
 speak with appreciation of men who without much 
 theoretical knowledge tried to upset theory by experi- 
 ment ; but the sole advantage he expected to accrue 
 from their labours was the mending of the theory and 
 he did not anticipate new facts being discovered lead- 
 ing to a revision of fundamental conceptions. When 
 I told him of the discovery then made in England, 
 that light falling on the surface of a bar of selenium 
 altered its electrical conductivity, he remarked : " I 
 am surprised that so curious a phenomenon should 
 have remained undiscovered till now." This represents 
 
lo Lecture I 
 
 the attitude of mind not only of Kirchhoff but of 
 the great majority of physicists at the time. Looking 
 back now on this period, when Roentgen rays and 
 radio-activity were undreamt of, we may well learn to 
 be cautious in our own predictions of the future. 
 
 It will be noticed that though Maxwell evidently 
 looked forward to further discoveries, he expected them 
 to appear as residual effects rather than by a direct 
 experimental attack deliberately made to break new 
 ground. A typical example of a great discovery brought 
 about by the method indicated by Maxwell is furnished 
 by Rayleigh's measurements, which culminated in the 
 discovery of argon through a research undertaken to 
 determine with accuracy the density of gases, and 
 shewing unexpectedly a marked discrepancy between 
 two samples of apparently pure nitrogen. The gas 
 derived from air was found to be heavier than that 
 generated chemically, because as was ultimately shewn, 
 it contained small quantities of a denser and new gas. 
 But though this may be quoted as an example of the 
 orthodox method of discovery, and also fully recognising 
 that a teacher is almost bound to point to this as the 
 soundest method, it is nevertheless indisputable that 
 the greatest discoveries, both formerly and in recent 
 years have not originated in the hunt of residuals. 
 Altogether I am doubtful whether any great discovery 
 has ever been made by anyone who has only aimed at 
 recording a number of facts. I do not believe, in spite 
 of what is sometimes asserted, that Schwabe was led 
 to the discovery of the periodicity of sunspots simply 
 by sitting down to record without ulterior motive the 
 number of spots he could see ; and it is to me unthink- 
 
Acctirate Measurement and Research 1 1 
 
 able that the determination of the position of planets, 
 which allowed Kepler to discover his laws, was made by 
 men who, solely for their own pleasure or satisfaction, 
 recorded their places among the fixed stars. We may 
 be certain that Schwabe had some idea on the possi- 
 bility of connecting sunspots with other phenomena, and 
 we know that the old astronomers were not doing their 
 work simply for the sake of curiosity, or in the vague 
 hope that something unforeseen should come of it. 
 Similarly we may be certain that Lord Rayleigh's object 
 in determining the density of nitrogen, was not that of 
 obtaining a more accurate value of a physical quantity 
 without some definite view, why an increased accuracy 
 in the determination of that constant was desirable. 
 The most important result of a research no doubt is 
 very often not that which was originally aimed at, but 
 this does not prove the absence of an intention, which 
 went beyond that of mere increased accuracy or an 
 amplified record of facts. 
 
 My object is not, however, to shew what students 
 ought or ought not to have been taught, but what they 
 actually were taught thirty years ago. In this respect 
 Maxwell's declaration must be taken to represent the 
 most progressive view of the time, but unfortunately 
 only few students came under his direct influence. 
 
 I shall endeavour to describe the atmosphere 
 which surrounded a student of Physics in the five 
 institutions in which I worked. The method of 
 testing a student's fitness for graduation differed then, 
 as it does now, in different countries and naturally 
 formed a potent factor in determining the course of 
 instruction. In Germany the presentation of a thesis 
 
12 Lecture I 
 
 was almost universally a necessity, and consequently 
 the Professor had to find some easy subject on which 
 his pupils had to exercise their powers. There is no 
 doubt that this system brings a student quickly to the 
 limits of knowledge in one special department and he 
 learns that science even in simple problems is pro- 
 gressive and unlimited. This is a decided advantage, 
 counter-balanced however to some extent by the 
 tendency it induces of exaggerating the value of small 
 and unimportant matters. The problem of a thesis 
 must be definite and limited. Unless a teacher is 
 himself in the front rank of research, he cannot avoid 
 suggesting some small piece of detailed work which 
 may be useful as a training and even as a statistical 
 registration of facts, but is apt to distort the student's 
 perspective. The type of problem to which, as his 
 first piece of research, he is apt to attach importance, 
 has a tendency to affect his future aspect of nature, 
 for the pleasures of studying minute detail are fasci- 
 nating and grow quickly to a point where accuracy 
 becomes an end in itself instead of a means to an end. 
 In so far as the educational system is national do uni- 
 versity requirements impress a national character on 
 research work ; but we must be careful not to mistake 
 this for an inborn racial difference in the manner of 
 looking at nature. 
 
 At the time to which my remarks apply, a thesis 
 in an experimental subject was often undertaken 
 without any previous training in the laboratory such 
 as is everywhere now introduced. Comparing the 
 results of the old and modern system we are led to 
 the conclusion that at present we probably attach 
 
Kirchhoff' s Laboratory 13 
 
 too much importance to such prehminary training. 
 Experience has shewn that some practice in the 
 most common physical manipulation is desirable, but 
 there is nothing to prove that a moderate time spent 
 on a comparatively few typical experiments is not all 
 that is required for the purpose. The need for some 
 systematic training is acknowledged and had made itself 
 felt everywhere and almost simultaneously in different 
 countries. In Germany, the first edition of Kohlrausch's 
 text-book appeared in 1869, and the practice of 
 laboratory training in our present sense, spread slowly 
 but steadily through all Universities. Kirchhoff, at 
 Heidelberg, attached great importance to a carefully 
 prepared scheme of observation accurately carried out. 
 Once a week he gave a lecture explaining the principle 
 of some experiment and the methods of calculation. 
 As he generally had not more than twelve students, 
 one morning or one afternoon in the week could 
 be appropriated to each of them, to perform the 
 manipulation which had been described in the 
 previous lecture. The student, at the time set 
 apart for him, found the apparatus ready but only 
 partially adjusted, and without further assistance was 
 expected to complete the adjustment and obtain an 
 accurate result. Once during the time of work 
 Kirchhoff came round to see whether the student had 
 got into serious difficulty. The answers had to be 
 written out and were carefully entered up in a book. 
 At the beginning of a subsequent lecture, they were 
 all written up on a blackboard, and the professor 
 commented on their agreement or disagreement, and 
 discussed generally the accuracy which could be 
 
14 Lecture I 
 
 expected in any particular case. In addition to 
 this so-called "Seminar," which only occupied part 
 of the year, a few men were admitted to the laboratory 
 to carry on research work. In the two semesters 
 over which my stay extended, Lippmann was completing 
 his celebrated research on the capillary electrometer, and 
 Kammerlingh Onnes was preparing his doctor's dis- 
 sertation, on a modified form of Foucault's pendulum. 
 I had told Kirchhoff that my main object was to 
 learn from him as much as I could, and that I did 
 not mind whether any definite result came out of my 
 work or not. He consequently set me to test some 
 instrument he had designed to determine the elliptic 
 polarisation of light reflected from a metallic mirror ; 
 a typical piece of work of the orthodox stamp. The 
 instrument was not a particularly successful one, and 
 there was nothing to publish when it had been tested, 
 but I learned something both theoretically and ex- 
 perimentally on elliptic polarisation and I was quite 
 content. In addition to a very elementary course on 
 Experimental Physics, Kirchhofif gave some theoretical 
 lectures which were of the greatest interest. He 
 was a very precise man who weighed every word he 
 said, and was commonly reported never to have 
 missed a lecture'. Whenever a student ventured 
 on any remark respecting some question in Physics 
 
 ^ I can speak however from personal knowledge of an exception. 
 " Gentlemen," he told us at the end of a lecture, "I regret to inform 
 you that circumstances compel me to omit my lecture to-morrow." 
 The circumstances referred to were, that he was going to get married. 
 The omitted lecture was on a Saturday, Sunday was reserved for the 
 wedding tour and on Monday the lectures proceeded as usual. 
 
Heidelberg and GotHngen 15 
 
 he had to be prepared to stand a cross examination 
 on the precise meaning of what he said — but with all 
 that, Kirchhoff was one of the kindest of men, who 
 judged others leniently, though he expected the same 
 standard of accuracy in others which he had set for 
 himself. 
 
 During the summer of 1874 I spent a few months 
 at Gottingen, the recollection of which is valuable to me 
 chiefly on account of the scientific intercourse I had 
 with Wilhelm Weber. That name probably conveys 
 but little meaning to the students of the present day, 
 yet two generations ago he was a leader of science, 
 and one of the founders of our present system of 
 electric units and measurement. Weber's laboratory, 
 housed in a private residence, carried the distinction 
 that the first electric message transmitted along wires 
 was sent from it to the astronomical observatory, then 
 in charge of Gauss. 
 
 The contemplation of that period takes us back far 
 beyond the time which I am now trying to recall, and 
 I will not dwell upon it. Weber, when I knew him, 
 was 70 years old, and had preserved an active and 
 elastic mind. I was endeavouring to find some effects 
 of the material through which an electric current was 
 passing on the intensity of that current in violation 
 of Ohm's law. Looking up the literature, I found that 
 this law rested on very slender experimental basis, and 
 my experiments actually seemed to indicate a deviation, 
 observable when an alternate current was superposed 
 on a direct one. Weber entered with great spirit into 
 the question, though the result apparently shewed a 
 breakdown of the law, in the opposite direction from 
 
1 6 Lecture I 
 
 that which he anticipated in consequence of his own 
 theoretical speculations. My experiments ultimately 
 led to an investigation by a British Association Com- 
 mittee, resulting in a vindication of Ohm's law to a 
 very high degree of accuracy\ The directorship of the 
 laboratory of Gottingen had, shortly before my arrival, 
 been handed over by Weber to Riecke, who still holds 
 it. There was only one student at work beside myself; 
 for the purposes of his doctor dissertation he was 
 magnetising ellipsoids and testing magnetic formulae 
 in the orthodox fashion. 
 
 I now turn to describe a laboratory of very different 
 character and ambitions. Helmholtz, who then stood 
 at the height of his power, naturally drew the most 
 promising students from all parts of Germany to Berlin, 
 where they worked in the two or three rooms, which 
 then constituted his laboratory. Most of them were 
 preparing their doctor dissertation on some subject 
 directly rising out of Helmholtz' work. As examples of 
 such work, in which the generally limited and some- 
 times moderate powers of a student were utilised to 
 the fullest extent to advance science, I may mention 
 Moser's work on the electromotive forces set up between 
 two solutions of the same salt in different states of 
 concentration, and E. Root's verification of the pene- 
 tration of hydrogen into platinum under the influence 
 of electric polarisation. Needless to say, no efforts 
 were made in this laboratory to push numerical measure- 
 ment beyond its legitimate limits, and though most of 
 the work done was quantitative in character, qualitative 
 
 ' The experiments were designed by Maxwell and performed by 
 Chrystal. 
 
The Laboratory of Hehnholtz 17 
 
 experiments were not discouraged. Goldstein was 
 working at his electric discharges in high vacua, and 
 trying to explain effects he discovered near the kathode 
 by a theory which I know to have been distasteful to 
 Helmholtz ; yet no word did he ever say to discourage 
 the purely experimental side of those experiments. I 
 always considered Helmholtz' laboratory, such as it 
 was at the time, the ideal of what a teaching laboratory 
 should be. There were no doubt some inconveniences 
 owing to want of space, each student having only a 
 table at his disposal, but the simplicity and compactness 
 of the arrangement allowed Helmholtz to make his 
 round every day and to speak separately to each 
 student. We profited not only by the advice given to 
 us individually, but also by the suggestions made to 
 our colleagues, and reaped the advantage of a detailed 
 study of one problem without losing the wider culture 
 of becoming interested in a number of different 
 questions. The lesson I then learned in the conduct 
 of a laboratory was still further impressed upon me 
 during a visit which I paid to Berlin a few years later, 
 when a new laboratory had been built under Helmholtz' 
 own direction. In place of a small number of badly 
 furnished and crowded rooms, there was now a noble 
 building, impressive on the outside and perhaps also 
 inside to the casual visitor, but all the soul and scientific 
 spirit of the old place had gone. The laboratory was 
 constructed on the principle so dear to beginners, that 
 every one should have a private room in which he could 
 set up his apparatus without fear of outside inter- 
 ference. In consequence, Helmholtz no longer found 
 the time personally to see and to advise the students. 
 
1 8 Lecture I 
 
 who worked without a common bond and generally 
 without scientific impulse, and mainly for the purpose 
 of completing a dissertation of sufficient merit for 
 their degree. Experience to my mind shews that the 
 system of closed compartments in a scientific laboratory, 
 which is still adhered to in many cases, encourages 
 secrecy and does not foster the true scientific spirit in 
 the community. It is significant that Helmholtz did 
 not long retain the directorship of the " Institut," but 
 resigned shortly after its opening. 
 
 The two laboratories at Berlin such as I knew 
 them represent two extremes, but how is the designer 
 of a new building to avoid the inconveniences and 
 combine the advantages of both ? How is he to give 
 increased space, better equipment, greater facilities of 
 research, without loss of that contact between the 
 students, and between the student and professor which 
 is the essence of the educational value of a laboratory ? 
 The problem is difficult, but not impossible to solve, 
 if we bear the main objects of practical instruction 
 in mind. 
 
 Though its educational value cannot be called into 
 question, it is doubtful whether the systematic training 
 in practical work, which has grown up during the last 
 twenty-five years, has had a material influence on the 
 progress of science. In entering into the question of 
 laboratory organisation, I am therefore laying myself 
 open to the charge of wandering from my subject ; but 
 in appreciating the different agencies that have been 
 at work to increase knowledge, we cannot altogether 
 ignore one, on which strong hopes were built, and which, 
 if it has hitherto failed, partially or completely, has done 
 
Value of Laboratory Training 1 9 
 
 so because it has not yet found its right methods. 
 We may hope to remedy this most quickly by facing 
 the facts : and the cardinal fact to bear in mind is that 
 previous to 1870, when laboratory instruction, if given 
 at all, was sporadic, the experimental skill of inves- 
 tigators was as great as it is now. We need only 
 mention the names of Faraday, Joule, Helmholtz, and 
 Regnault, in support of that contention. The con- 
 clusion is irresistible that an intelligent student possess- 
 ing a sufficient theoretical knowledge, is capable of 
 starting research work in Physics without previous 
 special training. It is not for him that complicated 
 laboratory courses have been designed, but for the 
 ordinary student, who requires a certain amount of 
 theoretical knowledge, which is acquired far more easily, 
 if the principal phenomena are brought to his direct 
 notice by practical work. The future investigator will 
 no doubt ultimately save time, if at an early stage he 
 acquires a certain technical skill, and becomes acquainted 
 with physical methods ; but otherwise the efforts of the 
 teacher should be directed to stimulating his scientific 
 activity rather than sending him through all manipula- 
 tions which he might possibly have to perform. If we 
 place this stimulation of mind in the foreground, we 
 must all the more insist on the necessity of securing an 
 intimate intellectual intercourse between students and 
 Professors. Accepting this as the correct view, nothing 
 can be more antagonistic to the true interests of a 
 student than his condemnation to solitary confinement 
 in a cell, while he is performing his first research. 
 This tendency of laboratory designers to favour solitary 
 work may be partly responsible for the small number 
 
20 Lectiiyc J 
 
 of experimental discoveries wliicli can l»e traced to the 
 direct inlluence of laboratory teachin^^ 
 
 The organisation of lahoratory work in I',ngland 
 has to some; extent been cnntrdllcd l»y examination 
 re(]uirements, tliougli in \\v (;arly part <jl the peri(jd 
 which I am considering, there was no such tiling as 
 laljoratory (examination. Nevertheless, the more rigid 
 method of instruction which resulted from the control 
 (il ex.imin.'itinn in other dep.irlments, imconsci(;usIy 
 alfected pr.ictical teaching, ;ind led mc;re (|uickly to a 
 system.itic organisatitjn with all its .advantages and draw- 
 backs. W'hei'c the number of students to be taught is 
 small, all organisation can b<; disp(:used with, lor llx; 
 student is most (|uickly taught to depend on his own 
 resources when the teacher is still tentatively f(,'eling his 
 way either in conse(|nen(:e ()\ his own immatur(;ni,-ss f;r 
 the unsettled state of his subject. /At any r.ite, 1 fc'lt in 
 my own case the aflvantag<: 1 derived from beginm'ng 
 my practical training at a time when then; were no 
 te.xt-books on the subject, lialfour Stewart introduced 
 laboratory teaching in the old buiMings of (^wens 
 LolI(;ge soon alter his ajj|xiintment in 1.S70. A terrible 
 railway accidc.-nt which nearly cost him his life int(;r- 
 rupted his work, and fjniy in the autumn of 1871, when 
 I entered the college, was he able to resmrie it. 
 1 hough Stewart was always njady to help us in all 
 serious dilliculties, the main part of the- instruction was 
 handed over to teachers, whose knowledge was tempered 
 by sufficient ignorance- to allow them to be dogmatic 
 without carrying conviction. In consefjuence we -the 
 students — not being able to distinguish between bluff 
 and solid teaching, took nothing on trust and started 
 
Instruction at Mmichester 21 
 
 iiu'fstigations of our own, to confirm or disprove 
 our instructors. To cite one case, I remember being 
 extremely puzzled by finding that in measuring the 
 refractive index of a glass prism, two widely separated 
 images of the sodium line appeared. I thought of 
 internal reflexions as an explanation, but the position 
 ot the lines tlid not support this and I ultimately con- 
 sult«,'d one of the teachers, who in a superior tone 
 rebuked me lor not knowing that the sodium line was 
 doulile. 1 le had not noticed that there was an angle 
 ot several degrees between the components and one 
 ol tlu^ images, tlu>ui'h \'ello\v was standing riuht over 
 the blue part ot the continuoLis spectrum accompanying 
 the other component. In conjunction with a fellow 
 student I pursued the enquir\', and ultimately we 
 louiid ill some I'Vench book that imannealed olass was 
 doubly retractive. It seems that the prism we experi- 
 mentetl with, was originalh' purchased to shew this 
 (_louble retract ii>n, and t(.> the end of my tenure of the 
 protessiirship 1 shewed it annual!)' to ilkistnite the 
 effect. 
 
 Ballour Stewart, whose work on the Theory of 
 Exchanges holds a high position in the history of 
 rhxsics, was an insj)iring teacher because he was one 
 ot tile tew who did mit discourage attempts to discover 
 new tacts. He himselt was always bus\' tr)ang to 
 open out new fields of enquiry. In a room partly used 
 tor laboratory instruction, he experimented with a disc 
 set \\\\.o rapiil rotation in \-acuo, with a \-iew to 
 finding a thermal effect. He was led to try this, because 
 he could not reconcile in his mind the continuance 
 of thermal equilibrium with the change in the radiation 
 
22 Lecture I 
 
 and absorption, which in consequence of the Doppler 
 effect would result when the velocity of a body changes. 
 Balfour Stewart experimented from the point of view of 
 one, who thought that a breakdown of the second law 
 of thermodynamics was conceivable when new con- 
 ditions were introduced. Wien at a later period was led 
 by the same idea, but basing himself on the truth of 
 the second law, he deduced the relationship between 
 the intensities of radiation for different periods which 
 is consistent with that law, when the bodies within the 
 enclosure are in motion. He thus made an important 
 addition to the science of Radiation. 
 
 I came into closer contact with Balfour Stewart 
 when I returned to Manchester in 1873 as Demon- 
 strator of Physics. His mind was then running on 
 the possible discovery of a minute variation of gravity 
 due either to chemical combination or to a screening 
 effect which might shew itself when the weight of a 
 disc is determined with its plane first in the vertical 
 and then in the horizontal position. He also made a 
 number of experiments to discover some kind of inter- 
 ference between two crossing electric currents, as when 
 currents are made to pass at right angles to each other 
 from two independent cells through an electrolyte. 
 In none of these experiments was he successful, but 
 the students benefited from his alertness and freshness 
 of mind. My own ideas at the time were dominated 
 partly by Weber's explanation of diamagnetism, through 
 electric currents circulating round molecules and partly 
 by the first impressions I received on reading Maxwell's 
 Theory of Electricity. I formed some idea that if light 
 is an electrodynamic disturbance, Weber's currents 
 
ReminisceMces 23 
 
 might be brought into connexion with it and con- 
 sequently tried a number of experiments (described in 
 a Note Book which is still in my possession) on the 
 influence of magnetism on the vibrations emitted by a 
 sodium flame. All this was only vaguely before me, 
 but I placed a sodium flame in the strongest magnetic 
 field which the laboratory could provide and looked at 
 the result through a spectroscope. To my delight I 
 found a widening of the lines. I sometimes wonder 
 what would have been the result if I had then published 
 what might now be considered as an anticipation of 
 Zeeman's discovery ; but as a matter of fact the effect 
 was spurious and with the spectroscopic resolving 
 powers used, I could not possibly have seen the real 
 effect. After a few days of exciting work, I found 
 indeed that the steel spring of the slit was drawn 
 aside whenever the magnetic field was excited, and that 
 the broadening of the line was simply due to a widening 
 of the slit. By arranging the experiment differently, 
 I could produce a narrowing of the line in place of the 
 broadening. An experience such as this ought to 
 shew the danger of taking it too easily for granted, 
 that an observation recorded in some old book is an 
 anticipation of a later discovery, when in reality it 
 is probably only the result of a careless experiment 
 accidently simulating an effect which is real, but can 
 only be detected by more refined methods. 
 
 The opening of the Cavendish Laboratory at 
 Cambridge marks the beginning of a new era of 
 physical discovery ; for though the results achieved in 
 actual research work were small during the first years 
 of its existence, the indirect effect of the laboratory 
 
24 Lecture I 
 
 and of Maxwell's personality on Cambridge thought 
 was very considerable. Nor was this influence con- 
 fined to Cambridge, for it was Maxwell's writings 
 which inspired Hertz in the work which laid the 
 foundation of a new school of physics in Germany. 
 
 The Cavendish Laboratory owes its existence to 
 the generosity of the Duke of Devonshire, who was 
 Chancellor of the University, and as Chairman of a 
 Royal Commission on Scientific Education had been 
 impressed by the need of institutions in which experi- 
 mental research could be carried on. The building 
 was formally opened on June i6th, 1874. Lord 
 Kelvin (then Sir Wm Thomson) had declined the 
 new Chair of Experimental Physics, and after some 
 hesitation, and only in consequence of considerable 
 pressure put upon him by his scientific friends, Clerk 
 Maxwell consented to be nominated. The formal 
 appointment was made on March 8th, 1871. The 
 introductory lecture, of which I have already quoted 
 a significant passage, shews us the ideal which was in 
 Maxwell's mind, but Universities do not alter their 
 habits very rapidly and the whole Cambridge system 
 of education did not fit in well with the new develop- 
 ment. The qualification for graduation on the scientific 
 side was dominated by the Mathematical Tripos, which 
 covered a range of studies too wide already to admit 
 of time being available for additional instruction in 
 experimental work. 
 
 Physics looked at from the theoretical point of view 
 always held a position and occasionally an important 
 one in the Tripos, though the policy of the University 
 with regard to that subject was not always consistent. 
 
Physics at Cambridge 25 
 
 In 1849, the mathematical theories of electricity, mag- 
 netism, and heat, were definitely excluded from the 
 examination, and capillary attraction soon followed, but 
 all these subjects were re-introduced in 1873. The 
 detachment of the theoretical study of Physics, as it 
 then was conducted at Cambridge, from the experi- 
 mental presentation of facts, led in some cases to a 
 remarkable antagonism to ocular demonstration which 
 is illustrated by a characteristic incident. Clerk 
 Maxwell who possessed an innate desire to see what 
 he could with his own eyes, had taken considerable 
 trouble in cutting and grinding a plate out of a doubly 
 refracting crystal to shew conical refraction. The ex- 
 periment is difficult, and delighted at its successful 
 accomplishment Maxwell met one of the mathematical 
 teachers of the University. "Would you like to see 
 Conical Refraction?" asked Maxwell. "No," replied 
 Todhunter, " I have been teaching it all my life, and I 
 do not want to have all my ideas upset by seeing it." 
 That this remark was not made jocularly is shewn 
 by a passage which occurs in an essay on the Conflict 
 of Studies, in which Todhunter discusses the advis- 
 ability of introducing experimental illustrations into the 
 lessons given in schools. He declares himself as 
 opposed to it on the ground, that an experiment which 
 is not intended to bring out a new fact is useless, and 
 proceeds as follows: "It may be said that the fact 
 makes a stronger impression on the boy through the 
 medium of his sight, that he believes it the more 
 confidently. I say that this ought not to be the case. 
 If he does not believe the statements of his tutor — 
 probably a clergyman of mature knowledge, recognised 
 
26 Lecture I 
 
 ability, and blameless character — his suspicion is irra- 
 tional, and manifests a want of power of appreciating 
 evidence, a want fatal to his success in that branch 
 of science which he is supposed to be cultivating." 
 
 Todhunter was a highly cultivated man and an 
 able mathematician ; he undeniably believed in the 
 mission of a University to advance knowledge. 
 "There appear to be," he writes, "three distinct 
 functions of the University. One is that of examina- 
 tion, one that of teaching, one that of fostering original 
 research : the first of these three has practically been 
 as yet most regarded, and many of us hope that it will 
 in future decline either absolutely or relatively by the 
 increased development of the other two'." Or again, 
 consider this passage : — 
 
 " Of that public professional instruction which 
 is often regarded as the essence of a University, 
 there is comparatively little in Cambridge. During 
 the last quarter of a century out of the whole 
 range of mixed mathematics lectures have been 
 regularly delivered by professors only on the following 
 subjects: Optics, Hydrostatics, Astronomical Instru- 
 ments and Lunar Theory. I have myself sometimes 
 received letters of enquiry from strangers who wished 
 to study certain branches of pure or mixed mathe- 
 matics, and contemplated spending a year or more at 
 the University for the advantage of professional 
 lectures. It was not quite satisfactory to be com- 
 pelled to reply that there was scarcely any of that 
 machinery for teaching of which the applicants seemed 
 to assume the existence." 
 
 ' Conflict of Studies, p. 238. 
 
Physics at Cambridge 27 
 
 While Todhunter believed in advancing knowledge 
 by experiment, he attached no value to one which is 
 made to improve the mind of the student. His frame 
 of mind was that of one satisfied with reading about 
 foreign countries in books, believing that the traveller 
 who goes to see them, only incurs trouble and expense 
 while he would gain morally by staying at home and 
 shewing his faith in the original explorer. It would be 
 wrong to take Todhunter as the type of a numerous 
 class ; he was a freak who differed from his type in 
 having the courage of his opinions. It is not surprising, 
 considering the manifold conflicting interests which must 
 always retard the progress of a University, that the 
 Cavendish Laboratory only slowly obtained its grip on 
 Cambridge education ; that it obtained it as quickly 
 as it did was a sign of the rapidly growing influence 
 of men animated by the true progressive spirit which 
 spread through the University, and is the cause of the 
 pre-eminent position which Cambridge now holds in 
 the world of science. 
 
 During the spring of the year 1876 I determined 
 to give up the position I held at Owens College, in 
 order to work a year or two at Cambridge : the time 
 of my stay there was gradually extended until my 
 final return to Manchester in 1881. I was well 
 received by Maxwell, but otherwise was looked upon 
 with friendly curiosity by a University that was not 
 accustomed to see anyone coming to it merely for the 
 purposes of research, without previously availing them- 
 selves of the benefits of degree courses and the prospect 
 of the usual University rewards. After a little trouble 
 and delay I was allowed to enter one of the colleges. 
 
28 Lecture I 
 
 My position was irregular, if not illegal, for though 
 inscribed as an undergraduate I was allowed to stay 
 five years without passing the " Little Go," though 
 subject, of course, to the discipline of the Proctor. 
 I remember meeting that formidable officer of the 
 University at dinner, and walking home with him 
 unchallenged and not fined, though I was without 
 cap and gown. Research students were only recog- 
 nised at a much later date, and though their numbers 
 are now large, I feel sorry that it should not have 
 been possible to attract them by the advantages to be 
 gained through working under a leader of science rather 
 than through the bribe of a degree. 
 
 The Cavendish Laboratory presented a different 
 aspect then, from that which it assumed at a later time 
 when students began to fill its rooms. As far as I 
 remember, William Garnett, who acted as demon- 
 strator, and George Chrystal, now Professor of 
 Mathematics at Edinburgh, were the only regular 
 workers. W. M. Hicks, now Professor of Physics 
 at the University of Sheffield, came occasionally ; but 
 shortly afterwards R. T. Glazebrook and W. N. Shaw 
 joined us, and we began to consider ourselves crowded. 
 The facilities of the laboratory were not what students 
 now expect. We had to charge our own batteries 
 and learn a little glass blowing and ordinary workshop 
 manipulations, as there were no instrument makers 
 nearer than London. But every one who worked in 
 the Cavendish Laboratory during these last years of 
 Maxwell's life must retain for ever the impression of 
 Maxwell's intellectual influence and charm of manner. 
 Except when prevented by his own ill-health or that of 
 
Clerk Maxwell 29 
 
 his wife, he visited the laboratory daily and made the 
 round of the rooms in which any work was going on. 
 He might ask a question about the progress of one's 
 experiments, but more generally spoke about the 
 subject which occupied his thoughts at the time, for 
 he was much absorbed in his own ideas and not 
 always quick in bringing his mind to bear on a fresh 
 subject. It would occasionally happen that he took 
 no notice of a question put to him, leaving it doubtful 
 whether he had heard it or not, but next day he 
 would probably start his conversation with : " By the 
 way, there was a question you asked me yesterday 
 
 and I have been thinking about it " Then would 
 
 follow a suggestive and fully considered answer. All 
 who were brought into contact with him must re- 
 member some of the quaint and humorous sayings 
 which gave such a charm to his conversation. His 
 thoughts at the time seemed to run much on what is 
 now called the equi-partition of energy. Boltzmann's 
 researches had recently been published, and Maxwell 
 seemed to accept them, though only with hesitation, 
 because he could not see how far they might lead us. 
 I remember especially his saying that Boltzmann's 
 theorem if true, ought to be applicable to liquids and 
 solids as well as gases. 
 
 Maxwell took an intense interest in the publication 
 of the papers of Cavendish which he was then editing, 
 and with great enthusiasm repeated most of the ex- 
 periments described in them. He seemed specially 
 fascinated by the manner in which Cavendish antici- 
 pated subsequent discoveries made with delicate 
 instruments by converting himself into a galvano- 
 
30 Lecture I 
 
 meter. No effects of electric currents known to 
 Cavendish could serve for the measurement of their 
 strength, and he therefore estimated the intensity 
 of the physiological shock which he felt when a 
 current was suddenly sent through his body, the 
 experiments being arranged so that two currents were 
 declared to be equal, when the shocks they produced 
 appeared to be equally intense. Every one who came 
 to the laboratory at that time had to submit himself 
 to a sometimes unpleasant test, in order to convince 
 himself that the method could give consistent results. 
 
 It is impossible to say whether Maxwell was 
 satisfied with the position the laboratory took in the 
 first years of its existence. It grew steadily in im- 
 portance, but the progress was slow and required time 
 to develop, but time — alas — was not vouchsafed to 
 him. Maxwell died in the autumn of 1879 at the age 
 of 48. His ideals soon began to be realised under 
 the professoriate of his successor. Lord Rayleigh. 
 
 Systematic instruction which now became necessary 
 in the preparation for the Natural Sciences Tripos 
 was excellently supplied by R. T. Glazebrook and 
 W. N. Shaw, who acted as demonstrators. For the 
 purpose of encouraging research, Lord Rayleigh 
 planned to take up a problem requiring extensive 
 series of measurements, in which a number of students 
 could take part, who would thus gain experience and at 
 the same time the power of combination which is only 
 possible in an organised laboratory would be utilised. 
 His first choice of subject, the redetermination of the 
 electrical unit of resistance was a most appropriate one, 
 because the revolving coil of the British Association 
 
The Cavendish Laboratory 31 
 
 which had served for the first determination of the Ohm 
 was available in the laboratory, and the accuracy of this 
 determination had been called in question by Rowlands 
 work. To its shame, be it said, no one in the University 
 was willing to help, and it was only to assist a good 
 cause and to set a good example that I regretfully 
 left my own experiments to take part in the redeter- 
 mination of the Ohm. I have never regretted, 
 however, being associated with Lord Rayleigh in the 
 early stages of the work in which Mrs Henry Sidgwick 
 soon joined, and which proved to be of permanent 
 value to electrical science. 
 
 If I have entered, perhaps too much, into a 
 discussion of laboratory organisation, it was because 
 many hopes were founded on the introduction of 
 practical work as an essential portion of degree 
 courses. There is no doubt that the general standard 
 of knowledge has been raised through the more in- 
 timate contact of a student with the facts of nature ; it 
 is also open to us to believe, though it may be difficult 
 to prove, that the progress of knowledge as tested by 
 the discovery of new facts was accelerated. 
 
 It is time, however, that we should return to our 
 main object, which is to trace the outline of knowledge 
 before the great revival of experimental discovery. I 
 have already indicated the position of the main 
 branches of Physics, but there were a few detached 
 portions in which experiment was ahead of theory, and 
 which to some extent were destined to form a con- 
 necting link between the old and the new aspect of 
 science. The discovery of spectrum analysis and its 
 application to the study of the chemical composition of 
 
32 Lecture I 
 
 celestial bodies, kept alive our faith that man had not 
 reached the end of his endeavours to bring new facts 
 to light, and we began to recognise that the radiations 
 of a luminous gas were likely to give us valuable 
 information on molecular structure. The first simple 
 conception that the spectrum of an element was 
 characteristic of an atom, and accompanied that atom 
 through its various combinations with other bodies, 
 was rudely disturbed when it was found that com- 
 pounds had their own spectra, and that even an 
 elementary body could under different circumstances 
 possess more than one spectrum. This last fact was 
 not generally admitted without opposition, extending 
 through several years, because it is always difficult to 
 answer objections based on a supposed impurity of the 
 material used. The facts however soon began to be 
 too strong for the doubters, and multiple spectra 
 established their position. The longer spectroscopy 
 was studied, the greater its complications appeared, 
 especially in the case of gases rendered luminous under 
 reduced pressure. To clear up the difficulties it 
 became necessary to examine more closely the dis- 
 charge itself and this cleared the way for one of the 
 great advances of which I shall have to speak in a 
 subsequent lecture. 
 
 An example of an experimental discovery arising 
 out of irregularities observed in the course of accurate 
 numerical measurements is to be found in Crookes' 
 experiments, which led to the construction of the 
 radiometer. Wishing to determine accurately the 
 atomic weight of thallium, Crookes designed a 
 balance placed in a receiver from which the air could 
 
The Radiometer 33 
 
 be removed by means of an air pump. When the 
 pressure had been reduced sufficiently to eHminate all 
 effects of air currents, certain anomalies were observed 
 indicating an apparent repulsive force acting on bodies 
 exposed to heat radiations. An experiment made 
 by myself proved that the forces depended on the 
 presence of the residual gas still left in the receiver, 
 and Crookes himself was ultimately able to improve 
 the vacuum sufficiently to diminish and finally to 
 destroy the effect. These experiments, which roused 
 considerable interest at the time, take their place in 
 the history of science not so much because they have 
 given us an instrument which has been useful in the 
 study of radiation, but because they shewed the im- 
 perfections of the so-called chemical vacua, which were 
 then believed to contain no traces of air. As soon as 
 this was recognised to be wrong, methods of removing 
 the air from a vessel, rapidly became more effective. 
 
 Our brief account of the state of science when the 
 Cavendish Laboratory was founded, would not be 
 complete without some reference to a fundamental 
 theory of matter, which absorbed a good deal of the 
 energy of Cambridge mathematicians, and though now 
 abandoned and neglected, will always stand as a 
 monument of a great effort. When the hydrody- 
 namical investigations of Helmholtz had shewn the 
 possibility of constructing a universe in which the 
 ultimate units were hydrodynamic entities. Lord 
 ■ Kelvin with characteristic energy developed a theory, 
 in which the atom was considered to be a vortex ring 
 in an incompressible fluid, and this theory fascinated 
 and satisfied the mind because it went to the very 
 
 s- 3 
 
34 Lecture I 
 
 foundation of the edifice of nature. Kirchhoff, a man 
 of cold temperament, could be roused to enthusiasm 
 when speaking about it. It is a beautiful theory, he 
 once told me, because it excludes everything else. 
 For the present the theory is set aside, but may yet 
 be revived in a different shape. In one respect its 
 point of view resembles that of later theories, because 
 it recognises only one substance, and makes the energy 
 of a moving atom and therefore its mass, depend on 
 the kinetic energy of the surrounding medium. The 
 study of electrical phenomena has led to the same idea 
 by a different road, and it will be my duty in the 
 next lectures to shew how a more perfect representation 
 of the ultimate constitution of ponderable matter is 
 foreshadowed by the more recent electrical discoveries. 
 
LECTURE II 
 
 When we consider two rival theories in any 
 branch of human knowledge, we are sometimes drawn 
 towards one or towards the other, not by any process 
 of reasoning, but by an instinctive feeling which may 
 be so strong, that we unhesitatingly reject one of the 
 alternatives. We flatter ourselves in such cases that 
 our choice is guided by convictions gained in previous 
 studies, and is therefore based on a sound though 
 unconscious application of the reasoning faculty. 
 This use of an instinct, which we like to call 
 "common sense," is justified when applied to one 
 portion of a subject, that has already been examined 
 from a wider point of view ; but can we trust to a 
 mere bias — which may be a prejudice as well as 
 a true instinctive guide — in deciding a question of 
 fundamental character ? 
 
 If I open my second lecture with this psycho- 
 logical query, it is because I should like you to 
 consider briefly what justification there is for that 
 strong instinctive feeling, which denies action at a 
 distance. By action at a distance we understand, 
 that two bodies can act on each other, without beine 
 connected, by some medium which transmits the force. 
 
 3—2 
 
36 Lecture II 
 
 The denial of action at a distance implies that we only 
 recognise forces, which directly push or pull the body 
 that is to be moved. There can be no question that, 
 as an article of faith, the "push and pull" theor}' of 
 force has dominated the scientific thought of the last 
 half centur)" ; and though there are signs ot a coming 
 reaction there is still an overwhelming preponderance 
 of feeling in favour of an almost axiomatic negation 
 of action at a distance. It would be interesting to 
 trace in detail the causes of such firm belief in the 
 two mutually exclusive doctrines that "a body cannot 
 act where it is not," and that " two bodies cannot 
 occupy the same space," but we must content ourselves 
 with a few brief reflections. To me it seems that the 
 dogmatical denial of action at a distance is a survival 
 of the ancient anthropomorphic explanation of natural 
 phenomena. Bodies are moved on the surface of 
 the earth by the pulling or pushing of living bodies, 
 and we therefore unconsciously associate a sense of 
 muscular effort with the torce that causes motion. 
 When a body drops down from a height, without any 
 apparent cause, we instinctively assign to its fall a 
 cause essentially identical with that which can lift it 
 up again ; for to connect two similar effects with similar 
 or identical causes, is the most elementary instinctive 
 principle to which probably all others maybe reduced. 
 Is not the fundamental instinct, which makes action 
 at a distance seem unnatural to us, identical with that 
 which placed the sun in a chariot drawn by horses in 
 order to explain his motion across the heavens ? The 
 formulation of the principle of contact forces as an 
 act of faith is modern, but our pride in it mav to 
 
Action at a Distaitce 37 
 
 some extent be chastened, if we recognise its want of 
 originality. 
 
 Let us look at the question a little more closely. 
 When we push a body with our hands, we take the 
 effect to be an example of a force acting by contact 
 and not "at a distance." But if our senses were much 
 more acute than they are, and we could distinguish 
 by sight the separate molecules of a body, we should 
 probably see no contact anywhere, the molecules being 
 always separated by small intervals, and our conclusion 
 in witnessing the effect of the pressure exerted by our 
 hand would be that the body that is pushed begins to 
 move away before the molecules of our hand have 
 come into contact with it. Our primitive feeling then 
 would be in favour of action at a distance, which shews 
 that our present instinctive conviction against it, may 
 only be due to defects of our senses. 
 
 If I am anxious that you should realise how weak 
 are the grounds on which we deny on principle action 
 at a distance, it is only to lay the greater stress on the 
 real advance which science has made in consequence 
 of the belief that all action between two bodies is 
 transmitted through a connecting medium. Two 
 lessons may here be learned. One is, that the tem- 
 porary success of a doctrine does not necessarily justify 
 the grounds of its foundation, and the other that pro- 
 gress in science is more often achieved by a definite 
 hypothesis, which may be followed up and tested, than 
 by a wider and perhaps more philosophic doctrine, 
 which cannot be disproved, because it does not en- 
 deavour to go deeper than the mere descriptive 
 classification of phenomena. At present, we may take 
 
38 Lecture II 
 
 it, that the doctrine of " no action at a distance " has 
 secured its successes, mainly because it opened out a 
 new field, which could be brought to the test ot ex- 
 periments. These successes must be placed to the 
 credit of that school of science which explains the 
 unseen by means of models such as we can construct 
 with material bodies. 
 
 It is interesting to trace the gradual development 
 of scientific ideas on the properties of the medium 
 which transmits the vibrations of light. \A'e must, for 
 this purpose, go back to the foundation of modern 
 astronomical science by Sir Isaac Xewton. The sim- 
 plicity and success of his treatment of gravitation 
 tempted philosophers to a similar proceeding in in- 
 vestigating the problems of molecular forces. The 
 ultimate constituents of matter came therefore to be 
 looked upon merely as centres of force and the elastic 
 properties of solids had to be deduced from the pro- 
 perties of such centres. When it was realised that the 
 vibrations ot the luminiferous aether are similar to the 
 vibrations of an elastic medium, the aether was endowed 
 with an atomic constitution, and its power to transmit 
 transverse undulations had to be traced back to central 
 atomic forces. This was a logical proceeding, but it 
 failed because it took scientific men too far bevond 
 what could be ascertained by observation and experi- 
 ment. The next step restricted the problem, in ac- 
 cordance with the general experience, that the difficulty 
 in going ahead does not lie so much in making a step 
 sufficiently large to constitute a real advance, as in 
 finding one small enough to be taken with safety. If 
 the aether behaves like an elastic body, it is not 
 
Elastic Solid Theory of Light 39 
 
 necessary to make the investigation more difficult by 
 enquiring what it is that gives a body the property of 
 elasticity, but it is sufficient to deduce the phenomena 
 of light, postulating as an assumption that the ajther 
 possesses the elastic properties of known solid bodies. 
 The elastic behaviour of a finite portion of a body may 
 be studied without reference to any molecular forces, 
 and assuming these properties to persist when that 
 portion becomes indefinitely small, we may calculate 
 the effects of disturbing forces which are applied to 
 any portion of the body. The result will be substan- 
 tially correct, so long as the assumed elastic properties 
 hold for portions of the body as small as the length of 
 the wave. 
 
 It was the great work of Green and Stokes to 
 have treated the problem of light as one of elasticity, 
 by assuming no properties of the aether beyond those 
 which we know elastic solids to possess. The merit 
 of this work is by no means diminished by the ultimate 
 failure of the theory, for its failure could only be 
 established by the strict investigation of its con- 
 sequences. 
 
 While the elastic solid theory of light was being 
 developed by the great mathematicians of seventy years 
 ago, Faraday's genius gave us a new and lofty stand- 
 point for the outlook on electrical phenomena. Whether 
 his treatment of electric action, as transmitted through 
 a medium, permeated the scientific mind and affected 
 contemporary thought, is difficult to decide, but at any 
 rate it convinced Maxwell, and that was sufficient. 
 
 The strongest of our scientific " instincts " is our 
 ultimate belief in the simplicity of nature. If both 
 
40 Lecfinr. II 
 
 light and electrical attractions are transmitted through 
 a medium, it would revolt our feelings — that is to sav 
 our non-reasoning faculties — to assume, without strong 
 evidence to the contrary-, that two different media exist 
 for the two manitestations. But while it was possible 
 to explain light by means of a medium having the 
 properties of bodies known to our senses, this medium 
 failed at the same time to explain electric actions. 
 ^Maxwell's genius inverted the problem. He asked 
 himself what the properties of a medium should be in 
 order to account for the transmission of electrodynamic 
 effects, and he then discovered that such a medium 
 would transmit waves with a velocity of propagation, 
 exactly equal to that with which light was known to 
 be transmitted. Hence any medium which could ex- 
 plain electrical action could also explain light. This 
 is [Maxwell's greatest achievement and the foundation 
 of the electrodynamic theory of light. 
 
 Although I may go over ground that is familiar to 
 many of you, I shall briefly explain the main features 
 of Maxwell's theory, because I am anxious to point 
 out, that success was achieved mainlv through the 
 clearness and detlniteness of the physical conception, 
 on which the mathematical development was based. 
 
 Steady electric currents, if the current be defined 
 in the usual way, satisfy the laws of flow of an incom- 
 pressible fluid. If for instance, water flows down a 
 river bed which may be wide in some parts, narrow 
 in others, steep or nearly level in different places, the 
 volume of water which crosses each section in a certain 
 time is the same ever)'where. Hence, where the river 
 is deep or wide the velocity of the water must be corre- 
 
Maxwell's Theory 41 
 
 spondingly less than where it is shallow or narrow. 
 The same law holds in the case of electric currents 
 which are permanently maintained : the same amount 
 of electricity flows through every part of the circuit. 
 This suggests as a first hypothesis that a current of 
 electricity means the flow of something which always 
 behaves like an incompressible fluid. The essential 
 generalisation here lies in the word "always," which 
 extends what we observe in the case of certain electric 
 currents, where the observation can be carried out, to 
 all classes of currents, some of which differ very mate- 
 rially from those, for which the proof can be given. 
 Such generalisations are characteristic of the usual 
 procedure of scientific discovery when we apply what 
 can be proved in special cases to the general case and see 
 whether the wider proposition leads to contradictions. 
 To give, as an example, the case of a flow of electricity, 
 which seems to violate the conditions of an incom- 
 pressible fluid, imagine an electric battery E (Fig. i) 
 and connect its poles N and P to the 
 plates of a condenser, A and B. While 
 the condenser plates are charging, an 
 electric current passes through the battery 
 and the connecting wires. According to 
 the older view, while electricity thus 
 accumulates on A and B, there is no 
 current through the intervening space. 
 But if the flow of electricity is always like "' 
 
 that of an incompressible fluid, the current cannot 
 stop at the condenser plates, but must be completed 
 through the space between them. There should there- 
 
42 Lecture II 
 
 fore be an electric current through the non-conducting 
 space between the plates A and B when the battery is 
 connected, but only while the plates are charging. 
 When this is done there is no further current through 
 any part of the circuit. The question which now arises, 
 is whether the current which can temporarily pass 
 through a non-conducting medium has all the pro- 
 perties of an ordinary electric current, more especially 
 its electro-magnetic properties. [Maxwell assumes that 
 it has, and this carries with it the consequence, that a 
 chano-e of magnetic force causes an induced current in 
 a non-conducting as well as in a conducting medium. 
 Here then we have a generalisation which is 
 sufficiently precise to allow of its being followed up 
 mathematically and being confirmed or disproved by 
 experiment. 
 
 If at some point of space a change of electric force 
 takes place, as for instance when an electrified con- 
 ductor be moved backward and forward, there must, if 
 ?*Iaxwell is right, be electric currents in that part of the 
 medium which immediately surrounds the conductor ; 
 and these again must act by electro-magnetic induction 
 on the other adjacent parts of the medium. The 
 problem is quite definite and leads to a svstem of 
 equations for the propagation of an electrical disturb- 
 ance, giving a constant rate of propagation, which 
 can be calculated. The numerical data which are 
 necessary for the purpose are obtained by comparing, 
 e.g. in the above example (Fig. i) the electro-magnetic 
 force of the charging current with the electrostatic 
 attraction between the plates, when fully charged. It 
 
Maxwell's Theory 43 
 
 is found that the velocity of propagation of electro- 
 magnetic effects is equal to that of light. Hence we 
 can calculate the velocity of light, making use only of 
 purely electric measurements. This is convincing, and 
 leaves no reasonable ground for doubting, that light 
 and electrical actions are both transmitted through the 
 same medium : that light is in fact an electro-magnetic 
 phenomenon. 
 
 In his first endeavour to develop Faraday's theory 
 of electric action. Maxwell went beyond the necessities 
 of the case, so far as the electro-magnetic theory of 
 light is concerned. The scientific period of which I 
 am speaking was essentially one of model making. 
 Physicists still felt the necessity of having clear con- 
 ceptions, and distrusted mathematical symbols that had 
 no fully defined meaning. If the aether be called upon 
 to transmit electric and magnetic actions, scientific men 
 felt uncertain of their ground, unless they had some 
 notion how the necessary stresses and strains were 
 brought about. Hence Maxwell, who unquestionably 
 belonged to this school, which requires a mechanical 
 model for the conception of physical phenomena, 
 devoted a good deal of time to the specification of 
 electrostatic and magnetic strains in his medium. It is 
 not necessary to enter into that part of his work here, 
 because the question has for the moment been pushed 
 into the background ; nor is it necessary to enter into 
 the successes, which the electro-magnetic theory soon 
 achieved by overcoming the difficulties of the older 
 theories of light. 
 
 The issue on which Maxwell's results went consider- 
 ably ahead of experimental knowledge was the trans- 
 
44 Lecttire II 
 
 mission of electro-magnetic effects with finite velocity. 
 Everybody knows that the position of a delicately sus- 
 pended magnetic needle is affected by a magnet which 
 is placed in its neighbourhood, and if this magnet be 
 shifted the needle will in general turn to one side. So 
 far as our senses can judge, the action is instantaneous 
 and the needle turns at the moment the magnet is 
 displaced. But Maxwell says: No! not at the same 
 moment but after a certain interval of time only, 
 though this interval is too short to be apprehended by 
 our senses. If the distance of the magnet be one 
 metre the time of transmission of the effect would be 
 the three hundred millionth part of a second. Can we 
 obtain an experimental verification of this time of 
 transmission ? 
 
 Surprise may reasonably be expressed that while 
 Maxwell was surrounded at the Cavendish Laboratory 
 by a number of young physicists, who firmly believed 
 in his electro-magnetic theory, no attempt was made 
 by them to furnish an experimental proof of their 
 master's theoretical deductions. The explanation lies to 
 a great extent in Maxwell's habit of letting his students 
 go their own way and find their own problems. Unless 
 a student had asked him directly to suggest a problem, 
 I doubt whether it would have occurred to him to give 
 advice in the selection of a subject for investigation. 
 Nevertheless, and I speak from personal knowledge, 
 the desirability of an experimental proof of Maxwell's 
 theory was realised by Cambridge men, and other 
 British physicists, who were in contact with them, such 
 as FitzGerald. But the experimental difficulties seemed 
 formidable, notably as regards the emission of electro- 
 
Electro-magnetic Waves 45 
 
 magnetic waves of sufficient intensity to give a measure- 
 able effect at a distance. Had there been two equally 
 probable theories in the field, I doubt not the attempt 
 to carry out a crucial experiment would have been 
 made ; but we were perhaps over confident in the 
 inherent truth and simplicity of Maxwell's conception. 
 An extended experimental investigation would cer- 
 tainly have absorbed much time and labour which we 
 did not consider worth undertaking, considering the 
 indirect evidence in favour of the electro-magnetic 
 theory, which seemed to make the result a foregone 
 conclusion. We were wrong, because we forgot that 
 the great body of scientific thought abroad, and 
 to some extent in this country, was apathetic and 
 even reluctant to abandon an elastically solid aether 
 which had done good service, and to accept in its 
 place a medium, the properties of which were unlike 
 those of any known body. Even Lord Kelvin con- 
 tinued to place the weight of his great authority on 
 the scale of .the older views. 
 
 In the meantime Helmholtz, always on the alert 
 on the side of progress, suggested to his pupil, Heinrich 
 Hertz, to take up the experimental investigation of the 
 problem. Everybody knows the splendid manner in 
 which Hertz accomplished his task : how he over- 
 came the formidable experimental difficulties and 
 succeeded in convincing the scientific world of the 
 truth of Maxwell's theory. 
 
 The " Hertzian " wave carries a name which Hertz 
 would have repudiated, but it commemorates a man 
 who is associated with Maxwell in one of the greatest 
 scientific achievements. This habit of attaching names 
 
46 Lecture II 
 
 to laws, physical methods, or instruments, where the 
 law, method or instrument can be equally well described 
 by some descriptive name, is not one that should in 
 general be encouraged. Personal names are not merely 
 stumbling-blocks to the student, but— and this is my 
 main objection — they often establish false scientific 
 history. In the present instance and without wishing 
 to detract from the great merits of a personal friend 
 I see no reason why "electro-magnetic wave" is not 
 a sufficient designation. 
 
 Through the work of Hertz it was finally established 
 that electro-magnetic effects, radiant heat and light 
 are all transmitted through the same medium by dis- 
 turbances which in all respects are identical, and only 
 require different receiving appliances to render them 
 apparent to our senses. The recording or observing 
 instrument naturally has to adapt itself to the scale and 
 type of the disturbance. Just as we should use different 
 instruments, according as we wish to record the tidal 
 change of level of the ocean, or the waves engendered 
 by a storm or the ripples on the surface of the wave, 
 so are different appliances necessary to register different 
 disturbances of the luminiferous sether. A wave which 
 is many metres long can only be perceived by its heat- 
 ing effects, or by electro-magnetic means which fail, 
 when the scale is diminished so that the wave-length 
 is reduced to a fraction of a millimetre. When the 
 length is equal to about a thirteen-thousandth part of 
 a centimetre, the physiological function of our retina 
 begins to be effective and we can use optical apparatus, 
 until the length is still further reduced to about half 
 that value, when our eyes fail and we must have 
 
Electro-magnetic Waves 47 
 
 recourse to photographic means. There is still a gap 
 between the shortest wave which has been produced 
 electro-magnetically and the largest wave which has 
 been measured thermally', but this gap is gradually 
 being- bridged over. While the " Rest-strahlen " of 
 
 o o 
 
 Rubens and Aschkinass have very substantially in- 
 creased the range of measurement for heat radiations, 
 electric waves have been reduced in length by Righi, 
 Bose, Lampa and others. 
 
 The generation of electric waves on a large scale 
 by its far reaching application in wireless telegraphy 
 has spread in the public mind an almost awe-inspired 
 sense of the wonders of science. To many persons 
 wireless telegraphy appears to be the greatest of 
 recent scientific discoveries, and in the public mind 
 it seems more marvellous than the breaking up of 
 atoms through their own internal instability. I do 
 not wish to detract from the practical importance of 
 the results achieved, or underrate the high merits 
 of those, who by overcoming most serious practical 
 difficulties, have furnished us with a means of com- 
 munication which is likely to have an increasing 
 importance. I insist, however, that as a scientific 
 principle, the invention has taught us nothing of 
 conspicuous novelty. In the first place, wireless 
 telegraphy though historically arising out of the in- 
 vestigation of "Hertzian" waves, involves no scien- 
 tific facts which were not known to Faraday. It 
 
 ' Thermal effects accompany of course all disturbances of the 
 medium and give us the only true measure of their intensity, the 
 difficulty lies in the production of a quasi-homogeneous disturbance 
 the wave-length of which can be ascertained. 
 
48 Lecture II 
 
 depends on electro-magnetic action at a distance, 
 independently of the time of transmission which may 
 be instantaneous or not. If in the year i860 a 
 physicist had been asked, whether it were possible 
 to transmit electro-magnetic signals to a distance, he 
 would have answered that he actually does it every 
 day in his laboratory, when he closes an electric circuit 
 and deflects a galvanometer needle. If his questioner 
 had continued to ask whether these signals could 
 be transmitted beyond the range of his immediate 
 surroundings, he would no doubt have replied that 
 this depended on the scale of the experiment. He 
 might, quite likely, in an incautious moment, have 
 permitted himself to fix a distance, and to express 
 his conviction, that no power was ever likely to be 
 employed, and no receiving instrument of sufficient 
 delicacy, ever likely to be invented, which would allow 
 signals to be transmitted to more than a quarter 
 of a mile ; but I do not think he would ever have 
 acknowledged that any limit of distance was fixed 
 except by practical difficulties. Telegraphy without 
 wires is older than that through wires, and was 
 witnessed by the person who first saw bits of iron 
 attracted, when a loadstone was brought near to them. 
 The induction effects of strong currents rapidly 
 changing in intensity are also examples of wire- 
 less telegraphy which were long known to affect 
 instruments at considerable distances. The man who 
 looks upon his "wire" as on a common-place 
 phenomenon, while he considers his " Marconigram " 
 an almost super-natural message, is not very far 
 removed in scientific education from the old woman 
 
Wireless Telegraphy 49 
 
 who hung a pair of boots on the telegraph wires, so 
 that they should reach her son more quickly than if 
 they were sent by parcel post. An eminent man of 
 science, to whom 1 once expressed my astonishment 
 that wireless telegraphy was considered to be such 
 a novel and extraordinary achievement, illustrated this 
 general attitude by the remark, that, if people had 
 been accustomed to speak all their lives through 
 speaking tubes they would be greatly surprised by 
 the discovery that they could also speak without them. 
 This puts the case exactly, the " tubeless speech " is 
 the accurate counterpart of the "wireless telegram." 
 The experiments of Hertz repeated and extended 
 by Lodge and others, rapidly secured the final triumph 
 of the electro-magnetic theory, and models of the aether 
 illustrating the new ideas were soon forthcoming. One 
 essential point to be attended to was, that the aether 
 should be incompressible, and a further specification 
 which was not essential but generally accepted for 
 the sake of simplicity, made the displacement of the 
 aether consequent on an action of electric force linear 
 and in the direction of that force. This carried with 
 it the supposition, that an electric current through a 
 conductor means the flow of an incompressible aether 
 through that conductor. The mechanism of the for- 
 mation of an electrostatic charge at the surface of 
 the conductor would then be represented by a model 
 in which an incompressible liquid is forced through a 
 tube, which is closed at its ends by a flexible elastic 
 membrane. The liquid would push out the membrane 
 until the elastic reactions balanced the pressure of the 
 fluid ; and near the end of the tube, there would be 
 
50 Lectitre II 
 
 elastic strains and stresses in the partially yielding 
 medium outside. These strains and stresses repre- 
 sented the electric strains and stresses. I have entered 
 into this question because the mechanical representation 
 of the mechanism of the electric current adopted at 
 first as a picture of what might happen, soon became 
 an article of dogma which delayed further progress. 
 
 If we only had to consider the passage of electricity 
 through solid conductors, we might still be under the 
 sway of the simple representation which I have 
 sketched out ; but neither conduction through electro- 
 lytes nor through gases could be made to fit into this 
 simple scheme. Though neglected for a time, liquids 
 and gases had ultimately to be taken into account. 
 Faraday's laws of electrolysis had proved, that in the 
 case of liquid conductors the passage of a certain quantity 
 of electricity is always associated with the transference 
 of a definite quantity of matter, and had suggested 
 that each atom or radical which was set free by 
 electrolysis, carries a definite quantity of electricity. 
 Maxwell felt the difficulty of explaining Faraday's 
 laws by means of his own views on conduction, and 
 realised the importance of clearing up the apparent 
 antagonism. The Chapter on Electrolysis in his 
 Electricity and Magnetism forms very instructive 
 reading. "Of all electrical phenomena," he writes, 
 " electrolysis appears the most likely to furnish us 
 with a real insight into the true nature of the electrical 
 current, because we find currents of ordinary matter 
 and currents of electricity forming part of the same 
 phenomenon." After explaining Clausius' theory in 
 detail, Maxwell continues : " But if we go on and 
 
Molecules of Electricity 5 1 
 
 assume that the molecules of the ions within the 
 electrolyte are actually charged with certain definite 
 quantities of electricity, positive and negative, so that 
 the electrolytic current is simply a current of con- 
 vection, we find this tempting hypothesis leads us 
 into very difficult ground." Further on we find the 
 following significant passage: "Suppose, however, we 
 leap over this difficulty by simply asserting the fact of 
 the constant value of molecular charge, and that we 
 call this constant molecular charge, for convenience in 
 description, one molecule of electricity. This phrase 
 gross as it is, and out of harmony with the rest of this 
 treatise, will enable us at least to state clearly what is 
 known about electricity and to appreciate the out- 
 standing difficulties." 
 
 How the " molecule of electricity " finally triumphed, 
 and Maxwell's apostles who for a time persisted in 
 repudiating it, had to bring their views into harmony 
 with it, must now form the subject of our consideration. 
 It is doubtful whether the phenomena of electrolysis 
 alone would have been sufficient to lead to the present 
 electron-theory, because it might have been possible to 
 treat the quantitative relationship between electrolytic 
 decomposition and current as a secondary phenomenon. 
 The £ether current in the crude view adopted by the 
 early followers of Maxwell might pass through a liquid 
 as through a solid conductor, and produce the observed 
 results through the secondary action of electro-chemical 
 effects. I imagine that several physicists were trying 
 to develop some such idea; at any rate I did. (I have 
 some recollection of explaining my views at a meeting 
 of the Physical Society of London, but the paper was 
 
 4—2 
 
52 Lecture II 
 
 never published.) I soon convinced myself that the 
 "molecule of electricity" could not be explained away, 
 and turned my attention to the electric discharge 
 through gases. 
 
 The history of the investigation of the passage of 
 electricity through gaseous media, may serve to point 
 several lessons. One of them is the truth of the 
 statement, which has frequently been made, that the 
 history of scientific discovery is to a great extent the 
 history of scientific instruments and appliances. It is 
 illustrated in our case by the marked advances which 
 have accompanied each improvement in the construction 
 of air pumps. The spectroscopic work of Pliicker owed 
 a good deal to " Geissler," whose pump brought down 
 the pressure to a fraction of a millimetre. Later on, 
 Crookes felt the want of better vacua, which again soon 
 led to notable advances of our knowledge of gas 
 discharges. Finally the requirements of glow lamps 
 and Roentgen tubes made our demands still more 
 severe, and again new discharge phenomena were 
 brought to light. 
 
 The frame of mind with which the academic 
 physicist looked upon investigations of the passage of 
 electricity through gases, might be made the subject 
 of instructive comment. The facts so far as they had 
 been ascertained did not fit in with recognised views : 
 hence they were ignored and students were warned off 
 the subject. There was a feeling that perhaps in a 
 century or so, the question might be attacked, but that 
 in the meantime, it had better be left to be played 
 with by cranks and visionaries. No criticism was more 
 frequent at that time, than that of characterising as 
 
Condtiction through Gases 53 
 
 premature any new idea or fresh line of investigation 
 in this direction ; as if any advance of science has ever 
 been made which was not premature a fortnight before 
 it was made. It is perfectly true, and just as much 
 the effect as the cause of the attitude assumed by 
 academic science, that a vast quantity of time and 
 labour had been spent on the subject without making 
 any material advance. There were a great many 
 pretty experiments which seemed to defy all rational 
 explanations, and were too complicated to teach us 
 anything very definite. If we look back upon these 
 experiments now, they may be used to point the moral 
 that experiments conducted in what is sometimes 
 considered to be the true philosophic spirit, where the 
 investigator without any preconceived theories or 
 notions simply wishes to classify facts, seldom lead 
 to any valuable results. 
 
 Progress began when the subject was attacked 
 with some definite object in view, either some theory 
 however crude which had to be supported or some 
 numerical connexion which had to be investigated. 
 
 The combat between rival theories centred to a 
 great extent round the explanation of the phenomena, 
 which were observed in partly exhausted tubes round 
 the wire, which is connected to the negative pole of 
 the electric battery or other source of electricity. 
 Pliicker had found during his spectroscopic experi- 
 ments that as the gas was gradually exhausted, this 
 wire (the kathode) becomes surrounded by a luminous 
 glow, which expands as exhaustion proceeds (A'" Fig. 2). 
 Hittorf, who had been associated with Pliicker, found 
 subsequently that bodies placed in this glow could cast 
 
54 
 
 Lecture II 
 
 a shadow ; and Goldstein discovered further important 
 facts, more especially with respect to the effects observed 
 when two kathodes are placed side by side. 
 
 N 
 
 N 
 
 -A 
 
 Fig. 2. Fig. 3. Fig. 4. 
 
 When the vacuum is sufficiently perfect, it is found 
 that the glow separates off from the kathode, and leaves 
 a dark space between the electrode and the luminosity, 
 as shewn in Fig. 3. As exhaustion proceeds this dark 
 space increases in width and ultimately extends to the 
 boundary of the vessel. The transparent walls of the 
 vessel, where they cut into the dark space (between 
 N and F, Fig. 4), then become vividly fluorescent. It 
 is when the tube is in this state, that a body placed 
 in the dark space (e.g. at A Fig. 4) casts a shadow on 
 the fluorescent portion of the wall. 
 
Kathode Rays 55 
 
 Different investigators adopted different interpreta- 
 tions of the phenomena, which have just been described. 
 The shadow effect suggested radiation as the cause 
 of the glow, but radiation may be oscillatory or cor- 
 puscular in character. The natural inclination of those 
 who remembered how a similar antagonism between 
 two theories had, in the case of light, been decided 
 in favour of oscillations, was to discard corpuscular 
 theories as belonging to the middle ages. Hence 
 many attempts at artificial explanations were made, 
 while the simplest one was set aside. Goldstein's 
 experiments, in which a second kathode placed parallel 
 to the first was observed to repel the radiation coming 
 from the first, seemed absolutely conclusive in favour 
 of some theory of projected particles, but Goldstein 
 himself took a different view, in spite of the fact that 
 Helmholtz, in whose laboratory these experiments were 
 made, urged him — as he subsequently assured me — to 
 adopt the corpuscular hypothesis. Goldstein's experi- 
 ments were begun about the year 1874, and soon 
 afterwards Sir William Crookes, who had notably 
 improved the means of obtaining high vacua in con- 
 nexion with his radiometer experiments, began his 
 work on kathode rays. His beautiful experiments 
 soon attracted attention : they are too well known to 
 need description here as we are mainly concerned with 
 theoretical aspects. Crookes adopted the corpuscular 
 view — a view which probably was first put forward by 
 Varley in 1871 — and by means of accumulated evidence 
 of a most varied nature seemed to many of us to prove 
 his case. Nevertheless a good deal of apathy was 
 shewn, even in this country, with regard to the 
 
56 Lecture II 
 
 theoretical significance of the experiment, while in 
 Germany the opposition to the corpuscular view was 
 almost universal. The cause I believe to be this : 
 although Maxwell's electrodynamic theory had not 
 been generally accepted, the view that a current of 
 electricity was only a flow of sether, appealed generally 
 to the scientific world and was held almost universally. 
 The most absurd consequences were sometimes drawn 
 from this view, and Edlund's assertion that a perfect 
 vacuum was a perfect conductor found many adherents 
 even in England. Edlund's idea was, that a flow of 
 aether could take place without resistance, where there 
 is no matter, and that the high resistance of the most 
 perfect vacua we can produce, is due to a resistance 
 at the surface of the electrodes ; but this supposition 
 ignores the fact that no electrical force can be trans- 
 mitted through a perfect conductor however great the 
 resistance at the surface may be. Adopting the view 
 that a current of electricity simply means a flow of 
 sether, it was tempting to attribute the effect observed 
 under reduced pressures to secondary effects, accom- 
 panying longitudinal or other vibrations set up by the 
 discharge. Attempts were made altogether to dis- 
 connect the luminous effects observed, from the 
 primary effects of the discharge. Notably Heinrich 
 Hertz conducted experiments which were intended to 
 prove that kathode rays produced no magnetic effects, 
 and therefore could not form part of the main process 
 of conduction ; but these experiments, as I pointed 
 out at the time, did not support the interpretation 
 which Hertz gave them. 
 
 Before entering further into the discussion of the 
 
Conduction through Gases 57 
 
 nature of kathode rays, some other researches deahng 
 with the discharge of electricity through partially ex- 
 hausted vessels must be mentioned. A high place in 
 the historical development of the subject is deservedly 
 given to Hittorf, who first introduced accurate methods 
 of measurement, more especially in the investigation of 
 the relationship between current and electro-motive 
 force. In order to obtain steady currents it was 
 necessary to discard the induction coil, by means of 
 which experiments had till then generally been con- 
 ducted, and to replace it by primary cells, which were 
 however, then also open to the suspicion of giving only 
 intermittent discharges through gases. Gassiot was 
 the first to put together a battery of primary cells 
 having a sufficiently great electro-motive force (first 
 3620 zinc-water-copper elements, later 400 Grove 
 cells) to send a current through a gas, but when he 
 examined the Image of the discharge in a rotating 
 mirror he found It to be discontinuous ; hence he con- 
 cluded that the current was intermittent. The same 
 result was obtained by Warren de la Rue, who with 
 the help of a telephone satisfied himself that the 
 current obtained from his chloride of silver battery 
 was always discontinuous. Hittorf, however, after 
 a careful investigation, was able to formulate the 
 conditions under which an electric current can be 
 obtained in a gas, which, so far as all available tests 
 shew, is perfectly steady. In i874\ he experimented 
 with 400 bichromate of potash cells, and In i879-' 
 he described the construction of 1 200 additional cells, 
 
 ^ Poggendorff, Jubelband (1874). 
 ^ Wied. Ann. vn. (1879). 
 
58 Lecture II 
 
 giving him together with the previous set a total 
 electro-motive force of over 3000 volts. It would 
 lead me too far to describe the instrumental methods 
 which were introduced by Hittorf and generally adopted 
 afterwards; but from the point of view of these lectures 
 the following result is important. Hittorf shewed that, 
 in the positive column of the discharge (/', B Fig. 2) 
 so long as the pressure remains constant, the electric 
 force driving the current along the tube is independent 
 of the current. This result seemed quite paradoxical 
 to those who regarded the current as a flow of an 
 incompressible fluid, for when such an incompressible 
 fluid flows through a tube, the difference of pressure at 
 the ends, which corresponds to the electric force must 
 necessarily increase, when the flow is increased. Never- 
 theless it will be seen presently, how simple the ex- 
 planation of Hittorf 's experiment becomes, according 
 to our present view of the nature of electricity. Hittorf 
 also obtained important results, by measuring the total 
 electric force at the kathode. If the kathode is suf- 
 ficiently large, the glow does not generally cover it, and 
 he found that if the current be increased, while the glow 
 expands and covers an additional area of the kathode 
 the electric force does not alter. Only when the 
 current has increased sufficiently to allow the glow to 
 cover the whole of the kathode, is a further increase 
 of the current accompanied by an increase of electric 
 force. These researches were published in 1879. 
 
 The preceding account describes, I hope fairly and 
 sufficiently, the state of the question, so far as I was 
 acquainted with it, when I began experimenting with 
 a view to obtain an insight into the mechanism of gas- 
 
Conduction through Gases 59 
 
 discharges. I have already referred to the position 
 taken up by Varley, Crookes and others, who contended 
 that the kathode ray consisted of electrically charged 
 projected particles, of which the velocity was due to 
 electrostatic repulsion from the kathode, but the 
 question how the particles of gas became charged had 
 not been definitely raised. The most natural idea, 
 and I believed the one generally held, was that the 
 molecules of the gas took the charge by contact with 
 the electrode. 
 
 While trying to bring Faraday's laws of electrolysis 
 into harmony with Maxwell's theory, I became con- 
 vinced that the most hopeful direction of research 
 consisted in starting from the hypothesis that electricity 
 was always concentrated in definite quantities, charging 
 with definite amounts the carriers which conveyed the 
 current. This was a logical and as events proved a 
 correct step. In the case of electrolysis, it was further 
 necessary to stipulate that the charges should never 
 become detached from the atom, except at the surface 
 of the electrode. It seemed natural and even obligatory 
 to adhere to this restriction. The separate existence 
 of a detached atom of electricity never occurred to me 
 as possible, and if it had, and I had openly expressed 
 such heterodox opinions, I should hardly have been 
 considered a serious physicist, for the limits to allow- 
 able heterodoxy in science are soon reached. Looking 
 back now with the advantage of the wisdom which comes 
 after the event, I still think that in the first effort to 
 explain gas discharges, one was forced to adopt the 
 recognised explanation of electrolytic conduction, with 
 as small a number of additional hypotheses as possible. 
 
6o Lecture II 
 
 Hence it was necessary to retain the view that elec- 
 tricity is always attached to ponderable matter, so long 
 as the facts could be brought into harmony with this 
 view. 
 
 One fundamental distinction between the passage 
 of electricity through electrolytes, and that through 
 gases lies in the observation that an electro-motive force 
 however small produces a current in a liquid, while in 
 the case of gases a finite and sometimes a considerable 
 electro-motive force is required to cause a discharge. 
 But it must be remembered, that the behaviour of gases 
 is really the more easy to understand, and that to 
 explain the behaviour of liquids an additional and 
 rather artificial hypothesis, — that of dissociation — had 
 to be invented. The possibility of passing a current 
 through a liquid by means of the smallest electric forces 
 was originally a stumbling block in the theory of 
 electrolytic conduction, but when the same theory was 
 applied to gases this had been forgotten, and the 
 more natural view was now objected to. A more 
 real difficulty in the electrolytic view of gas dis- 
 charges, lay in finding a reason for the opposite 
 charges which it was necessary to ascribe to the two 
 atoms forming the molecule of the ordinary gases ; 
 one of these had to be looked upon as being 
 charged positively, while the other was negatively 
 electrified. It is true that it was not necessary to 
 assume that the polarity of the molecule was per- 
 manent, as it might only be established by the electric 
 field, tending to drive the negative electricity in one 
 direction and the positive in the other. Assuming the 
 polarity to be established it became natural to suppose 
 
Electron Theory 6 1 
 
 that with a certain critical value of the electric force, 
 the molecule broke in two, and that the decomposition 
 of a certain number of molecules then allowed the 
 passage of electricity to take place by a process of 
 diffusion. The view that electric conduction in gases 
 is due to a diffusion of ions, similar to that which 
 takes place in electrolytes, was first definitely proposed 
 by Giese\ in his investigation of the conductivity of 
 flames. His work had escaped my attention at the 
 time, but except in so far as the fundamental points of 
 the theory are concerned, our work did not overlap, 
 as I was mainly concerned with the discharge in 
 partially exhausted vessels. In order to bring the 
 phenomena which take place in the neighbourhood 
 of the kathode within range of the theory, it was 
 necessary to assume that the decomposition of the 
 molecule took place near the kathode, whence the 
 negatively charged particle was driven away with 
 violence. The projected particle then formed the 
 kathode ray. Such were the essential features of 
 the theory I proposed in a paper which was presented 
 to the Royal Society in 1884, and which, at the sug- 
 gestion of Sir George Stokes, was chosen as the 
 Bakerian lecture for that year. The view I had 
 formed of the kathode glow seemed supported by 
 experiments which appeared to shew that the most 
 prominent features of that glow were absent in 
 mercury vapour which is monatomic. Later, I con- 
 vinced myself that an abnormally large dark space 
 could form in mercury, and that its absence in my 
 
 ' Wied. Ann. xvii. (1882). 
 
62 Lecture II 
 
 first experiments was due to the narrowness of the 
 tubes which were used. 
 
 I reahsed at an early stage that in order to de- 
 monstrate the correctness of the theory of ionic charges 
 it was necessary to find a proof that the charge is a 
 definite quantity, and that a crucial experiment could be 
 devised by observing the magnetic deflexion of kathode 
 rays. I pointed out that the experiments then available 
 tended to support the view that the charge is indeed 
 constant. As regards the positive discharge I shewed 
 that Hittorf's laws were easily accounted for, by sup- 
 posing that an increase of current may be produced by 
 an increase in the number of particles which take part 
 in the discharge, without change in their velocity. As 
 it is the velocity which is determined by the electric 
 force, an increase of the latter does not necessarily 
 accompany an increase of current intensity. 
 
 Pursuing the subject I published a further paper 
 in the year 1887, in which it was shewn, that a gas can 
 be converted into a conductor by an independent 
 discharge which is made to pass through it. Hittorf 
 had already found that a column of gas through which 
 a current passes, responds to small electromotive forces 
 introduced at right angles, but his experiment left it 
 undecided whether this transverse conductivity is due 
 to peculiar conditions either of temperature or lumi- 
 nosity accompanying the primary discharge. In my 
 own experiments, the conductivity was found to exist 
 some distance away from the primary discharge. 
 The experiments were explained by the breaking 
 up of the neutral molecules in the primary discharge, 
 the charged atoms acting as we would now say as 
 
Ions and Ionization 63 
 
 ions capable of independent diffusion and therefore 
 converting the whole mass of gas into a conductor 
 of electricity. Similar experiments leading to the 
 same conclusion were made by Arrhenius, and 
 communicated by him to the Swedish Academy of 
 Science on September 14th, 1887. 
 
 As the word " ion " is so frequently used in our 
 discussion, it is well to make its meaning clear. Every 
 carrier of electricity of not more than molecular dimen- 
 sions is an ion, whether the carrier be an atom of 
 electricity, or a charged molecule, or a portion of a 
 molecule carrying a charge. The charge is measured 
 by its external effects so that a molecule containing an 
 equal number of positive and negative charges would, 
 for our present purpose, be counted as uncharged. In 
 the absence of any ionizing agent, the molecules of 
 ordinary gases would be neutral, so far as we know. 
 When through an external radiation or through any 
 other process, the neutral molecule loses a charge, or 
 splits up into charged portions, we say that the gas 
 has become "ionized." 
 
 This ionic theory of gas discharges, while ignored 
 in England, made good progress abroad ; Arrhenius 
 adopted it as well as Elster and Geitel, who were then 
 investigating the electric behaviour of gases in the 
 neighbourhood of incandescent electrodes. Warburg, 
 whose important researches on the constancy of 
 electric force at the kathode were made during this 
 period, also expressed himself in general agreement 
 with the views which I had expressed. In the 
 meantime I was pursuing the subject experimentally, 
 and was more especially trying to obtain evidence of 
 
64 Lecture II 
 
 the constancy of the charge carried by the ion, when 
 I received a request once more to deliver the Bakerian 
 lecture in 1890. It was intended that this lecture 
 should give a general account of our knowledge on 
 the conduction of electricity through gases. So far 
 as my own experiments were concerned, the request 
 came rather prematurely, as they were only in a 
 preliminary state. My primary object in writing out 
 the lecture, was to shew that the principal facts 
 observed could be explained by the hypothesis that 
 gases may be converted into conductors by ionization 
 and that the charge of the ion is a fixed quantity. 
 One of the great difficulties I found in discussing the 
 subject with my colleagues was to convince them of 
 the possibility of having a volume electrification, 
 meaning an electrification spread through a volume 
 instead of being confined to the surface of a conductor. 
 This was generally denied in a dogmatic fashion on 
 the ground of the incompressibility of the aether. In 
 my lecture, I tried to shew that volume electrification 
 must exist whenever a current passes through an 
 electrolyte, which is not homogeneous in composition, 
 and that even in solid conductors, which are not 
 rigidly at the same temperature everywhere, volume 
 electrification can only be avoided by introducing 
 something analogous to hydrostatic pressure for which 
 there is no experimental evidence. At any rate the 
 possibility of volume electrification is easily recognised, 
 if we adopt the hypothesis of an atomic charge, and 
 a dogmatic denial could not dispose of that possibility. 
 At the head of my paper, I was able to place 
 a sentence of far reaching importance taken from 
 
Atomic Charge 65 
 
 Helmholtz's Faraday lecture : " If we accept the 
 hypothesis that the elementary substances are composed 
 of atoms, we cannot avoid concluding that electricity 
 also, positive as well as negative, is divided into 
 definite elementary portions, which behave like atoms 
 of electricity." As a personal recollection I may add 
 that at that period, I occasionally met Helmholtz 
 during the summer holiday which he used to spend 
 at Pontresina, and he frequently enquired after the 
 progress of my experiments. I consistently received 
 helpful encouragement from him, as I did from no 
 one else, more especially in the prosecution of the 
 investigation of the magnetic deflexion of kathode 
 rays, which he quite realised would yield the key of 
 the position. ^ 
 
 If we assume that a particle having a mass in, 
 carries a charge e and moves with a velocity v, a 
 magnetic field at right angles to the direction of 
 motion will deflect the particle which then will no 
 longer move in a straight line but in a curve having 
 a radius r. If the intensity of the magnetic field be 
 M, the following relation must hold : — 
 
 e V 
 ni Mr' 
 
 If we assume that the velocity is that due to the 
 electric force V we can calculate v in terms of V and 
 in that case 
 
 m~ M-r""' 
 
 The quantities on the right-hand side are all 
 measurable, hence we may determine the value ejm 
 
 s. p 
 
66 Lecture II 
 
 and ascertain whether it is the same as that found by- 
 electrolysis. For nitrogen atoms we should expect 
 the value to be about 2000, while the number actually 
 obtained by the experiment was about three and a half 
 millions'. If this number be accepted as correct, the 
 conclusion I ought to have drawn from it is that either 
 the quantity of electricity carried in the gas discharge 
 is very much greater than that conveyed by the ions 
 in electrolysis or that the mass of the carrier is very 
 much smaller. Though perhaps I ought to have 
 had more faith in my own experiment, there was 
 certainly the possibility of a large error introduced 
 by the assumption that the velocity of the particle had 
 the full value which it would reach under the effect of 
 the electric forces only. This meant neglecting the 
 retardation due to mutual impacts which in my case 
 (the pressure being about "3 mm. of mercury) might 
 have been considerable. Viscosity would ultimately 
 reduce the velocity of the particle to that appertaining 
 to ordinary gaseous translation, and if this ultimate 
 velocity were substituted for v, the calculated value of 
 elm comes very near that which is found in electrolytes. 
 Even now I am not convinced that the conclusion 
 I drew from my experiment was not perfectly correct. 
 I worked at pressures which were considerably higher 
 than any at which the later experiments were conducted, 
 and it seems to me to be very likely that in the rays 
 which I investigated the electron had already attached 
 
 ' In the published account of the Bakerian lecture the value 
 I -I X 10" is given; but an error in the adopted value of the magnetic 
 field was subsequently discovered. The correction was made and 
 further details about the experiments given in Wied. lxv. (1898). 
 
Magnetic Deflexion of Kathode Rays 67 
 
 itself to the molecule of matter. But even taking the 
 experiment with all its imperfections, it was sufficient 
 to prove that the charge carried was much greater 
 than any that could have been received by the 
 molecule through contact with the electrode — and this 
 was pointed out. The above experiment did not 
 attract the attention of physicists ; at any rate four 
 years later Sir Joseph J. Thomson in a paper published 
 in the Philosophical Magazine wrote : — " I cannot find 
 any quantitative experiment on the deflexion of these 
 (kathode) rays by a magnet." The same paper contains 
 the description of an attempt to measure the velocity 
 of kathode rays, the result being found consistent with 
 the supposition that ejm has the electrolytic value for 
 hydrogen. 
 
 The next important advance was made by Sir Joseph 
 J. Thomson, in a lecture delivered before the Royal 
 Institution, and printed in full in the Philosopliical 
 Magazine, October, 1897, the essential portions having 
 already appeared in the Electrician for May 21st, 
 1897. Perrin had shewn by direct experiment that 
 the kathode ray carries a negative charge, but some 
 objections which might have been raised against the 
 form of the experiment were now removed. The ratio 
 of the charge to the mass was determined by com- 
 bining the magnetic deflexion of kathode rays with 
 their total energy, measured by the effect they pro- 
 duced in raising the temperature of a thermo-junction. 
 This ratio (^/w) was found to range between 3x10' 
 and 10'. The same ratio was determined by combining 
 the magnetic deflexion with the deflexion in an electric 
 field, acting at right angles to the rays. The value 
 
 5—2 
 
68 ._ Lecture II 
 
 obtained by this method was 77 x 10'. The new in- 
 terpretation given to these experiments was summed 
 up as follows : " Thus on this view we have in the 
 kathode rays matter in a new state, a state in which 
 the sub-division of matter is carried very much further 
 than in the ordinary gaseous state : a state in which 
 all matter — that is, matter derived from different 
 sources such as hydrogen, oxygen, &c. — is of one and 
 the same kind ; this matter being the substance from 
 which all chemical elements are built up." On May 21 
 of the same year Kaufmann communicated a paper 
 to Wiedemann s Annalen in which it was shewn that 
 the magnetic deflexion of the kathode ray is, as 
 required by the theory of projection, proportional to 
 the square root of the potential fall. The value found 
 for elm was 10', but the author expresses a difficulty 
 in understanding that this number is apparently in- 
 dependent of the nature of the gas and of the metal 
 used as material of the kathode. In the following 
 year Kaufmann in conjunction with Aschkinass shewed 
 that the electrostatic repulsion effects of Goldstein are 
 consistent with the corpuscular theory, and Kaufmann 
 now appears fully to accept that theory, repeating his 
 former measurements and obtaining the more accurate 
 value of I "86 x 10'. 
 
 In the interest of historical accuracy a lecture 
 delivered by Professor E. Wiechert to the " Mathe- 
 matisch-physikallsches Institut" of the University of 
 Konigsberg on January 7th, 1897, deserves to be men- 
 tioned. He describes experiments on the magnetic 
 deflexion of kathode rays which did not go beyond the 
 previous work of other observers, but he recognises 
 
Atomic Charge 69 
 
 for the first time the smallness of the carrier ; and 
 though perhaps on insufficient grounds, identifies the 
 carrier with the electron, and assigns a mass to it which 
 is between the four hundredth and four thousandth 
 part of the mass of the hydrogen atom. Two years 
 subsequently', an important research was published by 
 the same author. In it, he shewed how an experi- 
 mental method, first attempted by Descoudres, could 
 be made to give practical results in determining the 
 velocity of the particles in the kathode ray. The 
 method is of great originality, and the execution was 
 difficult and well carried out. Once this velocity is 
 determined, the ratio ejni can be measured, but though 
 arriving at a valuable confirmation of previous results, 
 Wiechert did not effect an improvement on the accuracy 
 of the numerical constant to be measured. 
 
 In the experiments, which so far have been men- 
 tioned the ratio of the charge to the mass was measured, 
 but the smallness of the mass could only be inferred if 
 it be assumed that the charge of the carrier is the 
 same as that we calculated from experiments with 
 liquid electrolytes. It was therefore an important 
 advance when Sir Joseph J. Thomson, in 1898, de- 
 scribed experiments, which allowed the charge to be 
 measured directly. He made use of a discovery of 
 C. T. R. Wilson, that the negative ions can act as 
 nuclei in the formation of water drops. If the air be 
 ionized by Roentgen rays and a cloud be formed by a 
 sudden chilling of the air, the small drops which make 
 up the cloud may be observed to descend slowly, the 
 rate of descent indicating their size. If the quantity 
 ' Wied. Ann. Lxix. (1899). 
 
70 ■ LecHire II . 
 
 of water vapour, which is condensed is known, the 
 number of drops, and therefore the number of nuclei, 
 is determined. On the other hand, if the ions formed 
 are set in motion by electric forces the current may be 
 measured, and this current must be equal to the product 
 of the charge, the number of ions, and the velocity 
 they require under the action of the given electric 
 force. The last of these quantities may be determined 
 independently, and therefore making use of the number 
 of nuclei found by the experiment in the falling cloud, 
 the amount of the charge can be calculated. In the 
 current electro-magnetic system of measurement the 
 charge found by Thomson was 7'3 x lo"'". Certain 
 sources of error were not sufficiently taken account 
 of in these earlier experiments, and H. A. Wilson, 
 improving on the method, obtained a value rather less 
 than half of that obtained by Thomson'. In order 
 to compare the numerical value found with the charge 
 of an ion in an electrolyte, it is necessary to introduce 
 a quantity which is not accurately known, for electrolysis 
 can only give us the value of the product of the charge 
 into the number of molecules, which a cubic centimetre 
 of hydrogen contains. If an estimate of this number be 
 made by other means a substantial agreement is arrived 
 at. No one could doubt, after the experiments which 
 have been described, that the electric charges we are 
 concerned with in the conduction through gases and 
 through liquids are identical. Even then, however, 
 the general body of physicists remained indifferent to 
 the fundamental importance of the results arrived at 
 
 ' The experiments of Rutherford and Geiger by a different and 
 more accurate method give 4'65 x iq-". 
 
J. J. Thomsons Measurements 71 
 
 until the meeting of the British Association at Dover 
 in 1899, when Sir J. J. Thomson described some 
 further experiments, and explained more fully the con- 
 clusion to which they inevitably led. The experiments 
 themselves only confirmed the previous results ; but 
 while in the previous investigations the ratio e\ni was 
 measured in the kathode ray, and the charge e was 
 measured on the ion formed by Roentgen rays, both 
 quantities were now measured for the same ion, formed 
 at the surface of a metal, when ultra violet radiation falls 
 upon it. The experiments proved, beyond any possi- 
 bility of doubt, the smallness of the mass of the electric 
 carrier, which is set free by ultra violet radiation. The 
 lecture in which the above experiments were described, 
 was delivered to the British Association on the occa- 
 sion of the visit of members of the French Association, 
 which met concurrently at Boulogne\ It at once 
 carried conviction, and though to those who had followed 
 the gradual development of the subject, it only rendered 
 more certain what previous experiments had already 
 plainly indicated, the scientific world seemed suddenly 
 to awake to the fact that their fundamental concep- 
 tions had been revolutionised. A new era of science 
 begins at this point, and I cannot do better than 
 conclude my lecture with the passage in which Sir 
 Joseph J. Thomson summarised his results : 
 
 " I regard the atom as containing a large number 
 of smaller bodies which I will call corpuscles ; these 
 
 ' The title which originally was " On the existence of masses 
 smaller than the atoms," but was changed when the paper was 
 published in the Philosophical Magazine, XLViii. p. 565 (1899) to 
 " On the masses of ions in gases at low pressure." 
 
72 Lecture II 
 
 corpuscles are equal to each other ; the mass of a 
 corpuscle is the mass of the negative ion in a gas 
 at low pressure, i.e. about 3 x 10""* of a gramme. In 
 the normal atom, this assemblage of corpuscles forms 
 a system which is electrically neutral. Though the 
 individual corpuscles behave like negative ions, yet 
 when they are assembled in a neutral atom the 
 negative effect is balanced by something which causes 
 the space through which the corpuscles are spread 
 to act as if it had a charge of positive electricity equal 
 in amount to the sum of the negative charges on the 
 corpuscles. Electrification of a gas I regard as due to 
 the splitting up of some of the atoms of the gas, 
 resulting in the detachment of a corpuscle from some 
 of the atoms. The detached corpuscles behave like 
 negative ions, each carrying a constant negative charge, 
 which we shall call for brevity the unit charge ; while 
 the part of the atom left behind behaves like a positive 
 ion with the unit positive charge and a mass large 
 compared with that of the negative ion. On this 
 view, electrification essentially involves the splitting 
 up of the atom, a part of the mass of the atom getting 
 free and becoming detached from the original atom." 
 
LECTURE III 
 
 During the first few days of January 1896, I was 
 returning to Manchester after a short Christmas hoHday, 
 and on my way home called at the Laboratory for my 
 letters. I opened a flat envelope containing photo- 
 graphs, which, without accompanying explanation, were 
 unintelligible. Among them was one shewing the out- 
 line of a hand, with its bones clearly marked inside. 
 I looked for a letter which might give the name of the 
 sender and explain the photographs. There was none, 
 but inside an insignificant wrapper I found a thin pamph- 
 let entitled: — "Uber eine neue Art von Strahlen" — (on 
 a new kind of rays) by W. C. Roentgen. This was 
 the first authentic news that reached England of a 
 discovery made at the end of the year 1895, which 
 both directly and indirectly gave a tremendous impulse 
 to experimental science. Before I left the room, and 
 although my family was waiting for me in the cold 
 outside, I had read and re-read Roentgen's account, 
 which concisely but convincingly described the experi- 
 ments, by which he had with remarkable ability, inves- 
 tigated and determined the main properties of the new 
 radiation. Briefly they are as follows : — the radiation 
 sets out from the place where the kathode ray strikes 
 an obstacle. In the original experiment this was the 
 wall of the glass tube, but as subsequently found, a 
 
74 Lecture III 
 
 metal target acts equally well or better. The rays are 
 more penetrating than any that were known at the time 
 of the discovery, but are absorbed more or less by all 
 substances. The radiation acts on a photographic plate 
 or on a phosphorescent screen, which becomes luminous 
 under its influence. If a leather purse containing coins 
 be held up between the source and the screen, the 
 shadow of the purse is only faintly indicated while the 
 shadows of the coins are very dark. Similarly, if a hand 
 be placed in the path of the rays, the bones are clearly 
 marked in the shadow, shewing that they absorb the 
 radiation more strongly than the fleshy parts. 
 
 The discovery of the rays was partly accidental, 
 which means that it was the consequence of one of 
 those hints, which sometimes an investigator receives 
 unexpectedly, and which according as they are taken, 
 or neglected, make or mar a reputation. In this case a 
 batch of photographic films lying near a discharge tube, 
 which was set up in order to repeat Lenard's experi- 
 ment on the transmission of kathode rays through 
 aluminium leaves, was found to be fogged. It might 
 have been merely a bad set of films, but the hint was 
 taken, and within a week the principal properties of 
 the new radiation were disclosed. 
 
 Roentgen proved that, unlike the rays of light, the 
 new rays are not refracted when they pass obliquely 
 from one medium to another, nor could he detect any 
 interference effects : hence they seemed to be different 
 in kind from rays of light. 
 
 The interest which the discovery roused in the 
 scientific world and the sensation it created generally 
 may be imagined, and there were few laboratories in 
 
Roentgens Discovery 75 
 
 which attempts were not immediately made to repeat 
 the experiment. This was not altogether easy, because 
 few institutions were then provided with the apphances 
 necessary to obtain so perfect a vacuum as that required 
 for the purpose, and also because the English lead 
 glass is much less suitable than the soft German glass 
 to excite and transmit the rays. Almost at once the 
 possibility of practical applications attracted the public, 
 and more especially the medical profession. In the 
 diagnosis of complicated fractures of bones, or of the 
 location of extraneous bodies, it was clear that a 
 method of great utility was now available. To me 
 this had an unfortunate consequence. My laboratory 
 was inundated by medical men bringing patients, who 
 were suspected of having needles in various parts of 
 their bodies, and during one week I had to give the best 
 part of three mornings to locating a needle in the foot 
 of a ballet dancer, whose ailment had been diagnosed 
 as bone disease. The discharge tubes had all to be 
 prepared in the laboratory itself, and where a few 
 seconds exposure is required now, half an hour had to 
 be sacrificed owing to our ignorance of the best con- 
 ditions for producing the rays. More difficult problems 
 also arose, as when I had to travel to a small manu- 
 facturing town in the north of Lancashire, in order to 
 locate a bullet in the skull of a poor dying woman who 
 had been shot by her husband. My private assistant 
 completely broke down under the strain and excite- 
 ment of all this work, and the experiments on the 
 magnetic deflexion of kathode rays on which I was 
 then engaged were seriously interfered with by this 
 interruption ; but I must return to my subject. 
 
76 Lecture III 
 
 The absence of refraction and of interference had 
 led Roentgen to reject the idea that the radiation was, 
 Hke light, due to transverse vibrations of the aether. 
 He therefore tentatively suggested that they might be 
 longitudinal vibrations. He was possibly led to this 
 suggestion by the belief that the kathode ray, according 
 to the then fashionable theory, was connected with a 
 wave-motion. Now in the theory of light there has 
 always been this difficulty, that if the aether were an 
 elastic body in the ordinary sense, a transverse wave 
 entering from one medium to another ought to be 
 partly converted into a longitudinal wave. These longi- 
 tudinal waves had never been traced, and Roentgen 
 may have thought that the new discovery might solve 
 a difficulty in the older theory of light. Be this as it 
 may, the suggestion never met with much favour and 
 there is not much to be said for it. As regards the 
 absence of interference and of refraction, I explained 
 in a letter to Nature, which appeared a few weeks 
 after the announcement of the discovery, why it is not 
 necessarily antagonistic to the view, that the Roentgen 
 rays are transverse vibrations of very short period. 
 The reason I gave was, that interference depended 
 on the regularity of vibration, which requires the pre- 
 dominance of waves of certain lengths over others, and 
 in such an impulsive wave as is likely to be generated 
 by the impact of a kathode ray, I pointed out that the 
 necessary regularity would probably not exist. I also 
 gave reasons, why the absence of refraction could be 
 explained by a theory of refraction, which was then 
 already in favour, the slower wave-velocities in trans- 
 parent media being connected with the resonance of 
 
Roentgen Rays 77 
 
 molecular vibration, responding to the periodic force of 
 the passing wave. 
 
 Experiments on the absorption of the Roentgen 
 radiation by different bodies give interesting results, of 
 which at present only one need be mentioned. The 
 rays emitted by different tubes, notably when they are 
 in different stages of exhaustion, differ considerably 
 in their penetrating power ; hence we must conclude 
 that the Roentgen rays are not "homogeneous" and 
 differ from each other, much as the rays emitted by 
 ordinary sources of light are found to differ. We must 
 defer the discussion of the nature of the Roentgen ray 
 until we can connect it with other radiations, but 
 it is necessary to keep in mind from the beginning 
 one marked distinction between a kathode ray and a 
 Roentgen ray ; the former is deflected by magnetic 
 force while the latter is not. This distinction was 
 insufficiently realised at first by some, who even went 
 so far as to deny the novelty of Roentgen's discovery. 
 Hertz and Lenard had already shewn — it was said — that 
 kathode rays could traverse metallic bodies, and there 
 was no reason why the effects observed by Roentgen 
 might not have been produced by some of the kathode 
 rays traversing the tube in which they were generated. 
 1 remember in this connexion an interesting discussion 
 on board an excursion steamer on the occasion of the 
 Kelvin Jubilee. A German Professor stood up for 
 Lenard's priority in the discovery of the new radia- 
 tion. The whole thing, he said, was in Lenard's 
 mind, when he carried out his researches. "Ah," 
 said Sir George Stokes with a characteristic smile, 
 " Lenard may have had Roentgen rays in his own 
 
7 8 Lecture III 
 
 brain, but Roentgen got them into other people's 
 bones." Nobody enjoyed the remark more, or 
 subsequently repeated it oftener, than the eminent 
 Professor who had called it forth. 
 
 It obviously suggested itself to every one acquainted 
 with the subject, that the new radiation might prove 
 an effective ionizer of a gas, and it is therefore not 
 surprising that several observers discovered inde- 
 pendently that this was indeed the case. This ionizing 
 property gave us for the first time, a method for 
 studying the properties of ionized gases in detail and 
 with accuracy. The previous methods did not easily 
 adapt themselves, partly because, as when ultra-violet 
 radiation is used, the ionization only takes place at the 
 illuminated surface and not throughout the mass of the 
 gas, and partly because the ionizers shewed too many 
 complicated irregularities. I had tried, for instance, 
 with small success, to study the law of conduction of 
 ionized gases in a vacuum, some distance away from 
 the kathode, finding it difficult to obtain consistent 
 results. 
 
 Very soon after Roentgen's discovery was an- 
 nounced, the matter was taken up in the Cavendish 
 Laboratory, by Sir J. J. Thomson, Rutherford and 
 others, and the main laws which regulated the effects 
 were soon brought to light. 
 
 To get a clear idea of their meaning, let us begin 
 by asking, what is an electric current ? We need not 
 go back to first principles and discuss whether it is 
 a transference of aether or not, but may rest satisfied 
 with what we can directly observe in every case in 
 which a current passes. An electric current means 
 
Convection of Ions 79 
 
 the transference of electricity just as a current of water 
 means the transference of water. If now we adopt 
 the view that electricity is concentrated at some points 
 in definite quantities, as has been explained in my 
 second lecture, the motion of these points must consti- 
 tute a current of electricity. These centres of electric 
 force possess what in ordinary language we call mass 
 and can act as "ions." The meaning of the word ion 
 has been explained in the previous lecture, and we 
 have been long familiar with the manner, in which an 
 electric current is transmitted through an electrolyte by 
 the convection of these charged carriers. If the positive 
 ions move in one direction, there will be an electric 
 current in that direction, and similarly a motion of the 
 negative ions constitutes an electric current, but it is 
 important to remember that as regards all external 
 effects the motion of a negative ion to the left is 
 identical with that of a positive ion to the right. 
 Hence, if negative ions move — say — towards the west, 
 this constitutes an electric current to the east. It 
 appears therefore that the measure of the intensity 
 of an electric current which depends merely on the 
 quantity of electricity transmitted, does not completely 
 specify the process of conduction ; because it leaves it 
 undecided, whether positive ions move in one direction 
 the negative ones being stationary, or negative ions 
 move in the opposite direction, the positive ones being 
 stationary, or whether both kinds take part in the 
 current. What has been said will make it clear, 
 I hope, that the current depends on the relative 
 velocities of the negative and positive ions, as they 
 move past each other, and that we have agreed to 
 
8o Lecture III 
 
 take the direction of the positive ions for the direction 
 of the current. 
 
 The positive and negative ions may differ in other 
 respects than that of the charge they carry. As a 
 rule they have different masses. The electron vi^hich 
 is the elementary constituent of electricity has, as 
 we have seen in the last lecture, a very small mass, 
 but positive ions so far as is known at present, have 
 masses of the same magnitude as ordinary molecules 
 of matter, and the negative ions which we find in 
 most cases in ionized gases are not simple electrons 
 but have masses which are not very much smaller 
 than those of the positive carriers. 
 
 Let us now take a gas which is subjected to some 
 ionizing radiation, such as that emanating from a 
 Roentgen tube. The rays traversing the gas are 
 absorbed to some extent, the absorbed energy being 
 partly utilised in producing a definite number of ions 
 in each second of time. Does this ionization then 
 increase indefinitely .'* No, because an ionized gas 
 does not, when left to itself remain ionized. There is 
 a steady recombination of the positive and negative 
 charges, and the gas tends to return to its original 
 neutral condition. I must remind you here of the signi- 
 ficant distinction between a liquid and a gas, which has 
 already been dealt with in the previous lecture. In 
 the liquid electrolyte, the ions are always present ready 
 to follow the smallest electric force that may be applied, 
 while in the case of a gas, ionization has to be set up 
 before an electric current can pass, and no permanent 
 ionization can be maintained without permanent ionizing 
 cause. The rate at which the ions recombine depends 
 
Ionization of Gases 8i 
 
 on the nature of the gas, but there is not much difference 
 in the case of the gases with which we are most 
 famiHar, one recombination taking place at atmospheric 
 pressure in each second of time for each thousand 
 positive and negative ions. The number of recombina- 
 tions is proportional to the number of ions of each 
 kind, so that if the number both of positive and of 
 negative ions is doubled we should get four times as 
 many recombinations in the same time. In the experi- 
 ments on recombination great care must be taken to 
 exclude the presence of dust, as small or liquid particles 
 seem to attract the ions and thus accelerate their 
 recombination. Similarly the presence of moisture 
 helps the recombination. 
 
 Returninof to the consideration of a eas which is 
 subjected to a constant ionizing agent, we see that the 
 ionization must reach a definite limit, when the number 
 of ions produced each second equals that which 
 disappears owing to recombinations. We obtain in 
 this way a gas ionized to a certain definite amount, so 
 long as the ionizing agent remains constant. When 
 the conductivity of such a gas is studied, it is found 
 that under the action of small electric forces, the 
 electric current produced is proportional to the force. 
 The positive ions will move to one side, the negative 
 ones to the other, and the total number which give up 
 their chargre to the two electrodes in unit time eives the 
 
 o o 
 
 current conveyed by the gas. This law of pro- 
 portionality between electro-motive force and current 
 is the law known as " Ohm's law," which holds in 
 solid and liquid conductors for all intensities of current 
 which have hitherto been studied. But in the case 
 s. 6 
 
82 Lecture III 
 
 of gases it ceases to hold when the electro-motive force 
 is increased, as it is found that a limit is soon reached 
 when for a given ionizing agent, the current remains 
 the same whatever electro-motive force we apply. It 
 is easy to understand why this should be. Let us fix 
 our ideas by imagining two equal parallel 
 plates A and B (Fig. 5), connected respec- 
 tively to the two poles of a battery E. If 
 a Roentofen radiation of constant strength 
 be sent through the space separating the 
 plates, the air will be ionized and there 
 will be a definite number of ions (say 1000 
 of each kind) formed in each cubic centi- 
 metre of gas, in each second of time. ^'^- 5- 
 It is clear that in that case, one thousand charges re- 
 present the greatest quantity that can be communicated 
 to each electrode in one second, and the maximum 
 current that can pass through the gas is therefore equal 
 to 2000 units, multiplied by the total volume of gas, 
 measured in cubic centimetres, which is included 
 between the plates. The maximum current will be 
 reached, when the electric force between the plates is 
 sufficiently strong to attract the ions, and remove them 
 as soon as they are formed. The current is then called 
 a "saturation current." 
 
 The discovery of the Roentgen rays stimulated 
 scientific enquiry in many directions, but the most 
 remarkable results were obtained through a process 
 of reasoning ultimately found to be mistaken, which 
 may serve to shew that the ideas which induce us to 
 conduct an experiment are often of small importance 
 compared to the results which, if it be carried out in 
 
H. BecgtierePs Discovery 83 
 
 a philosophic spirit, may be reached through it. Henri 
 Becquerel, who following the tradition of his emi- 
 nent father, had made himself thoroughly familiar with 
 all phenomena of fluorescence and phosphorescence, 
 had conceived the idea, that the emission of Roentgen 
 rays might always accompany the excitation of phos- 
 phorescent light. He was led to this supposition by 
 noting that in the original experiments of Roentgen, 
 the rays were emitted from the spot where the 
 kathode rays struck the glass rendering it phospho- 
 rescent. Having a variety of substances possessing 
 the required property at his command, he tried to test 
 his views, and after some unsuccessful trials placed two 
 crystalline laminae of a uranium salt on a photographic 
 plate, which was wrapped in black paper, so that no 
 ordinary light could fall on it ; between one of the 
 crystals and the photographic plates a piece of silver was 
 interposed. Strong sunlight was then allowed to fall 
 on the uranium salt, and after an exposure of several 
 hours an action on the photographic plate could be 
 detected, the outline of the crystal and the shadow of 
 the piece of silver being distinctly marked. Here 
 then, there seemed to be a complete justification of 
 the view which had suggested the experiment, for a 
 radiation had evidently penetrated the paper and 
 reached the photographic film. It looked as if sun- 
 light excited the uranium salt, just as the kathode ray 
 excites the glass wall of a Roentgen tube. But now 
 happened one of those chance occurrences, which 
 while demolishing Becquerel's theories, placed him 
 at the same time in the forefront of experimental 
 discoverers. A plate was prepared by him in the way 
 
 6—2 
 
84 Lecture III 
 
 described, but as the sun was only shining fitfully at 
 the time, the experiment was discontinued, and after a 
 short exposure the plate put in a drawer ready to be 
 used again on a clear day. The next few days were 
 cloudy, and the plate was developed without expecta- 
 tion of finding more than a feeble trace of the radiation : 
 on the contrary, the shadow was marked with great 
 distinctness. Becquerel was surprised but soon con- 
 vinced himself that the sunlight had nothing to do 
 with his previous results, but that his uranium salts 
 always emitted radiations, which could penetrate paper 
 and acton a photographic plate. It remained however 
 to be seen, whether this was a permanent property 
 of these substances, or only a temporary effect 
 due to their having been previously exposed to 
 some source of light. It might have been that the 
 salt of uranium which had been lying in diffuse day- 
 light for a considerable time, had absorbed a con- 
 siderable amount of energy, which it was able sub- 
 sequently to give out in the form of a penetrating 
 radiation. If this were the case, we should expect the 
 radiation to die out gradually though perhaps only 
 slowly ; but experiment shewed no diminution of the 
 effect with time. Hence Becquerel could claim with 
 justice to have discovered a continuous radiation, 
 involving an emission of energy by substances which 
 till then had been considered to be unalterable and 
 therefore containing a fixed amount of energy. The 
 discovery of such a property naturally created a great 
 sensation, and various observers examined a number 
 of bodies, including all known elements, in order to 
 see whether any of them emitted similar radiations. 
 
Discovery of Radium 85 
 
 All compounds of uranium shewed the effect to the 
 same degree, if amounts were compared which con- 
 tained equal weights of the metal, but thorium was 
 the only other known element which was also found 
 to be "radio-active." Madame Curie, who almost 
 simultaneously with E. Schmidt, discovered this fact, 
 made in conjunction with her husband, the significant 
 observation that while all artificially prepared uranium 
 compounds are equally active, pitchblende, which 
 is a mineral consisting chiefly of uranium oxide, 
 was weight for weight decidedly more powerful. To 
 find the reason why, the various metals, which the 
 mineral contains in small quantities, were separated 
 chemically and examined. It was found that the 
 extracted bismuth and barium were both strongly 
 active, more especially the barium. Now ordinary 
 barium shews no sign of any radio-activity, and the 
 conclusion was therefore inevitable, that pitchblende 
 contained some unknown element, having similar 
 properties to barium, and separating out with that 
 body in the chemical treatment to which it had been 
 subjected. By a lengthy process often repeated, the 
 new element was gradually separated from barium, and 
 ultimately a product was obtained which was about 
 a million times more active than uranium. The name 
 of "radium" was given to the element which was thus 
 obtained, as chloride or bromide almost free from 
 barium. Another radio-active element was found in 
 the precipitate, which separated out with the bismuth, 
 and was called "Polonium." Soon afterwards Debierne 
 obtained a further active substance from pitchblende to 
 which he gave the name of " Actinium." We had then 
 
86 Lecture III 
 
 apparently five elements possessing the new properties: 
 Uranium, Thorium, Polonium, Radium, and Actinium. 
 These elements, with the exception of Polonium give 
 rise to other radio-active bodies, so that the total 
 number known at present is much greater than five. 
 Amongst them "Ionium" recently discovered by 
 Boltwood is of special interest on account of its close 
 relationship to Radium. 
 
 I now interrupt the historical account of these dis- 
 coveries in order to discuss the entirely new aspect of 
 the constitution of matter which they opened out to us. 
 Clearness in perception of the fundamental facts and 
 their theoretical bearinsj is here essential. We must 
 begin by defining our terms, and making clear what is 
 novel in the properties of radium and of its associates. 
 We call a body radio-active if it permanently emits 
 energy, which belongs to it in virtue of its chemical 
 identity. The ordinary radiation of heat does not come 
 within this definition, because though a hot body placed 
 in a colder medium emits energy, it does not do so 
 because it is iron or because it is copper, but because 
 it is hotter than its surroundings. Similarly ordinary 
 phosphorescent bodies are not radio-active, because the 
 energy they emit is transient, and taken from a store 
 which has previously been absorbed. The character- 
 istic feature of the emission of energy from radio-active 
 bodies is, that it is entirely independent of physical 
 conditions. The three different kinds of " rays " which 
 are emitted by these bodies are designated by the 
 first three letters of the Greek alphabet, a, j8, y. The 
 a ray is a projected particle which is positively charged 
 and has a mass of about twice that of the atom of 
 
Magnetic Deflexion of Rays 
 
 87 
 
 hydrogen, the ^ ray, on the other hand, while also 
 corpuscular, has a much smaller mass, and is, in fact, 
 identical with the kathode ray, being negatively charged 
 and having a mass about the thousandth part of that 
 of the hydrogen atom. The y radiation is identical 
 with the Roentgen ray, and according to the generally 
 accepted view, consists of a disturbance of the ather 
 not unlike that of a group of extremely short waves. 
 Our knowledge of the charge and mass of the particles 
 which make up the rays, is mainly obtained from the 
 deflexions observed when they are submitted to electric 
 and magnetic forces. If we consider the sign of the 
 charges only, all we need observe is the direction in 
 which a ray is bent, when placed in a magnetic field, 
 and in that case we may represent the effect diagram- 
 matically by Fig. 6, in which R represents a small 
 
 Fig-. 6. 
 
 cup containing radium, and the magnetic force is 
 supposed to act at right angles to the paper in such a 
 way, that the north pole of a compass needle would be 
 driven towards the back of the paper. The diagram 
 illustrates how the a rays bend very slightly to one 
 
88 Lechire III 
 
 side, while the y rays are not deflected, and the /3 rays 
 bend round to the other side in varying degrees 
 according to the velocity with which they have been 
 projected. The magnetic effect on the a rays is ex- 
 aggerated in the figure, and is rather difficult to 
 observe, so that it was thought at first that these rays 
 carried no charge until Rutherford observed a slight 
 bending, when the magnetic field was sufficiently 
 increased. 
 
 The a rays are strongly absorbed by all substances 
 on account of their small velocities ; the yS rays are 
 more penetrating, but not so much as the y rays which, 
 when of sufficient intensity, can be traced through a 
 plate of lead over 20 cms. thick. The great difference 
 in this absorbing power may be shewn by comparing 
 the diminution of intensities of the different rays in 
 the same metal such as aluminium, when it is found 
 that 8 mms. of aluminium are required to reduce the 
 y rays sent out by a radio-active material to half their 
 intensity, while half a millimetre is required for the 
 /3 rays, and only the 2000th part of a millimetre for 
 the a rays. 
 
 The magnitude of the spontaneous emission of 
 energy which takes place from a radium compound is 
 strikingly illustrated by the rise of temperature of the 
 substance above that of its surroundings. This eleva- 
 tion of temjoerature which may amount to several 
 degrees was first discovered by Curie and Laborde, 
 and is accounted for by the absorption of the a rays in 
 the radio-active body itself 
 
 We now come to an important part of the lesson 
 that may be learned by studying these phenomena. 
 
Discovery of the Emanation 89 
 
 An irregularity in the ionizing power of the radiation 
 derived from thorium being traced to air currents, led 
 Rutherford to the discovery of a gas, which is constantly 
 being generated by thorium, and is responsible for part 
 of its apparent radio-activity. This gas, which was 
 called the "emanation," has all the properties of other 
 gases ; it diffuses at a definite rate which may be 
 measured, and it solidifies when cooled down sufficiently. 
 In every respect it behaves like an elementary body, 
 but it disappears after a short life, so that in less than 
 one minute after its first appearance only half the 
 original quantity is left. No radio-active gas is given 
 out by uranium, but radium also generates an emanation, 
 and this has a longer life than that of the thorium 
 emanation, so that half the original quantity is still left 
 after about three days and twenty hours. The physical 
 properties of the emanation of radium, owing to its 
 longer life, can be more easily studied than those of 
 thorium. Its spectrum has been measured, its boiling 
 point fixed at — 64° C. for atmospheric pressure, and 
 even its density in the solid state has been approximately 
 determined. 
 
 Let us be clear about the facts. Emanations are 
 gases which are continually being produced and con- 
 tinually vanish : both processes take place at a definite 
 rate which hitherto has proved to be independent 
 of all external conditions. A steady state may be 
 reached, when the amount generated in each second 
 of time (which depends on the quantity of the parent 
 substance present), is equal to the amount that 
 disappears which is proportional to the quantity of the 
 disappearing substance. 
 
go ' Lecture III 
 
 If we take a solution of a radium salt, and extract 
 all the emanation it contains, we must wait some days 
 before a further quantity has been formed ; but the pro- 
 cess of formation continues as regularly as the process 
 of self-destruction. What happens to the emanation as 
 it disappears ? Have we here really a destruction of 
 matter ? The answer to that question is suggested by 
 the observation, that the walls of the vessel which has 
 contained the emanation acquire radio-active properties, 
 strong in proportion to the amount of emanation which 
 has disappeared. 
 
 You are now in possession of the main facts, which 
 led Professor Rutherford and Mr Soddy to a simple 
 but far-reaching theory of radio-active phenomena. 
 According to this theory, an atom or molecule is only 
 radio-active while in the act of breaking up, and radio- 
 active bodies are those which are unstable to the 
 extent, that a certain definite fraction of their atoms 
 break up every second of time. The breaking up of 
 the atom is nearly always accompanied and probably 
 caused, by the ejection of an a or of a /S particle, or 
 of both particles simultaneously ; some of the charges 
 are further accompanied by the y radiation. The 
 radio-active process is a consequence of the instability 
 of the atom and consists of a succession of steps, one 
 body being formed as the other disappears. When 
 the ultimate product is inactive and does not give rise 
 to any further emission of rays, the process ceases to 
 come to our ken ; for the quantities of matter at our 
 command are, in most cases, so minute, that the 
 radiation is the only property which affects our in- 
 struments. It will be noted that in calling a body 
 
Theory of Radio- Activity 
 
 91 
 
 " radio-active," we are guilty of the same inaccuracy 
 as we should be in saying that the powder in a gun 
 projects the bullet. The powder is the ultimate source 
 of energy of the projection, but the projection itself 
 is due to the disintegration of the powder, and not 
 to the presence of it. Similarly a radium atom ceases 
 to be radium by the very process which we call 
 radio-activity. 
 
 The successive changes which radium undergoes, 
 so far as is known at present, are illustrated in the 
 diagram (Fig. 7) : — 
 
 Ra. Em. A B C D E F 
 
 neo s-se 3 26 i9 u s 740 
 
 t^ears days minutes miinites minutes years days days 
 
 Fig. 7- 
 
 The radiation accompanying the change and the 
 life of each intermediate substance is indicated in the 
 figure, the life being measured by the time it takes 
 for half the original quantity to disappear. Thus an 
 atom of radium sending out an a particle becomes the 
 atom of a gas (the emanation), and after the further 
 emission of a similar particle, changes into a substance 
 which is called "radium A," and has a life measured 
 in the above standard by three minutes. The main ^S 
 radiation only appears in the change from radium C to 
 radium D\ It will be noted that the change from 
 radium £> to radium £ is not accompanied by any 
 
 ^ Some /i rays of small penetrating power are already given 
 out by radium B. 
 
92 Lecture III 
 
 radiation. The existence of such rayless products 
 cannot be directly proved but must be inferred from 
 indirect indications. 
 
 If Rutherford's explanation be accepted as correct, 
 as it universally is at the present time, it means that 
 at least some of our chemical elements are in the act 
 of breaking up. To say that radium is not an element 
 does not dispose of the question, because if we deny 
 to radium the claim of being a chemical element we 
 must also do so to thorium and uranium and where 
 shall we draw the line ? Shall we revise our language 
 and say that these metals are really compounds and 
 not simple bodies ? It would be a pity if we did, for 
 there is no known distinction, except as regards their 
 radio-activity, between uranium, thorium or radium, 
 and other chemical elements. These bodies are cer- 
 tainly not chemical compounds in the common sense 
 of the word. No doubt the name " atom " in its 
 original meaning " uncut," and more especially in its 
 traditional meaning " uncuttable," is no longer strictly 
 applicable and this has given rise to critical comments 
 on the unscientific use of scientific language. The 
 late Lord Kelvin was a purist in such matters, and I 
 was once present at a discussion in an English country- 
 house, when he denounced, in forcible language to the 
 guests, what he called the absurdity of speaking of the 
 division of an undivisible body. He was interrupted 
 by the son of the host, who was no respecter of 
 authority, and turning round to the interested audience 
 remarked, " There you see the disadvantages of 
 knowing Greek." This remark derived some of its 
 point from the discussion which then occupied the mind 
 
Decay of the Atom 93 
 
 of Cambridge University regarding the advisability 
 of keeping Greek as a compulsory subject in the 
 entrance examination. It is impossible to modify our 
 language continually, in order to make it keep pace 
 with the advance of knowledge, and we must therefore 
 be allowed to retain a word defining a group of certain 
 entities, even if we have altered our ideas of the nature 
 of these entities. The original name may etymo- 
 logically be inappropriate, but as a label it retains its 
 meaning. We shall continue to speak therefore of 
 atoms just as we continue to speak of sunrise and 
 sunset, which according to our present view are also 
 incorrect expressions. 
 
 Let us return to the gradual transformation of the 
 radium atom. If radium be left to itself for a con- 
 siderable time, it must be accompanied by all its 
 products of decay in such quantities, that for each 
 atom of radium that disintegrates an atom of each of 
 the products must also disappear. This will readily be 
 understood if it be remembered, as already explained 
 that an equilibrium state in the amount of the emanation 
 is reached when one atom of the emanation that 
 disappears is replaced by one atom generated. But 
 each atom of emanation that is generated implies the 
 disappearance of one atom of radium. When the 
 same reasoning is applied to all bodies of the series, 
 our statement will be seen to be correct. If all 
 intermediate products be taken into account, it is 
 found that four a particles are ejected for each atom 
 of radium that disintegrates, and if we measure the 
 whole number of a particles which are ejected by 
 radium in its complete transformation, we obtain a 
 
94 Lecture III 
 
 measure of the number of radium atoms which break 
 up in a given time. Such measurements and cal- 
 culations have led to the result\ that if we start with 
 a certain quantity of radium, half of it will have 
 disappeared after 1760 years and the millionth part 
 only would remain after 30,000 years. 
 
 If this be true, how is it that there is any radium 
 on the earth ? Ought not all measurable quantities 
 to have disappeared over and over again in geological 
 times ? The answer is obvious. Radium must con- 
 tinuously be forming as well as disappearing. Our 
 endeavour, therefore, must be to discover its parent. 
 Thorium and uranium suggest themselves as possible, 
 both being present in pitchblende which is the main 
 source of our radium, but various lines of argument 
 point unmistakably to uranium and not to thorium as 
 the ancestor of radium. It may be said, no doubt, 
 that we are only pushing the difficulty further back, 
 as we must now enquire into the parentage of uranium. 
 We possess, at present, no information on that point, 
 but it is to be noted that the rate of change of uranium 
 is only about the one millionth part of that of radium, 
 so that one gram of uranium would only lose one 
 milligram in a million years. MX science consists in 
 pushing difficulties further back, and we may therefore 
 be satisfied with having traced the ancestry of radium 
 sufficiently far back, to include in our range of know- 
 ledge all radio-active effects which have been appreciable 
 since the first solidification of the earth. 
 
 I must next speak of an important discovery 
 suggested by the theory of disintegration. What 
 ' Rutherford and Geiger, Proc. Royal Soc. lxxxi. p. 162 (1908). 
 
Helium and the a Ray 95 
 
 becomes of the a particle ? This is a question which 
 quite naturally presents itself. At its birth this particle 
 has a positive charge, but it soon becomes electrically 
 neutral either by losing its charge or by attaching 
 to itself a negative electron from outside. For the 
 purpose of identifying this neutral particle which must 
 have a mass about twice that of the hydrogen molecule, 
 Rutherford argued in this manner : — All final products 
 of the radio-active change must collect and accumulate in 
 quantity. The a particle is one of these products, and 
 to identify it, we must look for an element which invari- 
 ably accompanies radium or other bodies giving out a 
 rays. Such an element is helium, which is invariably 
 found in radio-active minerals. Now helium has a mass 
 which indicates that it might itself be the a particle. 
 
 That helium is actually being produced during the 
 decay of radium was subsequently confirmed by Sir 
 Wm Ramsay working in conjunction with Soddy, 
 and though the fact was at first received with some 
 scepticism, it has now been proved beyond doubt. 
 We may even measure the rate at which helium is 
 generated from radium, the most recent determinations 
 giving 158 cubic millimetres per year for each gram of 
 radium. 
 
 The Hon. R. J. Strutt' having measured the amount 
 of helium accumulated in rocks containing radio-active 
 substances, by comparing the quantity of the gas 
 present, with the amount of radium or uranium present, 
 was able to draw some interesting and important 
 conclusions. If no helium has escaped in the course 
 ' Froc. Royal Soc. Lxxxi. p. 272 (1908). 
 
96 Lecture III 
 
 of time, the quantity present would give us a measure 
 of the age of the rock since it was first deposited in 
 a sufficiently coherent form to retain the gas. If some 
 helium has escaped, we should still be able to form an 
 estimate of the shortest time that the rock can have 
 been in its present form. Unfortunately coprolites 
 and phosphatised fossil bones are the only substances 
 which allow an accurate investieation. As was to be 
 expected the younger deposits do not give so large a 
 quantity of accumulated helium as the older ones. 
 The longest life-time found was that of a haematite 
 overlying carboniferous limestone, which must have 
 been in existence for at least 141 million years. 
 
 I have already given reasons why we should keep 
 uranium and thorium in the list of bodies which we 
 continue to call elements ; the same reasons apply 
 equally to radium and to all products of radio-activity, 
 as no line can be drawn between uranium, radium, 
 the emanations and the other more or less fugitive 
 bodies which form the intermediate steps in the decay 
 of radium. Some of these no doubt are very short- 
 lived, but longevity has never been a decisive factor in 
 a fundamental classification. We should therefore 
 retain the word "element" but use it in connexion 
 with the building up of compound bodies, in the sense 
 in which bricks might be called the " elements " with 
 which houses are built, irrespective of the question, 
 whether bricks are themselves made up of smaller 
 constituents. 
 
 What is the ultimate result of the radio-active pro- 
 cess .'' The descendants of radium have been traced 
 
Modification of former Ideas 97 
 
 as far as radium F, which Rutherford has shewn to be 
 polonium. This body on ejecting an a particle changes 
 into an inactive substance, the identity of which is not 
 yet quite certain, though Boltwood is probably right in 
 thinking that it is lead, which thus can claim to be the 
 immortal descendant of a family whose pedigree has 
 been traced back to uranium. 
 
 With the facts and theories of radio-active change 
 before us, we are able to consider the modifications 
 of our former interpretations of physical phenomena, 
 which recent discoveries sug-aest. Ever since Galileo, 
 Huygens and Newton taught us the elements of dyna- 
 mics we have considered inertia to be an inherent and 
 fixed property of the centres of force, to which we have 
 attached the idea of matter. Owing to the apparently 
 rigorous proportionality between mass and gravitational 
 force and the predominant influence of gravitation on 
 all our doings in this world, we have acquired the 
 habit of connecting mass and weight so intimately to- 
 gether that we can hardly separate one from the other. 
 Nevertheless it is essential that we should do so, and 
 place — as indeed Newton did — mass on an indepen- 
 dent footing. The ratio of a force to the acceleration 
 it produces may serve the purpose, or we may have 
 recourse to the doctrine of energy. If a particle move 
 with a velocity v and we take its kinetic energy, relative 
 to the space in which it moves, in accordance with the 
 common definition to be \mv\ m is a factor which 
 on the views of 25 years ago, belongs to the particle, 
 as distinguished from the medium which surrounds it. 
 Whether these views be correct or not, it is important 
 to realise clearly, that energy cannot be expressed 
 
 .s. 7 
 
98 Lecture III 
 
 entirely by quantities, which merely depend on change 
 of position (kinematical factors), but must involve some- 
 thing which the old idea of inertia was intended to 
 cover. However much we may alter our views on the 
 nature of inertia, our conception of natural phenomena 
 must always, directly or indirectly, include this idea of 
 inertia. The belief in the constancy of the factor in 
 was justified by the experience, which taught us that 
 in the visible motion of material bodies, energy seemed 
 to be strictly proportional to the square of the velocity. 
 Without abandoning the old Newtonian dynamics, 
 we should note, however, that the above simple de- 
 rivation of mass from energy may lead us into error 
 when applied to a system of connected 
 bodies. Take, for instance, a heavy fly 
 wheel (Fig. 8) of mass M and radius r, and 
 let a string be wound round the central 
 shaft of radius a, one end of the string 
 hanging down with a mass in attached to 
 the end. Neglecting the masses of the strincr and 
 of the projecting portions of the shaft, the energy 
 
 of motion may be expressed by ~{M~ +m\v-, where 
 
 V is the velocity of the mass in. If now, the fly 
 wheel and the string were invisible, and we could 
 only observe and measure the energy of the mass m, 
 we might be misled into believing that its mass were 
 
 M ^, +7n\ . This fictitious mass, which may be great 
 
 compared to M, is something which, partaking no 
 doubt of the nature of a true mass, cannot be identified 
 with any particular body or system of bodies, because 
 
Electric Inertia 99 
 
 it involves the linear quantities r and a. If in the 
 extreme case m becomes zero, the mass which would 
 become identified with the particle is located entirely 
 outside it. If further we imagine that, perhaps through 
 centrifugal force, the radius r changes with the velocity, 
 we should have constructed a model shewing how mass 
 may depend on velocity. 
 
 Keeping the above in view, let us now estimate the 
 energy associated with the motion of an electrified 
 particle. Consider a sphere of radius a, uniformly 
 charged with a quantity of electricity e. The expen- 
 diture of energy connected with the act of charging 
 need not be taken into account, if for the sake of 
 simplicity we assume the distribution of the charge on 
 the sphere to remain constant during the motion. The 
 moving charge excites a magnetic field in the sur- 
 rounding space and, as first shewn by Sir J. J. Thomson, 
 the total energy of the field is e-v-JT,a. It follows that 
 even though the sphere had no mass in the ordinary 
 sense, it could not be set into motion without work 
 being done upon it, and it would generally behave, as 
 if it had a mass equal to 2e-l T,a. It may be said that 
 this mass is fictitious because we cannot locate it, but 
 just as in the case of the invisible fly wheel of the 
 above example, this does not abolish the idea of inertia, 
 but only transfers that idea from the centre of force to 
 the surrounding region. 
 
 Assuming the electron to be a charged sphere, we 
 can calculate its radius by equating Thomson's ex- 
 pression for the mass to its value as determined from 
 experiments on kathode rays. We obtain in this way 
 a number approximately equal to 2x10"'" cms. As 
 
 7—2 
 
loo Lecture III 
 
 the molecular diameter is, approximately, io~''cms., an 
 electron is something much smaller than a molecule. 
 The phenomena of radio-activity have proved, that 
 the atom of such a body as uranium must contain a 
 number of positive and negative electric charges and a 
 certain part of the atomic mass must therefore reside 
 in the electro-magnetic field immediately surrounding 
 the atom. Whether that part is considerable, or per- 
 haps even predominant, remains to be proved. But 
 without waiting for such proof it is tempting as a 
 plausible, though perhaps bold and risky hypothesis, 
 to deny the existence of a Newtonian mass altogether, 
 and to reduce everything to electric inertia. This 
 leads us to a novel theory of matter, according to which 
 an atom is entirely made up of negative electrons and 
 positive charges. As the mass of an electron is about 
 a thousand times smaller than the mass of the atom of 
 hydrogen, it would follow that, assuming positive and 
 negative charges to have equal masses, an atom of 
 hydrogen is composed of about 500 electrons of each 
 kind. But there is some doubt as to the proper way 
 of treating positive electrons which have not hitherto 
 been isolated. If the a particle consisted of a single 
 positive charge, its large mass could only be accounted 
 for, by assuming it to have a diameter a thousand 
 times smaller than that of the nerative electron, and 
 the possibility of this seems to be excluded by the 
 close relationship between the a particle and the atom 
 of helium. It is reasonable, on the contrary, to con- 
 clude that the a particle consists of a large number of 
 negative and positive charges, with an excess of one 
 positive charge. 
 
Nature of Atom loi 
 
 The great velocity with which electrons and a 
 particles are ejected from radio-active substances shews 
 that the ultimate components of the atom must be in 
 a state of violent motion. 
 
 This suggests an atom to be something of the nature 
 of a planetary system with the electric atoms revolving 
 round each other as constituents. The question of the 
 stability of such systems deserves, and has received, 
 considerable attention, but no completely satisfactory 
 solution has yet been found. A somewhat serious 
 difficulty is presented by the dissipation of energy 
 through electro-magnetic radiation, which accompanies 
 all except a strictly uniform motion of electric charges. 
 The electric inertia of a moving electron resides, as 
 has been explained, in the electro-magnetic field sur- 
 rounding the charge. If the velocity of the electron 
 be constant in direction as well as magnitude, a steady 
 state is reached when the magnetic field follows the 
 charge in its motion, but as soon as either the velocity 
 increases or the direction changes, a new field has to 
 be established, and the change is effected by means of 
 electro-magnetic radiations, propagated outwards from 
 the moving charge with the velocity of light. The 
 electron is thus called upon to supply a certain amount 
 of energy which is dissipated into space, and this energy 
 must have its equivalent in the work done on the 
 electron itself, by a force resisting the change of motion. 
 There is consequently always something in the nature 
 of a retarding force whenever the motion of a system of 
 electrons is such as to cause a change in the magnetic 
 field. A single electron rotating round a fixed centre 
 would quickly dissipate its energy into space by 
 
I02 Lecture III 
 
 radiation, but if there were a continuous electrified ring 
 involving no change in the magnetic field duringrotation, 
 there would be no such dissipation. A ring formed 
 by separate corpuscles will approach the state of the 
 continuous ring the more nearly the closer the electrons 
 are together. Hence the dissipation of energy into 
 space depends on the internal configurations. I must 
 refer you to the original memoirs dealing with this 
 subject for a further discussion of these and other 
 matters, which have to be well weighed before an 
 opinion on the all-electric origin of mass can be 
 formed. The writings of Larmor, who first shewed 
 the possibility of constructing an atom by means of 
 moving electric charges, and of Lorentz, will shew 
 that difficulties are not shirked, though they are not 
 always solved. 
 
 While the scientific world is recovering from the 
 shock of a great experimental discovery, there is 
 generally a reaction, when explanations are showered 
 upon it intended to demonstrate that so far from being 
 surprising, the new discovery is exactly what ought to 
 have been predicted. It was to be expected, therefore, 
 that we should now be instructed how we might have 
 foreseen radio-activity, if we had only adopted the 
 electric theory of matter a little sooner. For if, as 
 explained above, any motion in which the magnetic 
 field does not remain strictly constant must gradually 
 dissipate its energy, the rotating systems of electrons 
 must sooner or later reach a state at which instability 
 occurs. There may be the germ of truth in this and 
 we may look forward to further developments of the 
 idea, but many monsoons will pass over this country 
 
Dissipation of Energy 103 
 
 before the mechanism of an atom is laid bare to our 
 understanding. 
 
 An important appHcation of the dissipation of 
 electric energy through radiation has been made by 
 Sir J. J. Thomson in his theory of Roentgen rays. 
 When an electron projected from the kathode meets 
 with an obstacle so that its motion is strongly retarded, 
 an electro-magnetic pulse spreads out from it into the 
 surrounding space. This accounts for Roentgen's 
 discovery in so natural a manner, that it has readily 
 been accepted by men of science, and will not lightly 
 be abandoned, though Professor Bragge has recently 
 advocated, and given some grounds for the view, that 
 the Roentgen radiation is, like the a and /8 ray, corpus- 
 cular in its nature and electrically neutral only, because 
 two oppositely electrified particles are united in it. 
 
 Connected with the dissipation of electro-magnetic 
 energy through radiation, because dependent like it on 
 the finite velocity of light, is the change in the apparent 
 mass of the electron when its velocity alters. The value 
 I have given above for the mass of an electron only 
 applies when its velocity is small compared with that 
 of a wave of light, but we may conclude from the 
 theory, that with increasing velocity the apparent mass 
 increases until it becomes infinite, when the velocity of 
 light is reached. This would mean that ^no velocity 
 greater than the velocity of light is possible, which is 
 an important deduction on which experimental con- 
 firmation would be welcome. We possess indeed 
 experiments by Kaufmann and G. Bucherer which, in 
 the case of electrons projected from a kathode, shew 
 this increase of mass with velocity. A velocity of 
 
I04 Lecture III 
 
 more than nine-tenths of that of light was obtained, 
 and the apparent mass was found to be more than 
 three times as great as that for small velocities. 
 Experiments of this nature may furnish an answer 
 to the question whether the masses of electrons are 
 entirely electro-magnetic, or whether there is a remnant 
 of the old Newtonian mass. The answer, which at 
 present is perhaps not altogether decisive, is generally 
 taken as favouring the extreme view. 
 
 The state of plasticity and flux — a healthy state, in 
 my opinion — in which scientific thought of the present 
 age adapts itself to almost any novelty, is illustrated 
 by the complacency with which the most cherished 
 tenets of our fathers are beino- abandoned. Thouo^h 
 it was never an article of orthodox faith that chemical 
 elements were immutable and would not some day be 
 resolved into simpler constituents, yet the conservation 
 of mass seemed to lie at the very foundation of creation. 
 But now-a-days the student finds little to disturb him, 
 perhaps too little, in the idea that mass changes with 
 velocity; and he does not always realise the full meaning 
 of the consequences which are involved. Does he 
 know, for instance, that the total mass of two electrons 
 placed side by side are not equal to the sum of their 
 two separate masses, but greater if they moved in the 
 same direction, and smaller when moving in opposite 
 directions. That electrons circulatinsf inside the atom 
 never get sufficiently near for this change of mass to 
 become appreciable, does not seem to me an answer 
 to the difficulty, for it leaves out of account the 
 cumulative effect of the large number of electrons 
 concerned. Were it allowable to neglect this cumu- 
 
Variability of Mass 105 
 
 lative effect, the inertia of an ordinary electrical circuit, 
 would be proportional to its length, irrespective of its 
 shape and the inertia of coil proportional to the number 
 of its windings, while we know of course that it is 
 proportional to the square of that number. The reason 
 of the discrepancy becomes apparent when we consider 
 that the electro-magnetic energy varies as the square 
 of the electro-magnetic force, and that therefore, a 
 thousand electrons arranged along an arc of a circle 
 produce at its centre a field which per unit volume is 
 a million times greater than that due to the single 
 electron. Though we may therefore neglect the field 
 beyond the immediate surroundings of a moving 
 electron when dealing with it alone, we are no longer 
 justified in doing so when a large number of electrons 
 are concerned, even though their distances may be 
 great compared with their dimensions. 
 
 If I draw attention to this change of mass, which 
 according to the present theory must necessarily, to 
 some extent, accompany all chemical combination, it is 
 not in order to discredit that theory, but rather to point 
 to a field of experimental research, which some day 
 may lead us to a fresh experimental confirmation. So 
 far however in spite of many careful efforts, mass as 
 determined by weight has never shewn any tendency 
 to be affected by chemical combination. 
 
 I have not yet finished my tale of the surprises 
 sprung upon us by the theoretical pioneers, who at the 
 present moment rival the experimental discoverers in 
 boldness. One startling conclusion not quite the most 
 daring, but running it pretty closely, arose out of an 
 experiment made by the American man of science, 
 
io6 Lecture III 
 
 Michelson, and repeated later with much increased 
 accuracy by him, jointly with E. VV. Morley. The 
 experiment was intended to clear up an obscure point 
 in all our theories of light. If the aether is sufficiently 
 solid to transmit torsional vibrations with a considerable 
 velocity, how is it that matter can move freely through 
 it. This difficulty frequently occupied the thought of 
 Sir George Stokes, and previously Fresnel had tried to 
 find out experimentally, whether, when a transparent 
 body is in motion, it carries the aether filling the space 
 between its molecules, with it, or whether that aether 
 remains at rest. Fresnel's efforts remained undecisive 
 because, according to our present ideas on the cause of 
 refraction, the transmission of light through a moving 
 body must be affected by the motion of the molecules, 
 even though the aether remain at rest. The intro- 
 duction of the electro-magnetic theory, getting rid of 
 all idea of "solidity," seemed to lighten, and according 
 to the views of Sir Joseph Larmor, altogether to 
 remove this particular difficulty, if the aether be assumed 
 to remain absolutely at rest. Michelson's experiment 
 was intended to serve as a delicate optical test of the 
 fixity of the aether by comparing the velocity of light 
 in the direction in which the earth, as a whole, moves 
 with the velocity at right angles to the motion. It is 
 the orbital motion of the earth round the sun which 
 comes into play, because it is much greater than that 
 of the diurnal rotation, while the effects of a translation 
 of the whole solar system through space may be 
 eliminated by repeating the observations at different 
 periods of the year. 
 
 The result of Michelson's experiment was the 
 
The Michelson-Morley Experiment 107 
 
 velocity of light between two points on the earth's 
 surface is the same whatever relation the direction of 
 the line joining the points may have to the direction 
 of the earth's orbital motion. At first sight this 
 seems intelligible only on the view that the aether near 
 the surface of the earth takes part in the orbital 
 motion, but if this were true a large number of other 
 difficulties would at once arise. The importance of 
 the question makes it desirable to indicate the nature 
 of the experiments, at any rate in outline. Let A and 
 B (Fig. 9) be two points in the direction of the earth's 
 motion, and C a point in a line at right 
 angles to it and at the same distance """ 
 from A as B. If a wave of liafht startingf 
 from A travels with velocity V, the . 
 
 distance / is traversed in time l\V, but A b 
 
 by the time the wave reaches the point 
 where B was, when the wave left A, that point has 
 moved forward, owing to the bodily motion of the 
 earth, which we imagine to be in the direction from 
 A to B. The problem of calculating the time at 
 which the wave reaches an object placed at B, moving 
 with the earth, with a velocity v, is the same as that 
 of finding the time when one body leaving A with a 
 velocity V, overtakes another body leaving B with 
 velocity v. This gives at once 
 
 Vt = l+z>i 
 
 and hence t is equal to {IjV—v), and the actual length 
 of the path traversed by the light in passing from A 
 to B instead of being / is now /F/( V—v). If there be 
 a mirror at B which returns the light to A, we find 
 
io8 Lecture III 
 
 similarly that the path traversed in the return passage 
 is IVjV+v, so that the total path, including the two 
 journeys, is : 
 
 ,../ I I \ 2IV- 
 
 If the fraction vjVhn sufficiently small to allow its 
 fourth power being neglected, the expression for the 
 total path may be taken as equal to 
 
 2i{V' + zr)IV\ 
 
 On the other hand, if a ray of light passes from A to 
 a mirror at C and is reflected back to A, the path is 
 found to be equal to 
 
 The difference in length between the paths is lv'IV\ 
 and it is this difference in length that Michelson set 
 himself to observe. The value of vjV is about one 
 divided by ten thousand, so that the experiment, to be 
 successful, requires the measurement of the hundred 
 millionth part of the distance /; the resources of optics 
 are equal to this task, thanks in great part to Michelson 
 himself It is of course impossible to adjust the 
 distances AB and ^C to absolute equality, but this 
 does not matter provided that the whole apparatus can 
 be rotated through a right angle, so that either length 
 can be placed in the direction coinciding with that 
 of the velocity of the earth. The conclusion finally 
 arrived at by Michelson and Morley was, that so far as 
 their experiments could decide, the difference in path, 
 calculated on the hypothesis of a quiescent aether did 
 not exist; but a quiescent aether is one of the few 
 
The Michelson-Morley Experiment 109 
 
 necessities of modern physics. How are we to meet 
 this difficulty ? Mathematicians are always sufficiently 
 resourceful to cope with any problem set to them by 
 the experimentalist, and as Poincare likes to tell us, 
 we can always find a new hypothesis to fit a new fact. 
 The hypothesis in the present instance was supplied 
 independently by FitzGerald and Lorentz. The 
 mathematical reasoning on which the above expression 
 for the difference in optical path of the rays going re- 
 spectively with the earth's motion and at right angles 
 to it, is based, assumes that the length / remains fixed 
 as the apparatus is rotated through a right angle. It 
 seems possible at any rate that the dimensions of what 
 we call a rigid body may alter according to the direction 
 of its motion through space. If a contraction of appro- 
 priate amount took place along all lines in a moving 
 body parallel to the direction of motion, Michelson's 
 negative result could be explained. Further, Lorentz 
 and Larmor have given grounds for believing that 
 if molecular forces have ultimately an electric origin, 
 such a contraction ought to take place. The matter 
 did not rest there, mainly perhaps because the explana- 
 tion was felt by some to be a little artificial. Looking 
 at the experiment in a broad way it shews, that at any 
 rate with our present methods of observation, we are 
 unable to detect any relative motion between matter 
 and aether, but only relative motion between different 
 portions of matter. Let us take the point of view 
 that this is not a consequential effect depending on a 
 peculiar influence of velocity on molecular forces, such 
 as is required by FitzGerald's explanation, but a law 
 of nature holding universally. Einstein, in a paper of 
 
I lo Lecture III 
 
 great interest and power, has developed this idea, 
 calhng his imagined law " The principle of relativity," 
 because it stipulates — a priori — that only relative 
 motion between material bodies can be detected. It 
 is impossible for me to discuss in detail the reasoning 
 by which this principle is justified, and an account 
 without explanations of its consequences would lay 
 me open to the charge that I was pla)'ing with your 
 credulity. Suffice, therefore, it to say that strict 
 adherers to the principle cannot admit the existence 
 of an aether, and yet may speak of the transmission of 
 light through space with a definite velocity. They 
 must further accept, as a consequence of their dogma, 
 that identical clocks placed on two bodies moving 
 with different velocities have different rates of going 
 and that, even on the same body, identical clocks 
 indicate different times, when the line joining their 
 positions lies in the direction of motion. The motion 
 must be determined relative to another body, which is 
 supposed to be at rest, and a clock placed on that body 
 must serve as the ultimate standard of time. The 
 theory appears to have an extraordinary power of 
 fascinating mathematicians, and it will certainly take 
 its place in any critical examination of our scientific 
 beliefs ; but we must not let the simplicity of the 
 assumption underlying the principle hide the very 
 slender experimental basis on which it rests at present, 
 and more especially not lose sight of the fact, that it 
 goes much beyond what is proved by Michelson's 
 experiment. In that experiment, the source of lio-ht 
 and the mirrors which reflected the light were all 
 connected together by rigid bodies, and their distances 
 
Principle of Relativity 1 1 1 
 
 depended therefore on the intensity of molecular forces. 
 Einstein's generalisation assumes that the result of the 
 experiment would still be the same, if performed in a 
 free space with the source of light and mirrors dis- 
 connected from each other but endowed with a common 
 velocity. This is a considerable and, perhaps, not 
 quite justifiable generalisation. I am well aware that 
 Bucherer's experiments with kathode rays are taken 
 to confirm the validity of Einstein's principle, but if 
 we say that they are not inconsistent with it, we 
 should probably go as far as is justifiable. 
 
 In bringing the principle of relativity to your notice 
 I have characterised it as one of the most startling 
 developments of recent science, and you may ask why 
 I have qualified the superlative ; have I reserved some 
 even more surprising fancy of the scientific imagination ? 
 That is indeed the case, though it may be difficult 
 to place modern philosophic speculations in order of 
 revolutionary merit. When in the presidential address 
 to the Physical Section of the British Association at 
 Edinburgh, I alluded to certain views on the trans- 
 mission of energy which had been advocated, I stated 
 that they could only be true if energy had, like matter, 
 an atomic constitution. I thought I had thus finally 
 disposed of the matter by a rediictio ad absitrditvi, but 
 now-a-days such atomic constitution is openly advo- 
 cated. It all arose out of the theoretical investigations 
 intended to account for the energy radiated by a hot 
 'and ideally black body, and more especially the dis- 
 tribution of that energy among the waves of different 
 frequencies. The law of partition of energy which 
 holds in this case is correctly represented by a formula 
 
112 ■ ■ Lecture III 
 
 which we owe to Max Planck, who deduced it 
 theoretically by assuming that the radiation takes place 
 in such a way that a molecule can only radiate either 
 a certain definite minimum quantity or some multiple 
 thereof. This means that as far as this type of energy 
 is concerned, it behaves as if it were made up of a 
 number of finite bits of equal value. As in the case of 
 Michelson's experiment, this result has been generalised 
 and the hypothesis has been formulated that all energy 
 is made up of finite bits. That such a hypothesis 
 should be advocated by men whose opinion deserves 
 the most serious consideration, shews the restless 
 turmoil which agitates the scientific thought of the 
 present day. During the last century we felt sure that 
 we were building our scientific edifice on a secure 
 basis ; to-day many have become suspicious of the 
 soundness of its foundations, while some are already 
 digging for new ones or tinkering with the old ; there 
 are still however a few optimists left who try to go on 
 building upon the old structure in the perhaps mistaken 
 belief that it still stands on firm ground, and will 
 remain standing just as in the material world we find 
 some edifices sufficiently strong to survive the ravages 
 of an earthquake. 
 
 I confess to a feeling of relief in stepping now 
 from the steep and slippery slopes of universal theories 
 unto more modest but safer ground. Whatever may 
 be the ultimate constitution of matter, there is no doubt 
 that electrons are attached to the molecules and are 
 responsible for the radiations of light and heat which 
 they emit. It has already been briefly alluded to in 
 the first lecture that Zeeman discovered a change in 
 
Further Results of Electron Theory 1 1 3 
 
 the waves emitted by an incandescent body, when that 
 body is under the influence of a strong magnet. From 
 the point of view of the electron theory this is easily 
 explained, because the origin of the wave of light is 
 an alternating motion of an electrified particle, which 
 must obey the well-known laws of electro-magnetic 
 force. Indeed, the discovery of Zeeman was in- 
 complete until Lorentz applying theory pointed out to 
 him that he ought to look for circular and elliptic 
 polarisation. The magnitude of the effect cannot 
 easily be predicted by theory, as it depends on the 
 mutual interaction of the electrons which take part in 
 the emission of light. In the simplest possible case, 
 observation and theory combined gives us the relation 
 of the quantity of electricity concerned, to the apparent 
 mass of the vibratory system, and as this for some of 
 the spectroscopic lines is found to be nearly the same 
 as the ratio of quantity to mass deduced from experi- 
 ments on kathode rays, we must conclude that in the 
 simplest cases, the vibrating system consists of an 
 electron which can vibrate independently of others. 
 
 It is not possible within the limits of this lecture 
 to give an account of the various applications of the 
 electron theory, which has been found capable of 
 solving many difficulties, but some reference must be 
 made to the work of Drude, which involves a new 
 idea, and gives a rational explanation of the parallelism 
 between electric and thermal conductivities. Drude 
 imagines a number of free electrons within a conductor, 
 which acquire a velocity in virtue of the internal 
 thermal energy of the body. The average energy 
 of these electrons should be equal to the average 
 
 s. 8 
 
1 1 4 Lecture III 
 
 energy of a gaseous molecule at the same temperature 
 and this is known. Heat conduction, according to 
 Drude, is not due to the interchange of energy from 
 atom to atom, as previously believed, but takes place 
 through the intermediate action of these electrons. 
 The reflexion of light at the surface of metals was 
 also examined by Drude, and an appropriate explana- 
 tion of the facts given by taking account both of the 
 free electrons and of others attached to the molecules 
 but free to vibrate. 1 shewed subsequently how an 
 estimate may be formed from optical phenomena of 
 the number of electrons which are capable of free 
 motion in the metals, their number being from two 
 to four or five for each molecule of the substance. 
 
 I have during these lectures contrasted on several 
 occasions the former tendency to base our theoretical 
 explanations of natural phenomena on definite models 
 which we can visualise and even construct, with the 
 modern spirit which is satisfied with a mathematical 
 formula, and symbols which frequently have no strictly 
 definable meaning. I ought to explain the distinction 
 between the two points of view which represent two 
 attitudes of mind, and I can do so most shortly by 
 referring to the history ot the electro-dynamic theory 
 of light, the main landmarks of which I have already 
 pointed out in the second lecture. The undulatory 
 theory — as it left the hands of Thomas Young, Fresnel 
 and Stokes — was based on the idea that the ajther 
 possessed the properties of an elastic solid. Maxwell's 
 medium being quite different in its behaviour, its author 
 at first considered it to be necessary to justify the 
 possibility of its existence, by showing how, by means 
 
Model Making versus Evasive School 1 1 5 
 
 of fly wheels and a peculiar cellular construction, we 
 might produce a composite body having the required 
 properties. Although later Maxwell laid no further 
 stress on the ultimate construction of the medium, his 
 ideas remained definite and to him the displacements 
 which constituted the motion of light possessed a con- 
 crete reality. In estimating the importance of the 
 support which Maxwell's views have received from 
 experiment, we must distinguish between the funda- 
 mental assumptions on which Maxwell based his 
 investigations and the mathematical formulae which 
 were the outcome of these investigations. It is clearly 
 the mathematical formulae only which are confirmed 
 and the same formulae migfht have been derived from 
 quite different premises. It has always been necessary, 
 as a second step of a great discovery, to clear away 
 the immaterial portions which are almost invariable 
 accessories of the first pioneer work, and Heinrich 
 Hertz, who besides being an experimental investigator 
 was a philosopher of great perspicacity, performed this 
 part of the work very thoroughly. The mathematical 
 formula instead of being the result embodying concrete 
 ideas, now became the only thing which really mattered. 
 To use an acute and celebrated expression of Gustav 
 Kirchhoff, it is the object of science to describe natural 
 phenomena, not to explain them. When we have 
 expressed by an equation the correct relationship be- 
 tween different natural phenomena we have gone as 
 far as we safely can, and if we go beyond we are 
 entering on purely speculative ground. I have nothing 
 to say against this as a philosophic doctrine, and I 
 shall adopt it myself when lying on my death-bed, if I 
 
 8—2 
 
1 1 6 Lecture III 
 
 have then sufficient strength to philosophise on the Hmi- 
 tations of our intellect. But while I accept the point of 
 view as a correct death-bed doctrine, I believe it to be 
 fatal to a healthy development of science. Granting 
 the impossibility of penetrating beyond the most super- 
 ficial layers of observed phenomena, I would put the 
 distinction between the two attitudes of mind in this 
 way : One glorifies our ignorance, while the other 
 accepts it as a regrettable necessity. The practical 
 impediment to the progress of j^hysics, of what may 
 reluctantly be admitted as correct metaphysics, is both 
 real and substantial and might be illustrated almost 
 from any recent volume of scientifiic periodicals. 
 Everyone who has ever tried to add his mite to 
 advancing knowledge must know that vagueness of 
 ideas is his greatest stumbling-block. But this vague- 
 ness which used to be recognised as our great enemy 
 is now being enshrined as an idol to be worshipped. 
 We may never know what constitutes atoms or what 
 is the real structure of the cCther, why trouble therefore, 
 it is said, to find out more about them. Is it not safer, on 
 the contrary, to confine ourselves to a general talk on 
 entropy, luminiferous vectors and undefined symbols ex- 
 pressing vaguely certain physical relationships? What 
 really lies at the bottom of the great fascination which 
 these new doctrines exert on the present generation is 
 sheer cowardice : the fear of having its errors brought 
 home to it. As one who believes that metaphysics is 
 a study apart from physics, not to be mixed up with 
 it, and who considers that the main object of the phy- 
 sicist is to add to our knowledge, without troubling 
 himself much as to how that knowledge may ultimately 
 
Danger of Generalities 1 1 7 
 
 be interpreted, I must warn you against the temptation 
 of sheltering yourself behind an illusive rampart of 
 safety. We all prefer being right to being wrong, but 
 it is better to be wrong than to be neither riofht nor 
 
LECTURE IV 
 
 Difficulties cease to trouble us either when they 
 are surmounted, or when we have become accustomed 
 to them. As soon as we observe a rare phenome- 
 non, our brain starts working to find some plausible 
 explanation, while matters of daily occurrence, which 
 may be the most puzzling of all, are treated with 
 indifference. Every one, for instance, wants to know 
 why a comet has a tail, yet how few ever trouble to 
 think why the sky is blue or the sunset red ; and many 
 a man wants to have wireless telegraphy explained to 
 him, though he does not know — or care — how the 
 ordinary telegraph works. This tendency affects not 
 only those who take only a spasmodic interest in 
 science, but perhaps even to a higher degree the 
 professional investigator. For every hundred learned 
 men who are deeply interested in an experiment which 
 can only be shewn with the most delicate and costly 
 apparatus, perhaps one troubles to think why the earth 
 is a magnet or charged with negative electricity, and 
 as regards the mystery of gravitation, we have long 
 given it up as a hopeless subject to worry about. It 
 is not that we should not like to know the causes of 
 all these things, but we are so accustomed to meet the 
 
Observational Sciences 1 1 9 
 
 questions which are daily before us, that they cease to 
 assert that spell of curiosity which is the necessary 
 antecedent to their being successfully attacked. I 
 propose to-day to bring to your notice a few subjects 
 which have suffered on account of the contempt, — 
 neglect would perhaps be a more polite word, — 
 which is proverbially bred by familiarity. 
 
 The progress of knowledge which I described in 
 the three previous lectures was made by means of what 
 is called the experimental method of investigation. 
 Here we fix our attention on some, perhaps, casual 
 observation, and by varying artificially the condi- 
 tions we subject nature, as it were, to a severe 
 cross-examination, which, if it prove successful, may 
 divulge some deeply hidden secret ; but when, for 
 instance, we wish to investigate the connexion of 
 sun-spots with terrestrial magnetism, we are unable to 
 affect the phenomena to be investigated, and are 
 reduced to the purely observational method. We 
 can do nothing but sit, watch, and note carefully, and 
 be satisfied with basing our reasoning on comparatively 
 few facts, because the work of an individual is limited 
 to the time of his life. Yet this patient observation 
 of facts which happen without our interference, often 
 carries us further than what can be achieved by 
 experiment on account of the larger scale in which the 
 phenomena present themselves. In our laboratories 
 we can reach and measure temperatures up to about 
 3000° C, but substances near the luminous surface of 
 the sun are at a temperature which is as high as 
 7000° C. and probably higher. Here nature gives us 
 the opportunity of extending experimental methods. 
 
I20 Lecture IV 
 
 Similarly, we are able, by experiment, to subject 
 matter to a pressure of a few thousand atmospheres, 
 but this is as nothing compared with the pressures 
 which must exist in the interior of the earth. Here 
 observation allows us to supplement our knowledge 
 of the properties of matter at high pressures, because 
 whenever an earthquake of sufficient intensity occurs, 
 waves pass through the interior of the earth and their 
 rate of propagation indicates the rigidity and com- 
 pressibility of its inner core. 
 
 The question of scale is often of paramount 
 importance, and more especially in questions which 
 affect our atmosphere. Though we believe we know 
 something about the formation of clouds, we should 
 get to know vastly more, if we could arrange 
 experiments on a scale sufficiently large, to form an 
 artificial cloud under conditions similar to those which 
 hold in the actual case. But scale is also important 
 as regards time, and the extent to which our scientific 
 activity and our scientific judgment are affected by 
 the duration of our lives has never been sufficiently 
 recognised. We are by nature individualists and 
 consequently wish to see the results of our own work. 
 If we can substantially improve an investigation by an 
 additional week's labour, most of us would do so, but 
 if the same improvement would require ten more years, 
 the policy of waiting would not commend itself to the 
 same extent. Similarly, when we have followed the 
 course of an experiment for a few months we are 
 proud of our patience, and few of us only would 
 undertake one extending over several years. When 
 centuries are required, we give up the attempt ; yet 
 
Effect of Scale 1 2 1 
 
 how grateful should we not be, if the old Egyptians, 
 six or seven thousand years ago, had sealed up in 
 air-tight vessels different substances, all carefully com- 
 pared as regards their weight, so that we could now 
 judge whether in the interval they had preserved the 
 same relationship of mass. Our ancestors have done 
 nothing for us, should we, having felt the want, not 
 do somethinof for our descendants ? 
 
 I have spoken of the scale of time as affecting our 
 scientific judgment, and perhaps I ought to explain 
 my meaning. If the sun shines on three successive 
 Fridays with cloudy weather in between, we should 
 laugh at any man bold enough to predict on the strength 
 of this observation that the sun will always shine on 
 Fridays : yet supposing that in the year at which 
 sun-spots are at their maximum, which occurs at 
 intervals of about eleven years, the monsoons were 
 exceptionally large or exceptionally small on three 
 successive occasions, how many of us would resist 
 the temptation to believe that there is a true connexion 
 between sun-spots and the weather .■* What is the 
 difference in the two cases ? To my mind it is only 
 the relationship of the period to the duration of our 
 own lives. In both cases, the mathematical probability 
 of an accidental coincidence is the same, but with the 
 weekly occurrence the true nature of the coincidence 
 is bound to be revealed within a short time, and the 
 risk of being wrong is more deeply impressed on our 
 minds, because a few weeks would be sufficient to find 
 us out. In the case of sun-spots on the other hand, a 
 decisive test can only be obtained after eleven, and 
 possibly twenty-two or thirty-three years; and the length 
 
122 Lecture IV 
 
 of time which must lapse before we can be shewn to 
 be wrong- in assertins: a true connexion, increases the 
 confidence with which we assert it. I am far from 
 intending to throw ridicule on those who in attempting 
 to relate definite long period phenomena base their 
 reasoning on comparatively few data. We are bound 
 to do so. The progress of science is achieved by 
 individual reasoning, which cannot be transferred from 
 one brain to another ; and the average span of a life- 
 time must therefore be an important factor in the rate 
 at which science progresses. This is a tempting 
 subject to dilate upon, but I must refrain from further 
 moralisino;. 
 
 The science of Terrestrial Magnetism began with 
 Gilbert, who taught us that the earth as a whole behaved 
 as if it were a magnet. If you imagine a large sphere 
 of magnetised steel, and investigate the magnetic field 
 in its neighbourhood by means of a small compass 
 needle placed at different points on its surface, this 
 needle would, roughly speaking, point towards the 
 magnetic poles of the sphere. We call a " north " 
 pole, a magnetic pole, which tends towards a point not 
 far removed from the north geographical pole of the 
 earth ; but, as you know, the north pole of one magnet 
 attracts the south and repels the north pole of another 
 magnet. If therefore the "north" pole of a compass 
 needle is attracted towards the north geographical pole 
 of the earth, this must mean that the earth's magnetic 
 pole which is situated in the northern hemisphere, 
 must be of the same kind as the pole of our compass 
 needle which we call the "south" pole. To avoid 
 the confusion which may arise from forgetting that 
 
Terrestrial Magnetism 123 
 
 geographical north means magnetic south, Gilbert 
 took the bold step of calling that end of a compass 
 which points north, the "south" pole of the magnet, 
 and British men of science followed that practice until 
 Faraday introduced the term "north seeking" pole, 
 and this has been gradually replaced by " north " pole, 
 thus bringing our nomenclature into accordance with 
 continental practice. Unfortunately at the same time 
 that this change took place in England the opposite 
 change was made in France, so that now Gilbert's 
 nomenclature is very generally adopted in that country, 
 and uniformity has again been disturbed. 
 
 The earth possesses to a great extent that simple 
 form of magnetisation, which is technically called 
 "uniform" ; consequently the forces on the surface of 
 the earth are nearly the same as those which would 
 hold, if there were a small magnet at the centre of the 
 earth possessing one north and one south pole. The 
 lines of force, which at each point indicate the direction 
 in which a magnetic pole would be driven, are shewn 
 in Fig. 10, the arrow-head pointing in the direction of 
 the force acting on the north pole. We have no means 
 of knowing how they run inside the earth, but outside 
 and near its surface, the agreement of the actual 
 magnetic forces with those of a small magnet at 
 the centre holds approximately, though not perfectly. 
 The line joining the north and south poles of our 
 small imaginary magnet is called the magnetic axis 
 and is shewn in the figure. If a small compass 
 needle, pivoted so as to be able to turn freely in all 
 directions, were placed at the places where the mag- 
 netic axis cuts the surface of the earth, it would 
 
124 
 
 Lecture IV 
 
 point vertically downwards, and these places would 
 then be what we call the magnetic poles of the earth. 
 From the scientific point of view these magnetic poles 
 defined as being the spots where a magnetic needle 
 points vertically downwards, are of no importance, 
 because as is easy to see, their position may greatly 
 
 Fig. lo. 
 
 be affected by magnetic rocks, placed accidentally 
 near P^ or P„. What is of paramount importance 
 in describing the magnetic state of the earth is the 
 direction of the magnetic axis, which may be defined 
 in a strictly scientific manner, independently of the 
 position of the poles. 
 
 This magnetic axis is observed to shift slowly, and 
 our observations, so far as they reach, are consistent 
 with a slow turning of the magnetic round the geo- 
 graphical axis in the direction from east to west. 
 The change of the magnetic forces accompanying the 
 shift of the axis is called the secular variation of 
 terrestrial magnetism. Its study forms one of the 
 
Secular Variation 125 
 
 most interesting branches of the subject and one of 
 the most difficult, because accurate measurements at a 
 sufficient number of stations only date back a century. 
 The magnetic poles must shift with the magnetic axis, 
 but they probably do not keep pace with it in a regular 
 manner. If indeed the actual position of the poles 
 is affected by magnetic rocks, we can imagine a 
 considerable shift of the magnetic axis to take place 
 without a corresponding change of the magnetic pole, 
 which, when the axis is displaced sufficiently, may 
 then follow with a rush. Hence we should attach 
 but little importance to the exact determination of 
 the magnetic poles in arctic and antarctic expeditions. 
 What is important to the general theory of terrestrial 
 magnetism is the accurate determination of magnetic 
 elements in the region of the poles without undue 
 regard to one particular point. 
 
 Our magnetic observations are practically confined 
 to the surface of the earth, but it is remarkable how 
 much information we may obtain from these observa- 
 tions, by applying theoretical considerations derived 
 from a general knowledge of magnetic and electro- 
 magnetic forces. The general principle of the con- 
 servation of energy which holds universally, allows us 
 to conclude that when a small magnet is shifted from 
 one place to another, the work done can only depend 
 on the initial and final positions, but not on the path 
 taken during the displacement. When the magnet is 
 moved in a closed path so that the initial and final 
 positions are coincident, the total work must be zero. 
 Can we apply this principle also to the displacement 
 of a single pole ? So far as the theory of energy is 
 
126 Lecture IV 
 
 concerned we must answer " No," because it is not 
 allowable to go beyond what can be tested by experi- 
 ment, and it is not possible to divide a magnet so that 
 its two poles can be moved separately. But in the 
 general analysis of magnetic forces, we resolve the 
 total effect on finite magnets into elementary forces, 
 between two poles, or between one pole and an 
 electric current. The complete agreement between 
 the derived and observed forces, justifies us in using 
 these elementary forces as stepping stones. We are 
 thus able to assert, that the work done in allowing a 
 pole to describe a closed circuit is zero, when the 
 magnetic forces are all derived from magnets, but not 
 necessarily so when they are due to electric currents. 
 If AB (Fig. ii) represents a portion of an electric 
 current, and a magnetic pole at P 
 be moved round the current in a 
 closed path (e.g. PHKLMP or 
 PKLMP) until it is brought back 
 to P, the total work done is not zero, 
 but a definite quantity proportional 
 to the total electric current which 
 has been enclosed by the path. If 
 the path (e.g. PHKP) does not 
 include the current, the work is zero. This is im- 
 portant, because it allows us to decide by experiments 
 made on the surface of the earth, whether electric 
 currents of sufficient intensity traverse its surface. 
 If we take the observations now available and apply 
 the test to a circle of latitude, we should find indeed 
 that these observations can only be reconciled, if we 
 admit a .substantial vertical electric current. On the 
 
 B 
 
 Fig. II. 
 
Existence of Potential 1 27 
 
 other hand, when we apply the test to regions of the 
 earth, which have been surveyed with great care, such 
 as the British Isles (Riicker and Thorpe), we find that 
 they do not indicate the existence of such vertical 
 currents. There is therefore a strong- suspicion that 
 in the former case, the observations extending over 
 regions including the ocean basins are not sufficiently 
 accurate to allow any certain conclusions to be drawn, 
 and this suspicion is strengthened by other lines of 
 reasoning based on the direct observations of the 
 electric currents traversing the earth's surface, which 
 shew that these currents are much too weak sensibly 
 to affect our magnetic instruments. It is not always 
 recognised that what the magnetic observations give 
 us is the current which actually passes from the inside 
 to the outside of the earth. These currents are 
 sometimes spoken of simply, as "vertical currents," 
 and the practical man who takes his theory from 
 "hearsay," frequently believes that he can reconcile 
 electric and magnetic observations by imagining 
 vertical electric currents entirely outside the earth, 
 ascending in one region, descending in another and 
 completed horizontally near the earth's surface, with- 
 out actually cutting it. Such explanations deserve no 
 consideration. For the present we must leave the 
 question of vertical currents open until the general 
 magnetic survey of the earth, undertaken by the 
 Carnegie Institution of Washington, has furnished us 
 with more accurate determinations in the oceanic 
 regions. 
 
 Gauss, to whom after Gilbert we owe the ground- 
 work of the theory of Terrestrial Magnetism, has 
 
128 
 
 Lecture IV 
 
 taught us how observations, made on the surface of 
 the earth may be used to decide whether and in what 
 proportion the observed magnetic forces are due to 
 causes lying inside or outside the surface. Neglecting 
 the small effect possibly due to electric currents 
 traversing the surface, it is always possible to imagine 
 two independent magnetic systems which can both 
 equally well explain the horizontal forces. Of these, 
 one lies inside and the other outside the surface. This 
 means that if we only observe magnetic needles swung, 
 like a compass, so as to confine their movement to a 
 horizontal plane, we can explain the observed de- 
 flexions equally well by forces coming to us from 
 outside and by forces which have their seat inside 
 the earth. But the vertical components of the forces 
 would be quite distinct in the two cases, being in 
 general in opposite directions. If for instance the 
 earth were placed in a magnetic field produced by 
 external magnets in such a 
 way that the magnetic lines 
 of force were all parallel as 
 in Fig. 12, the horizontal 
 forces on the surface of the 
 earth would exactly be the 
 same as those caused by a 
 small magnet of proper in- 
 tensity placed inside the 
 earth, as illustrated in Fig. 
 JO. In order to make the 
 north pole at the equator 
 point northwards, it would, 
 in the case now considered, be necessary to make the 
 
 Fig. 12. 
 
Internal and External Causes 129 
 
 lines of force run in the direction indicated by the 
 arrows in Fig. 12, which means that at P^ the north 
 pole would point upwards and not downwards as is 
 actually observed. Hence, while the horizontal forces 
 are the same in the two cases considered, the vertical 
 forces are in opposite directions, and this general 
 reasoning is quite sufficient to prove that the main cause 
 of the earth's magnetism lies inside. If however we 
 want to decide whether any portion of it comes to us 
 from outside, we must replace the general reasoning 
 by an exact and troublesome calculation. It is then 
 found that our present observations, assuming them to 
 be correct, indicate an effect amounting to about 5 "/„ ot 
 the total having its origin outside the earth's surface ; 
 but here again one is bound to speak with caution until 
 the Carnegie Institution has furnished us with accurate 
 figures. Before leaving this portion of the subject, I 
 must however insist on a result of Gauss' theory which 
 is not always sufficiently recognised. There can only 
 be one solution of the problem, and if we can explain 
 any magnetic effect on the earth's surface by outside 
 forces, it follows that it can not at the same time be 
 explained by internal forces. This remark disposes 
 of a good deal of the criticism lavished on pioneer 
 attempts to open out this region of science. Though 
 this criticism is often confined to a judicious shrugging 
 of the shoulders, it stops scientific progress more 
 effectually than active opposition, and is apt to become 
 a constitutional habit with those who give way to its 
 undoubted temptations. 
 
 I have hitherto spoken of magnetic forces as 
 invariable at every point, except for the slow secular 
 
 s. 9 
 
130 Lecture IV 
 
 variations. In reality when a magnetic needle is 
 delicately suspended it is never at rest, shewing that 
 the earth's forces are always changing both in magnitude 
 and direction. These changes are partly regular and 
 partly irregular. Among the regular variations, that 
 depending on the time of day is the most important. 
 We call it the diurnal variation. It is not my intention 
 to describe this feature of terrestrial magnetism in 
 detail, but I should like to explain what we have 
 learned of its origin. The greater portion of it is 
 undoubtedly due to electric currents situated in the 
 upper portions of the atmosphere. That the cause 
 lies mainly above the surface can be proved, and that 
 it lies in the atmosphere must be inferred as probable. 
 The discussion of the problem which I gave some 
 years ago, shewed that there is a' substantial remnant 
 which comes to us from the inside of the earth, and 
 this is, as it should be, because the earth being a 
 conductor, any change of the magnetic forces must 
 give rise to induced internal currents. The intensity 
 and phase of these internal currents allow us to draw 
 interesting conclusions on the conductivity of the earth, 
 but the observational data used were probably not 
 sufficient to do more than indicate that the deeper 
 layers must be more highly conducting than the outer 
 shell, which is in agreement with the conclusion de- 
 rived from other considerations. 
 
 If we admit that the primary cause of the diurnal 
 variation is electro-magnetic, we may easily formulate 
 that system of electric currents in the atmosphere, 
 which gives the required magnetic forces. In this 
 system the currents are everywhere parallel to the 
 
Diurnal Variation 131 
 
 earth's surface and the question now arises whether 
 we can explain them plausibly. The first attempt in 
 this direction was made by Balfour Stewart, who forty 
 years ago suggested that any motion of the atmosphere 
 in which the particles of air are driven across the 
 earth's lines of magnetic force would set up an 
 electro-motive force. One great difficulty however 
 stood in the way of this hypothesis : so far as was 
 known at the time, air was a non-conductor requiring 
 a considerable electro-motive force, to start a discharge 
 and this discharge, when it did take place, was dis- 
 ruptive and discontinuous like a flash of lightning 
 in a thunderstorm. It was only after I had shewn, 
 that gases can by various means, now called pro- 
 cesses of ionization, be put into a sensitive state in 
 which they behave like ordinary conductors, that the 
 main obstacle to an acceptance of Balfour Stewart's 
 theory was removed. I have recently^ worked out 
 this theory so as to obtain numerical results which 
 may be tested by observation. 
 
 If the magnetic effect be due to electro-magnetically 
 induced currents, it should be connected directly with 
 some diurnal change in the flow of air, and if we 
 investigate, what must be the flow of air, which 
 reproduce the magnetic effect, we find that it is exactly 
 of the kind required to cause the observed diurnal 
 change of the barometer. We are therefore led to 
 inquire whether one and the same oscillation of the 
 shell of air which surrounds us is sufficient to explain, 
 on the one hand the variations of atmospheric pressure, 
 
 ■■,..,: . .',„ . " Phil. Trans, ccvni., p. 163 (1908). 
 
 9—- 
 
132 Lecture IV 
 
 and on the other the oscillations of the magnetic 
 needle. 
 
 The pressure change is of two kinds, one of them 
 reaching its highest value once a day and having 
 therefore a period of 24 hours, while the second and 
 more important one comes to a maximum twice in 
 the day — like the ocean tides — having a period of 1 2 
 hours. It may at first sight seem curious, that, while 
 the varying altitude of the sun, the main outside 
 influence which can cause a variation in the height 
 of the barometer, is diurnal, the oscillation of the 
 atmosphere gives such a large preponderance to the 
 12 hourly change. It must be remembered, however, 
 that the amplitude of an oscillation does not entirely 
 depend on the intensity of the periodic force causing 
 it, but also on the degree of agreement between the 
 natural period of the oscillating system and that of 
 the acting force. We may compare the oscillation 
 of the atmosphere with that of a resonator excited 
 by an outside sound. The resonator will, as a rule, 
 have several independent periods in which it can 
 freely vibrate, and ordinarily when forcibly excited its 
 note will contain a number of these free periods, 
 though one of them, the fundamental note, will pre- 
 dominate. But when forced to resonate by an outside 
 sound coinciding in period with one of its free periods, 
 it will respond to that period alone, the fundamental 
 not appearing at all. It was therefore suggested first, 
 I believe, by Lord Kelvin and then worked out more 
 in detail by Lord Rayleigh, that the natural period 
 of the shell of air which surrounds the earth may be 
 nearly equal to 12 hours, and in that case the daily 
 
Diurnal Variation 133 
 
 change of temperature, which contains a semi-diurnal 
 variation, though its predominant period is diurnal, 
 might set up an oscillation of 12 hours, exceeding 
 in amplitude that of the purely diurnal oscillation. 
 Calculations of the natural period of the free oscillations 
 of the atmosphere, in which the effect of the rotation 
 of the earth is taken account of, have justified this 
 explanation. 
 
 If we calculate the motion of air, which is re- 
 quired to account for the amplitudes of the 24 hourly 
 (diurnal) and 12 hourly (semi-diurnal) variation of the 
 barometer, we find that, by properly adjusting the 
 electric conductivity each of them may be made to 
 explain satisfactorily the corresponding variation of 
 the magnetic needle, but if we fix on any definite 
 conductivity, we find too great an amplitude for 
 the diurnal or too small an amplitude for the 
 semi-diurnal magnetic change. We also find that 
 the magnetic variation is much more affected by the 
 season of the year than the barometric change. A 
 better agreement may be obtained by making the 
 plausible hypothesis, that the electric conductivity of 
 the higher regions of the atmosphere is due to solar 
 radiation, and is therefore greater in summer than in 
 winter and also greater in day time than at night. As 
 regards the absolute value of the conductivity, it is 
 found that it must be about ten times as great as that 
 which is found in the flame of a Bunsen burner. This 
 is greater than perhaps we have any right to expect, 
 though we know that for equal ionizing powers, the 
 conductivity should increase with altitude. Theory 
 would lead us to expect the conductivity to be inversely 
 
134 Lecture IV 
 
 proportional to the pressure, but observations made 
 during balloon ascents by Gerdien have led to an in- 
 crease of conductivity at an altitude of six kilometres to 
 more than ten times its normal value near the surface, 
 while the pressure was only reduced to one half. 
 Without wishing to lay too great a stress on the 
 results which I have brought to your notice, they are 
 sufficient to shew that a careful study of terrestrial 
 magnetism may allow us to draw important conclusions 
 on the state of the upper layers of our atmosphere. 
 
 While the normal magnetic changes which go on 
 from day to day have received comparatively little 
 attention, more interest has been shewn in the larger, 
 more sudden and more irregular changes which are 
 known as magnetic storms. This is no doubt due to 
 the remarkable connexion between the frequency of 
 their occurrence and the number of sun-spots which are 
 found on the solar disc. The storms occur principally 
 when the spots are numerous, though no certain con- 
 nexion between a particular spot and a particular 
 disturbance has yet been traced. Attempts have been 
 made to explain the connexion, but at present they 
 are a little more than more or less plausible guesses. 
 It is, however, certain that we must reject the possi- 
 bility of any direct magnetic action between the sun 
 and the earth. Explanations based on such a direct 
 effect which were never probable, were finally slain 
 by Lord Kelvin, who shewed that the active portions 
 of the sun's surface, if directly responsible for a mag- 
 netic storm of moderate intensity, would emit nearly 
 four hundred times as much energy as they do in the 
 regular supply of heat and light. 
 
Magnetic Storms 135 
 
 The popular view, at the present moment, is to 
 ascribe magnetic storms to the emission by the sun 
 of electrified particles which, passing with enormous 
 velocity by or through our atmosphere, affect our 
 magnetic needles and produce the storm. Neither 
 qualitatively nor quantitatively does it seem to me 
 that this explanation is probable, though the more 
 refined form in which it has been presented by Pro- 
 fessor Birkeland overcomes some of the difficulties 
 which previously seemed fatal. Lord Kelvin's ob- 
 jection which, mutatis mutandis, seems to me to apply 
 to this case also, must, however, for the present block 
 the way to any theory which makes the sun responsible 
 for the energy of magnetic storms. In a discussion 
 of the subject', presented to the Royal Astronomical 
 Society of London, I arrived at the conclusion that 
 the store of energy residing in the earth's rotating 
 mass is the only one sufficiently great to provide for 
 the energy of the magnetic storm. It is large enough 
 to do so without being sensibly affected, for after a 
 million years, the total loss through magnetic storms 
 would diminish the rotational velocity so little that, as 
 a time-keeper, the earth would only lose one second 
 per year. 
 
 I should like to explain briefly the view regarding 
 the solar influence on terrestrial magnetism, which 
 I have advocated for some time. I believe it to be 
 purely an indirect one, affecting the conductivity of the 
 upper regions of our atmosphere, either by direct 
 radiation or by the injection of particles which ionize 
 the air through impact. The increase in conductivity, 
 ' Monthly Notices. R.A.S. Lxv., p. i86 (1905). 
 
136 Lechire IV 
 
 in conjunction with electro-motive forces which are 
 always present, are, if I am right, sufficient to account 
 for the observed effects. 
 
 What are these electro-motive forces ? We have 
 already discussed those due to periodic displacements 
 in the atmosphere, which are responsible for the 
 diurnal variation. There is further, that due to the 
 general atmospheric drift which is such as to equalise 
 angular momentum in different parts of the atmo- 
 sphere, or from east to west in the equatorial and from 
 west to east in the polar regions. This drift sets up 
 an electro-motive force tending to drive positive 
 electricity towards the equator in the polar regions, 
 and in the opposite direction in the equatorial regions. 
 If the direction of motion along circles of latitude be 
 alone considered, and the conductivity be uniform, the 
 electro-motive force can be equalised by a redistribution 
 of electro-static force, and there is no electric current, 
 but if, suddenly, the conductivity be increased on one 
 side of the earth, a magnetic storm results which would 
 present in character marked similarities to that of a 
 certain type of magnetic storms. 
 
 Finally, there is one effect of electro-magnetic in- 
 duction deserving more careful study, which has its seat 
 at the extreme outer limits of the atmosphere. It is 
 difficult to say what the nature of the induced currents 
 should be, because we do not know how our atmo- 
 sphere ends. If it comes to an abrupt conclusion, and 
 if the surrounding parts of .space are conducting, as 
 no doubt they are, induced electric currents would 
 pass under the influence of the rotating and magnetic 
 earth from the atmosphere into space, and back again. 
 
Causes of Disturbances 137 
 
 If, on the other hand, there is a region on the out- 
 skirts of the atmosphere whicli circulates with planetary 
 velocities, the electro-magnetic induction would, roughly 
 speaking", be in the opposite direction to that holding 
 in the previous case. If this electro-magnetic induction 
 be the cause of the aurora borealis, a more detailed 
 study may decide between the alternatives, as to rest 
 on orbital circulation of matter in space. 
 
 As the magnitude and irregularity of all our 
 magnetic effects are exaggerated at times of maximum 
 sun-spots, we conclude that at these times the upper 
 atmosphere and probably also interplanetary space 
 has an increased electric conductivity. Whether this 
 increased conductivity of space is the cause of the 
 solar disturbances which shew themselves as spots or 
 their consequence, is a matter on which we are free to 
 speculate. At any rate, if there is a variable con- 
 ductivity of interplanetary space, it is to the magnetic 
 needle that we must look for information, and the 
 science of terrestrial magnetism may be expected to 
 throw much light on whatever corpuscular matter may 
 be contained in interplanetary space as well as on its 
 possible relationship to the solar drift through space. 
 
 It has often been suggested that finely divided 
 matter is being projected outwards from the sun, 
 and the appearance of the solar corona which is a 
 luminous envelope of the sun, visible only when the 
 solar disc is covered by the moon in a total solar 
 eclipse, lends some countenance to that view. We 
 may imagine these projections to be due to electrical 
 repulsion, or to the peculiar repulsive force which 
 accompanies the incidence of any radiation on ab- 
 
138 Lecture IV 
 
 sorbing or scattering particles. This repulsive force 
 was first brought to our notice by Clerk Maxwell, 
 who deduced it as a theoretical consequence of his 
 electro-magnetic theory of light. Owing to the small- 
 ness of the effect, it seemed hopeless, at the time, to 
 verify it experimentally, and the theoretical calculation 
 was not convincing to so high an authority as Lord 
 Kelvin. But, as in so many cases, Maxwell proved 
 to be right, and we have now the experiments of 
 Nicols and Hull, those of Lebedew and those of 
 Poynting, all shewing that the propagation of radiant 
 energy may be considered to carry a momentum 
 along with it. Whenever this energy is deviated 
 from its straight path, or absorbed or scattered in any 
 way, the law of conservation of momentum gives us 
 at once the magnitude and direction of the resultant 
 repulsion, which acts whether the body on which the 
 radiation falls be large or small, and must even affect 
 the individual molecules of a gas. If the effects are 
 observed chiefly in the case of solid and liquid 
 particles, it is because the same quantity of matter 
 absorbs more in the solid and liquid states than in the 
 gaseous form. If we had a mixture of gases, a radia- 
 tion transmitted through it would have a tendency to 
 separate the gases according to their absorbing power, 
 but it is questionable whether the effect would ever 
 become appreciable. Absorption of the more refrang- 
 ible rays would in general count less than absorption 
 in the infra-red, because most of our sources of light 
 and heat emit their greatest energy in the region of 
 low frequency. The repulsion by radiation fills an 
 important function in interplanetary space, explaining 
 
Ejection of Electrons from Sun 1 39 
 
 possibly the appearance of comet tails and of the 
 solar corona. It has further to be taken into con- 
 sideration when discussing the stratified arrangement 
 of substances in the solar atmosphere, but in the 
 absence of numerical data the danger of being led 
 astray is great. 
 
 Those who believe that the ejection of negative 
 corpuscles explains the connexion between solar and 
 planetary phenomena, require an occasional reminder 
 that a body cannot indefinitely emit negative electricity 
 without becoming charged positively, and this excess 
 of positive electricity must be disposed of. If the 
 sun emits negative electrons, the electric forces in its 
 neighbourhood must be such, that there is a repulsion 
 of positive electricity outwards. We should therefore 
 expect a perhaps slower but equally effective discharge 
 of positive ions, to accompany the emission of electrons. 
 
 Nothing has been said as yet on the explanation 
 of the secular variation of terrestrial magnetism, and 
 in our ignorance of the causes, which make the earth 
 behave Hke a magnet, it is perhaps wisest to put the 
 question aside for the present. But sometimes a hint 
 given by the accessories of a fundamental phenomenon, 
 throws light on the phenomenon itself, and it is 
 therefore permissible to consider the alternatives which 
 our present knowledge places before us. A body may 
 act as a magnet, either because it contains magnetised 
 iron or because electric currents circulate in it. The 
 possibility that terrestrial magnetism may be explained 
 by masses of magnetic iron is sometimes put aside as 
 impossible on account of the high temperature of its 
 interior. For iron loses its power of being permanently 
 
140 Lecture IV 
 
 magnetised at about 800' C, and this temperature 
 will be reached at comparatively moderate depths 
 below the surface. This argument neglects the possible 
 effects of pressure, which may considerably raise the 
 critical temperature at which iron loses its magnetism. 
 I have had made in the Physical Laboratory of the 
 University of Manchester a number of experiments, 
 in order to test whether an increase of 600 or 700 
 atmospheres had an effect on the magnetic properties 
 of iron near its critical temperature. The experi- 
 ments, which present very great difficulties, are not 
 yet concluded and I am unable to give any results. 
 If we reject the view that iron is the cause of 
 terrestrial magnetism, we are reduced to ascribe them 
 to electric currents, unless we wish to bring some 
 altogether new phenomenon into action. The objection 
 which might be raised, that in the absence of perma- 
 nent electro-motive forces such electric currents would 
 not be permanent, but die out in consequence of the 
 resistance of the earth. Is not as serious as it looks, 
 because Professor Horace Lamb has calculated that 
 a system of electric currents such as is necessary to 
 give to the earth the field of a uniformly magnetised 
 sphere would require several million years before it 
 is reduced to half its value. If electric currents are 
 the cause of the earth's magnetic state, the currents 
 producing it must have been started by some unknown 
 cause, and their effect must die out gradually. It is 
 perhaps interesting to note that a system of electric 
 currents circulating in a rotating body about an axis 
 inclined to the axis of rotation, as the earth's magnetic 
 axis is inclined to the geographical axis would be 
 
Cause of Secular Variation 1 4 1 
 
 subject to a secular variation such as is observed ; 
 but unfortunately this secular variation would be far 
 too slow to account for the actual phenomenon. 
 
 Some years ago I made an attempt to explain the 
 secular variation on the view that the earth contains 
 magnetic masses of iron. If we admit for the sake of 
 argument an inside core of the earth and an outside 
 shell, which though connected together allow a slow 
 relative displacement ; if we further take interplane- 
 tary space to be electrically conducting and the inner 
 core of the earth to be masfnetised alongf an axis 
 inclined to the axis of rotation, then magnetic forces 
 due to electro-magnetically induced currents in inter- 
 planetary space would act on the system in such a 
 way, that its magnetic axis would tend to become more 
 parallel to the axis of rotation, and at the same time 
 turn round it in the manner actually found in the secular 
 variation. 
 
 If, ultimately, we should have to reject both the 
 theory, that the earth's magnetism is clue to masses of 
 iron and that it is due to electric currents, we should 
 be compelled to look for some other cause. In 1892 
 I suggested, and the suggestion has I think also been 
 made by others, that every rotating body may behave 
 as a magnet, but we have at present no experimental 
 evidence for such a hypothesis, though it may be 
 defended on theoretical grounds. As a suggestive 
 argument in favour of the view, it may be mentioned, 
 that both the main facts of terrestrial magnetism and 
 its secular variation may be derived from the same 
 hypothesis on the molecular constitution. 
 
 I have already mentioned in speaking of the diurnal 
 
142 Lecture IV 
 
 variation that any change in the magnetic field must 
 be accompanied by induced electric currents in the 
 crust of the earth. These " Earth Currents " reveal 
 themselves whenever two metal plates are inserted into 
 the ground and connected through a galvanometer, 
 and at one time it was hoped that through them we 
 might be able to learn what part they play in the 
 observed magnetic disturbances. This hope has not 
 been fulfilled, partly because the distance between the 
 plates must be considerable, if errors are not to be 
 introduced by the action of the system as a primary 
 battery, and partly because the connexion between 
 the earth current and the current passing through the 
 galvanometer is not at all obvious, and depends on the 
 electric conductivity of the soil. It must also be borne 
 in mind, that while the magnetic effect depends on 
 currents probably reaching down to a considerable 
 depth, the observed earth currents are surface currents 
 depending both in magnitude and direction on the 
 distribution of electric conductivity in the uppermost 
 layers of the soil. 
 
 Without entering into any details on the methods 
 adopted for its experimental investigation, I will now 
 describe and discuss the effects which are due to the 
 electro-static charge which resides on the surface of the 
 earth. The earth as a whole is charged negatively, 
 and on the average we find that there must be nearly 
 a million electrons on every centimetre of its surface. 
 This negative charge is not constant, but shews 
 irregular as well as regular diurnal and seasonal 
 variations ; it may even be reversed in sign and reach 
 a high value, when the sky is clouded over. The 
 
The Earth negatively charged 143 
 
 electrification of the earth, as a whole, is difficult to 
 explain, especially when it is remembered that the 
 atmosphere conducts electricity, and that therefore 
 if the charge is permanently maintained it must be 
 continually reproduced, otherwise it would quickly be 
 dissipated. 
 
 The first question that arrests our attention is 
 whether the earth as a whole permanently discharges 
 negative electricity into space. This would be the 
 case if the electro-motive force due to the charge on 
 the surface reached the outer regions of the atmo- 
 sphere. Observations made in balloons shew that 
 the electro-static effect due to the charged earth is 
 continually diminishing, and at heights of between 
 4 and 6 kilometres, it is reduced to one-tenth of its 
 normal value at the surface. This means that the 
 effect of the negative charge of the earth at great 
 altitudes is counteracted by a nearly equal positive 
 volume charge of the atmosphere. If it is reasonable 
 to suppose that the compensation is complete in the 
 uppermost layers, we are justified in believing that 
 the primary cause of the earth's electrification is 
 atmospheric and not cosmical, and that there is no 
 dissipation into space. Though personally I share 
 this view, I am bound to point out that the above 
 reasoning is by no means as conclusive as it appears. 
 Imagine a charged sphere surrounded by a medium, 
 the conductivity of which is small near the surface 
 of the sphere, but increases outwards. As in the 
 steady state the quantities of electricity passing 
 through the successive layers of the spherical envelope 
 must be equal, it follows that the electro-motive force 
 
144 Lecture IV 
 
 must be continuously diminishing outwards and a low- 
 value of that force is therefore quite consistent, with a 
 permanent dissipation into space of all the electricity 
 which passes from the earth into the atmosphere. In 
 order to prove that there is no such dissipation, it 
 would be necessary to shew that the electro-motive 
 force diminishes with altitude more rapidly than the 
 conductivity increases, and our observations though 
 pointing in that direction are not yet sufficiently 
 precise to remove the possibility of doubt. 
 
 The conductivity of the air depends mainly on its 
 ionization, and a handy apparatus has been constructed 
 by Ebert to count at any time or place the number of 
 ions present. This number amounts on the average 
 to about looo per cubic centimetre, there being 
 generally an excess of positive ions present. It 
 follows indeed from what has been said above, that an 
 excess of negative ions would indicate either that the 
 charge of the earth at the point of observation is 
 reversed, or that the atmospheric conditions are such 
 that the conductivity diminishes upwards instead of 
 increasing as it usually does. As the negative and 
 positive ions are continually recombining, a permanent 
 ionizing agent must be at work. Near the surface of 
 the earth an effective ionizer is supplied by the radio- 
 active products escaping from the earth. At high 
 altitudes there may be additional cosmic causes such 
 as the ultra-violet light from the sun, but it is difficult 
 to believe that these can penetrate through 76 centi- 
 metres of mercury, which represents the mass of the 
 atmospheric layer above us. 
 
 The rate at which ions recombine varies consider- 
 
Ionization of Air 145 
 
 ably. Dust or mist accelerate combination, so that 
 the number of ions, and hence the electric conductivity, 
 depends on a balance between the rates at which they 
 are generated and disappear. It is in agreement with 
 this argument that mist increases the observed electric 
 force, because it helps recombination and therefore 
 diminishes the conductivity. Similarly Elster and 
 Geitel, to whom together with other important contri- 
 butions to the subject, we owe the first decisive proof 
 that atmospheric air is conducting, have shewn that 
 when the air is very transparent the electrification is 
 considerably diminished, doubtlessly owing to the 
 absence of dust with its recombining power. 
 
 It may, at first sight, appear strange that the elec- 
 trification is not more uniform over the earth, and may 
 even differ considerably at places which are near to 
 each other. If we were to neglect the atmospheric 
 charges, we should expect indeed the conductivity of 
 the earth to be sufficient to establish equilibrium with- 
 in a few seconds ; that this is not the case is itself a proof 
 that the positive electrification neutralising the earth's 
 negative charge resides at no great distance from the 
 surface. This positive charge being bound to matter 
 which can only move slowly, keeps the lines of force 
 in position and thus prevents a uniform distribution 
 from establishing itself quickly. 
 
 Regular observations from which the electrifica- 
 tion of the earth may be deduced are made in many 
 observatories, and the number of ions in the atmo- 
 sphere are also being regularly counted in a few 
 places. The diurnal and annual changes of these 
 elements shew many features of interest, but the 
 
 S. 1 n 
 
146 Lecture IV 
 
 methods used are not well adapted to present the 
 phenomena in their clearest light ; for, as has already 
 been pointed out, the ionization of the air is not itself 
 a primary phenomenon. An increased ionization may 
 either mean that the ions are generated at a quicker 
 rate, or that they recombine more slowly, and we 
 shall be unable to interpret the observations which 
 have accumulated in various observatories until these 
 two factors are separated and we trace independently 
 the diurnal changes in both of them. 
 
 The fundamental and, at the same time, the most 
 difficult of all the problems of atmospheric electricity 
 is to discover the origin of the negative charge which 
 covers the earth. This electrification is constantly 
 being dissipated into the air, at a rate sufficient to 
 reduce the total charge to half its value in the 
 course of four or five minutes, if some counteracting 
 agent were not at work. What is the agent which 
 continuously drives negative electricity into the earth 
 from outside ? We might think of some surface action, 
 or, going further afield, of some cosmical cause ; finally, 
 we might take an intermediate view and look to the 
 atmospheric layer of moderate altitudes as the primary 
 separator. All three views find their advocates. The 
 cosmic theory is rendered very improbable by the 
 balloon observations, and such contact theories, as have 
 hitherto been proposed, are mainly based on the 
 observed fact that in a closed vessel containing ionized 
 air the inner surface has a tendency to take up a 
 negative charge, owing to the adsorption of negative 
 ions. Fatal objections have been urged against this 
 being a sufficient cause to explain the negative charge 
 on the earth. 
 
Origin of Atmospheric Electricity 147 
 
 Other explanations having failed, the most promising 
 theory seemed to be that which made the rainfall re- 
 sponsible for the charge. A theory put forward by 
 Dr C. T. R. Wilson, met therefore with a very favour- 
 able reception. This theory is based on Dr Wilson's 
 discovery, mentioned in a previous lecture, that in 
 supersaturated air water condenses more readily on 
 negative than on positive ions. When an upward 
 current of moist air cools by expanding under the 
 reduced pressure, the first condensation takes place 
 on the particles of dust which act as nuclei, but it is 
 likely that the upward current passing through the 
 cloud which is thus formed, is not absolutely dry. 
 Further condensation, the dust-free air above the 
 cloud does not then begin until the air which is still 
 supposed to rise becomes supersaturated ; the drops 
 formed under these circumstances are negatively electri- 
 fied. If these drops collect and fall to the ground 
 they would bring their charge to earth, the neutralising 
 positive charge remaining in the air. Gravity here 
 acts as separator. An experimental confirmation of 
 this view would be obtained, if rain could be shewn to 
 bring down an excess of negative electricity. 
 
 Let us first form an estimate of the quantities 
 involved. Though the rainfall is only known accu- 
 rately over limited regions, we may form some rough 
 estimate of its total amount. The average all over 
 the earth is not likely to be less than 100 or more 
 than 200 cms. per year. If we put it at 150 cms. we 
 shall, so far as I can judge from the published maps 
 of rain distribution, somewhat underestimate the total 
 amount. We may note here, in passing, how the 
 
148 Lecture IV 
 
 same fact may be represented in two ways which 
 appeal very differently to the imagination. A rain- 
 fall of 1 50 cms. per year means an average deposit 
 of the twenty-thousandth part of a millimetre per 
 second, which seems small. But summing up all 
 over the surface of the earth it means that 26 million 
 tons of water fall down each second of time — which 
 creates quite a different impression. If we combine 
 together the observed loss of charge per second of 
 time with the quantity of water brought down, we find 
 that for equilibrium each cubic centimetre of rain would 
 on the average have to be charged with one-tenth of 
 an electro-static unit of electricity. 
 
 Observations on the electric charge of rain have 
 been first made by Messrs Elster and Geitel, and 
 whenever similar experiments have been repeated it 
 has been found that both positively and negatively 
 charged rain occurs, the average charge being about 
 one electro-static unit per cubic centimetre, which, if 
 the charge were always of the same sign, would be 
 ten times more than the quantity required to replenish 
 the earth. It follows that such theories as Wilson's 
 can only be tested by continued series of observations, 
 determining the average excess of one charge over 
 another. The conclusion arrived at by Elster and 
 Geitel and some other observers, was that there is 
 indeed an excess of negative charge. 
 
 In opposition to this are the more recent experi- 
 ments made by Dr George Simpson', who found that 
 
 ' The passage regarding Dr Simpson's experiments {Phil. Trans. 
 ccix., p. 319) has been added in writing out these lectures (June, 
 19 10). The experiments were made after the lectures were delivered. 
 
Electric Charge of Rain 149 
 
 during the monsoon the rain which falls at Simla 
 is more often electrified positively than negatively. 
 During the period investigated the total quantity of 
 rain which fell was 76 cubic centimetres for each 
 square centimetre of surface, bringing down 2 2 '3 
 electro-static units of positive electricity and 7 '6 of 
 negative electricity ; there was therefore an excess of 
 ■2 units of positive electricity for each cubic centimetre 
 that fell. These experiments seem to have been 
 confirmed in other localities, and should they be found 
 to supersede the earlier experiments of Elster and 
 Geitel, we should have to abandon what at one time 
 appeared to be the probable source of the earth's 
 electric charge. But, as at present the rainfall has 
 been investigated at only a few places on the earth's 
 surface, we must not too hastily reject the possibility 
 that the total balance of electrification is negative 
 although it must be conceded that the evidence at 
 
 o 
 
 present inclines towards the opposite view. 
 
 The deadlock which for the present seems to bar 
 our ability to explain the fundamental phenomena of 
 atmospheric electricity, renders it advisable to look in 
 all directions for any cause which might tend to elec- 
 trify the air positively as compared to the earth. 
 
 Professor Ebert has sus^gested that the radio-active 
 and therefore ionized gases, which are known to escape 
 from the soil, precipitate some of their negative charges 
 while filtering through the surface layers, the com- 
 plementary positive ions being discharged into the 
 atmosphere. The effect is a real one, as experiments 
 have shown that ionized gases driven through narrow 
 channels deposit negative charges on the walls of the 
 passages until a definite positive charge is imparted to 
 
150 Lecture IV 
 
 the escaping gas. There are, however, difficulties in 
 the way of accepting Ebert's theory as sufficient to 
 account for the magnitude of the earth's charge. 
 These have been mainly urged by Dr Simpson, whose 
 observations shew that an increased amount of radio- 
 active gas in the air is, in general, accompanied by a 
 diminution of the earth's charg-e as derived from the 
 potential-gradient. This diminution is obviously due 
 to the increased conductivity of the air. It would 
 seem, therefore, that while admitting the reality of the 
 Ebert effect, the equilibrium point would be reached 
 with a lower electrification than that which is observed. 
 But there are other effects tending in the same direction. 
 Such is the spraying of sea water which, according to 
 Lenard, imparts a positive charge to the atmosphere in 
 which the spraying takes place. I am not aware of 
 any observations carried on to determine the electrical 
 condition of the atmosphere over the ocean, when its 
 wave crests break into drops, but it is quite possible 
 that the earth's electrification may be partly accounted 
 for by the Lenard effect. The objection which may 
 be raised against this and Ebert's theory, that it only 
 accounts for the positive electrification of the air in the 
 immediate neighbourhood of the ground, is only a 
 superficial one, because the ordinary processes of 
 diffusion are sufficient to cause the electrification to 
 spread in time through the whole atmosphere. 
 
 It is not possible at present to estimate with any 
 degree of accuracy the part played by lightning dis- 
 charges in the electrical economy of the earth. 
 F. Pockels' has determined in two cases the ap- 
 proximate maximum current in a flash of lightning, 
 ' Met. Zeitschrift, 1901, p. 40. 
 
Cause of Atmospheric Electricity 151 
 
 and found it to be 10 and 20 thousand amepres 
 respectively. If we take the corresponding duration 
 to be a thousandth part of a second, we should get 
 10 coulombs for the cjuantity of electricity brought 
 down by the discharge, and we can then calculate the 
 average number of flashes per year which would be 
 required to charge the earth negatively to the re- 
 quired extent, if all flashes brought down negative 
 electricity. I find in this way that on the average all 
 over the earth there should be seven lightning dis- 
 charges per year for each square kilometre of surface. 
 Even making allowance for parts of the earth where 
 thunderstorms are of almost daily occurrence, it does 
 not seem to me, judging by present Information, that 
 lightning discharges from cloud to earth can play an 
 important part in increasing or diminishing the charge 
 of the earth. It would nevertheless be interesting to 
 collect information as to the direction of the passage 
 of electricity in a flash. Instruments intended to 
 measure the electric condition of the earth are as a 
 rule fine weather instruments, and are thrown off their 
 balance during a thunderstorm, otherwise a comparison 
 of the indications just before and after a flash would 
 materially assist us to decide the question. That the 
 earth is generally electrified positively when the sky is 
 covered by rain clouds would argue in favour of the 
 view that electricity which leaves the ground is positive. 
 Theories of the origin of thunderstorm electricity are 
 almost as numerous as those accountincr for the afeneral 
 electrification of the earth. The majority of them 
 must be received with considerable scepticism, but Dr 
 Simpson has recently made an important contribution 
 
152 Lecture ir 
 
 towards a rational explanation of what goes on in a 
 thunderstorm. He has proved that, contrary to what 
 was hitherto believed, the splashing and breaking up of 
 drops of water shew the "Lenard" effect, that is to say, 
 cause the drops to be charged positively while negative 
 electricity is dissipated into the air. A thunder cloud 
 is known to form in a rapidly ascending current of air. 
 Lenard, in an important research, has shewn that a 
 drop of water as it falls, never reaches a velocity 
 greater than 8 metres per second however large the 
 drop, while drops having a diameter of i "5 millimetres 
 fall with a velocity of about half that amount. An 
 ascending current of 8 metres per second will there- 
 fore keep the largest drops stationary in suspension, 
 while the smaller drops will be carried upwards. The 
 larger drops break up in the air, and doing so become 
 positivel}' electrified according to Dr Simpson. If the 
 ascending current spreads out laterally near the top of 
 the cloud, the vertical \'elocity is diminished, the drops 
 will grow and fall, but only to break up and be carried 
 upwards again. A quantity of electricity large enough to 
 account tor the lig;htninfr discharge can thus accumulate 
 in a cloud. Lightning can pass from the positively 
 charged raindrops to the negative charge, which is 
 carried away to the ascending and spreading current 
 of air, or from either of the charges to earth. 
 
 The interesting position in which the science of 
 atmospheric electricity is placed at the present moment, 
 must be my excuse if I have treated the subject with 
 a oreater amount of detailed attention than I have 
 given to other parts of terrestrial physics. 
 
 Before concluding this course of lectures, I feel that 
 
Age of Earth 1 53 
 
 I ought to refer to a question on which we thought we 
 knew something, until the discovery of radio-activity 
 threw us back to the primitive state where no opinion 
 is absurd and every hypothesis permissible. That is 
 the question of the age of the earth as an inhabitable 
 globe. The calculations of Lord Kelvin were based 
 on the known laws of conduction of heat, and on the 
 observed increase in temperature from the surface of 
 the earth downwards. Taking the reasonable view 
 that the earth originally was fluid and cooled down 
 by radiation into space, a time must have been, 
 when the crust began to solidify. According to Lord 
 Kelvin, solidification then spread through the mass 
 and no further diminution of temperature took place 
 until the whole earth was solidified. From known 
 data, among which the one concerning the temperature 
 of solidification is the most uncertain, he estimated 
 that about 50 million years must have lapsed since the 
 solidification was complete. Geologists however were 
 not satisfied with the shortness of time allowed to 
 them by Lord Kelvin for getting their strata in order 
 and demanded at least double the time. They may 
 have it now, and as much more as they desire, because 
 the generation of heat supplied by the decay of radio- 
 active products may alone be sufficient to account for 
 the loss of heat by radiation. So far as the laws 
 of heat can supply us with information, the earth 
 may have been in its present state during an infinite 
 time, and Lord Kelvin's estimate now stands as the 
 smallest allowable time, consistent with these laws. 
 We owe the recent investigation of this subject to the 
 
154 Lecture IV 
 
 Honourable R. J. Strutt\ who measured the radium 
 contents of a large number of minerals, and found that 
 if the interior of the globe contained as much radium in 
 proportion to its volume, as its surface layers, the heat 
 generated would far exceed the loss of radiation from 
 the surface. The earth would therefore get hotter 
 and not colder with time. If there is equilibrium of 
 heat, the loss throug-h radiation being- counterbalanced 
 by the generation of heat in the outer crust which 
 surrounds an inner core assumed to be radium free, 
 Strutt finds for the outer crust a thickness of 45 miles, 
 and for the inner nucleus a temperature of about 
 1 500° C. These calculations must alter our views 
 materially, though they are based on the assumption 
 that the heat production of radium is not affected by 
 a temperature of i500°C., and a pressure which is con- 
 siderably higher than anything we can produce in our 
 laboratory. This is a reasonable assumption, because 
 so far as we are able to judge at present, the rate of 
 decay of radio-active products is entirely independent 
 of external circumstances. 
 
 Lord Kelvin's conclusions regarding the solidity 
 and rigidity of the earth stand however firmer to-day 
 than ever. He based his opinion on the phenomena 
 of precession and nutation and found the observed 
 facts to be inconsistent with the idea of a liquid 
 nucleus. This result has quite recently received a 
 striking confirmation by the measurements of Hecker 
 of the actual deformation of the earth consequent on 
 
 ' Proc. Royal Soc. Lxxvn. p. 472 (1906). 
 
Rigidity of Earth 1 55 
 
 tidal action^ These observations give us definite in- 
 formation that the rigidity of the earth must be about 
 the same as that of steel. This is in complete agree- 
 ment with results obtained by other methods, the 
 most direct of which is perhaps that deduced from 
 the rate of propagation of earthquake waves. When 
 dislocation of matter in the earth's crust takes place 
 with sufficient suddenness to cause a tremor all over 
 the surface of the earth, the disturbance is propagated 
 by means of waves, and we distinguish three types of 
 waves. One is the wave of compression, identical in 
 character with the sound wave ; the second is a wave of 
 distortion resembling more nearly an electro-magnetic 
 wave, while the third is a wave running along the 
 surface like the waves of the sea. The wave of 
 distortion cannot exist in a liquid body, and as we find 
 that it actually does pass right through the central 
 portions of the earth, we conclude that the earth must 
 be solid throughout, except perhaps in isolated regions. 
 The rate of propagation gives, if the density of material 
 be known, a value for the rigidity, and the number that 
 is found in this way is quite consistent with Hecker's 
 result. But the most curious of all proofs of the great 
 rigidity of the earth is furnished by the movement 
 of the earth's poles. The earth's axis, according to 
 astronomical observations, changes its position with 
 reference to the earth itself, so that the north pole is 
 not a fixed point. The displacement is very small, the 
 position of its poles always falling within a circle of 
 about 10 metres radius. It has been known for a lone 
 
 o 
 
 ^ This only became known to me after these lectures were 
 delivered. 
 
J 
 
 156 Lecture IV 
 
 time that theoretically such a "wabble" was possible ; 
 but it remained undiscovered, because, the theoretical 
 calculation which assumed the earth to be a perfectly 
 rigid body, gave 10 months for the period of revolu- 
 tion of the axis, and it was this period that was looked 
 for. It was only comparatively recently that Chandler 
 discovered a period not of 10 months, but of 14 
 months, and Newcomb proved that the difference 
 between 10 and 14 months could be explained if the 
 earth were yielding, but only so much as a body 
 having approximately the rigidity of steel would do. 
 It is always satisfactory when we find that different 
 lines of reasoning; lead to the same result, and at 
 present there is hardly one, in the domain of Geo- 
 Physics which stands on so firm a basis as that giving 
 to the earth an extremely high rigidity. 
 
 In concluding these lectures I realise, that I have 
 omitted many important subjects and none more so than 
 that universal force of nature which at present stands dis- 
 connected from all other forces : the force of gravitation. 
 Experiments were made by Michael Faraday, than 
 whom no one was better qualified, with the object of 
 finding some relationship between gravitation and other 
 effects of nature, but without result. Many theories no 
 doubt have been proposed, and one, that of Lesage, 
 has received some support, until Maxwell shewed it to 
 be untenable. It may be revived some day in a less 
 vulnerable form, but for the present we may dis- 
 regard it. 
 
 Lorentz suggested that if the repulsion between 
 two quantities of electricity of the same kind is a little 
 less than the attraction between the same quantities of 
 
Gravitation 157 
 
 different kinds, then adopting tlie electron theory of 
 matter, gravitation might be explained. Another view 
 which has always fascinated me, is based on the remark 
 made by Lord Kelvin, that two sources of fluid, in 
 other words two points of space from which a liquid 
 flows out in all directions, attract each other according 
 to the law of gravitation. Now is it possible that 
 all matter, whether built up of electrons or not, dis- 
 charges particles indefinitely in all directions, thus 
 acting like a source of fluid ? Our first impulse, no 
 doubt, would be to reject this view as untenable, for 
 it would result in a gradual dissipation of all matter. 
 But if all forces were to diminish in the same ratio 
 as the masses, we might live in a universe in which 
 matter was constantly destroyed, without our ever 
 noticing it, because however little were left, the effects 
 depending on the ratio of force to mass would remain 
 the same. But while recognising the possibility of a 
 continual dissipation of mass, we are not necessarily 
 forced to admit it, if we adopt the above theory of 
 gravitation. The universe must have begun by a 
 process which lies outside physical laws, and it seems 
 to me no easier to grasp the conception of a creation 
 which took place at one single time than a creation 
 which continues throughout all ages. Indeed, if we 
 come to think of it, the continuance of a physical law 
 like that of gravitation is as much a miracle as the 
 continuous uniform creation of matter would be. 
 
 One final thought in conclusion. We recognise 
 atoms of matter, such as atoms of helium or of hydro- 
 gen, and we can observe these atoms in stars so far 
 away that the light from them takes several hundred 
 
158 Lecture IV 
 
 years to reach us. Now how far are these molecules ot 
 the same substance identical with each other? If you 
 weigh measurable quantities, containing millions and 
 millions of them,, you get the same weight to a fraction, 
 but that does not prove identity between individuals. 
 You mio'ht as well aroue, that if the total weight of 
 two flocks of sheep, containing the same number, is 
 the same, therefore all individual sheep must be equally 
 heavy. Molecules might differ from each other when 
 weighed singly, but when weighed by the million the 
 total might agree, if there is a certain average weight, 
 which they all more or less approach. Now spectrum 
 analysis allows us to discover discrepancies between 
 individual molecules, and though we cannot speak 
 absolutel)', because our instruments are not perfect, 
 we find that the light which is emitted by one sample 
 of helium, which may be on a star, is so nearly equal 
 to that of another sample on the earth, that there is 
 no room for the possibility of any great differences 
 between individual molecules. We may form a nu- 
 merical estimate of possible discrepancies between the 
 molecules of two samples of terrestrial helium, by 
 assuming the observed want ot homogeneity of spectrum 
 lines to be caused by real differences in the periods of 
 vibration. This of course gives us an outside value. 
 To put the result in a form, which is readily ap- 
 preciated, I will compare different molecules to a 
 number of clocks, the period of vibration of the 
 molecule corresponding to the time of swing of the 
 pendulums. If the clocks were rated to the same 
 accuracy as the molecules are found to be, there would 
 after 23 days running be no more than 10 of them, 
 
Identity of Atoms 159 
 
 shewing an error of as much as one second. It is 
 more than probable that increased powers of observa- 
 tion will prove an even nearer approach to identity in the 
 periods of molecular vibration ! This close agreement 
 between the masses of molecules of the same kind, and 
 between the forces which regulate the luminous motion 
 of the electrons is an essential condition to be satisfied 
 by any theory of atomic structure. 
 
INDEX 
 
 Actinium, 86 
 
 Action at a distance, 36 
 
 Air, ionization of, 144 
 
 Arrhenius, ionization of gases, 63 
 
 Atmospheric electricity, 142 
 
 Atom, present meaning of word, 93 
 
 constitution of, loi 
 
 of electricity, (>=;> 
 
 vortex theory of, 33 
 Atomic charge, J. J. Thomson and 
 H. A. Wilson's measurements, 70 
 
 Kecquerel, H., discovery of radio- 
 activity, S3 
 
 Boltwood, B. B., discovery of ionium, 
 86 
 
 Bragge, theory of Roentgen rays, 103 
 
 Bncherer, experiments on inertia of 
 electrons, 103 
 
 Cavendish Laboratory, 23 
 
 Chandler, periodical displacement of 
 earth's axis, 156 
 
 Chrystal, G., verification of Ohm's law, 
 16 
 
 Clausius, foundation of science of 
 energy, 2 
 
 Conductivity (electric) of outer atmo- 
 sphere, 133 
 
 Crookes, Sir Wm, discovery of radio- 
 meter, 32 
 improvement of vacuum pumps, 52 
 kathode rays, 55 
 
 Curie, Mr and Mrs, discuvery of 
 
 radium, 85 
 radio-activity of thorium, 85 
 Curie, P., and Laborde, thermal effect 
 
 of radio-active change, 88 
 
 Deflexion of a and /3 rays in magnetic 
 
 field, 87 
 Discharge of electricity through gases, 
 
 5^ 
 llittorfs experiments, $^, 57 
 J. J. Thomson's experiments, 67 
 theory of, 61 
 Wiechert's experiments, 68 
 
 Diurnal variation of terrestrial magnet- 
 ism, 130 
 
 Drude, electron theory, 113 
 
 Karth, displacement of poles, 155 
 
 age of, 153 
 
 rigidity of, 154 
 Kbert's theory of atmospheric electricity, 
 
 J 49 
 
 Einstein, principle of relativity, 109 
 Elastic solid theory of light, 39 
 Electric inertia, 99 
 
 theory of matter, 100 
 Electro-magnetic theory, early history, 40 
 
 experimental verification, 44 
 Electron theory of matter, 100 
 Electrons ejected from sun, 139 
 Elster and Geitel, electric charge nf 
 rain, 148 
 
1 62 
 
 Index 
 
 Emanation, discover}- of, 89 
 propeities of, 89 
 
 Energj-, conservation of, 1 
 atomic theory of, 1 1 1 
 dissipation by radiation, loi 
 
 Faraday, electrical theor}', 39 
 FitzGerald, on IMichelson-Morley ex- 
 periment, 109 
 Fresnel's wave surface, 5 
 
 Uarnett, William, demonstrator at the 
 
 Cavendish Laboratory, 28 
 Gauss, theory of terrestrial magnetism, 
 
 129 
 Giese, theory of ions, 61 
 Cilazebrook, verification of Fresnel's 
 
 wave surface, 5 
 demonstrator at Cavendish I^abora- 
 
 tory, 30 
 Goettingen, physical laboratory in 1S74, 
 
 15 
 Goldstein's experiments on kathode 
 
 rays, J 5 
 Gravitation, theories of, 156 
 Green, wave-theory of light, 39 
 
 Melium, identification with a particle, 
 
 95 
 accumulation in rocks, 95 
 derived from radium, 95 
 Helmholtz, foundation of science of 
 energy, 1 
 atoms of electricity, ()^ 
 laboratories at Berlin, 16 
 Hertz, H., verification of electro-mag- 
 netic theor)', 45 
 experiments on kathode rays, 56 
 FlittorPs experiments on electric dis- 
 charges, S3 
 proof of continuity of discharge, 57 
 
 Inertia, electric origin of, 99 
 
 alteration with velocity, 103 
 Ionic theory of electric currents, 79 
 
 Ionium, 86 
 
 Ionization by Roentgen rays, 78 
 
 of Air, 144 
 Ions, definition of, 63 
 
 rate of recombination, 80 
 
 Kaufmann, experiments on inertia of 
 
 electrons, 103 
 Kelvin, Lord, attraction of sources and 
 sinks of fluid, 157 
 contractile jether, 4' 
 foundation of science of energy, 2 
 on the age of the earth, 153 
 on the word atom, 92 
 vortex theory of atom, 33 
 Kirchhoft', G., on the probability of new 
 discoveries, 9 
 method of laboratory instruction, 13 
 object of scientific investigation, 115 
 on the vortex theory of atoms, 34 
 
 Laboratory Training, discussion of 
 methods, 19 
 early history of, 12 
 in England, 20 
 in Helmholtz' laboratory, 16 
 in Kirchhoff's laboratory, 13 
 Laborde, thermal effect of radio-active 
 
 charge, 87 
 Lamb, H., decay of electric currents in 
 
 spheres, 140 
 Larmor, Sir J., electron theory, 102 
 on the Michelson-Morley experi- 
 ment, 106 
 Lead, derived from radium, 97 
 Lenard, experiments on kathode ra)S, 
 77 
 on rain, 1 50 
 Lesage, theory of gravitation, 1 1;6 
 Lippman's capillary electrometer, 14 
 Lodge, O., electro-magnetic waves, 49 
 Lorentz, electron theory, 102 
 
 on the Michelson-Morley experiment, 
 
 109 
 on the Zeeman-effect, 113 
 
Index 
 
 163 
 
 Lorentz, theory of gravitation, 156 
 Loschmidt, theory of gases, 3 
 
 Magnetic dellexJon of kathode rays, 6,^ 
 
 of a and ^ rays, 87 
 Matter, electric theory, 100 
 Maxwell, Clerk, introductory lecture at 
 Cavendish Laboratory, 7 
 
 electrolysis, 50 
 
 electro-magnetic theory of light, 40 
 
 intercourse with students at the 
 laboratory, 29 
 
 properties of medium, 115 
 Michelson and E. W. Morley, 106 
 Molecules of electricity, 51 
 
 Xewcomb, ri., periodical movement of 
 eartli's poles, 156 
 
 Olim's law, verification of, 16 
 
 limits of validity in gas discharges, 81 
 
 Planck, Max, law of radiation, J12 
 Pockels, F., on lightning flashes, 150 
 Polonium, 86 
 
 identical with radium, 97 
 
 Radiation, dissipation by, 101 
 
 pressure due to, 138 
 Radio-activity, discovery of, S3 
 
 theory of, 90 
 Radiometer, 32 
 Radium discovered, 85 
 
 derived from uranium, 94 
 Rain, electric charges of, 14S 
 R.amsay, Sir VVm, helium derived from 
 
 radium, 95 
 Rayleigh, Lord, discovery of argon, 10 
 
 succeeds Maxwell, 30 
 Relativity, principle of, 109 
 Rest-Strahlen, 47 
 
 Roentgen, discovery of a new kind of 
 rays, 73 
 
 radiation nature of, 103 
 
 rays not homogeneous, 77 
 
 Rubens and Aschkinass on Rest- 
 Strahlen, 47 
 Rutherford, E., investigation on gas 
 discharge, 78 
 discovery of emanation, 89 
 identity of helium and a particle, 95 
 magnetic deflexion of a rays, 88 
 theory of radio-activity, 90 
 
 Schmidt, E., radio-activity of thorium, 
 
 85 
 Schwalbe, periodicity of sun-spots, 10 
 Secular variation of terrestrial magnet- 
 ism, 125 
 Shaw, demonstrator at Cavendish La- 
 boratory, 30 
 Simpson, G., electric charge of rain, 149 
 
 theory of thunderstorms, 152 
 Soddy, theory of radio-active change, 
 90 
 helium derived from radium, 95 
 Spectrum analysis, 32 
 Stewart Balfour, laboratory teaching, 20 
 
 experiments with rotating disc, 21 
 Stokes, elastic theory of light, 4, 39 
 
 on Roentgen's discovery, 77 
 Stoney, Johnston, theory of gases, 3 
 Strutt, Hon. J., accumulation of helium 
 in rocks, 95, 154 
 
 Terrestrial magnetism, secular varia- 
 tion, 125, 141 
 
 connected with sun-spots, 135 
 
 diurnal variation, 131 
 
 existence of potential, 127 
 
 Gauss' theory, 129 
 Thermal effect of radio-active change, 
 
 88 
 Thomson, J. J., txperiments on electric 
 discharge, 67, 78 
 
 measurement of atomic charge, 69 
 
 on the constitution of the atom, 71 
 Thorium, radio-activity of, 85 
 Todhunter, on conical refraction, 25 
 
 on conflict of studies, 25 
 
1 64 Index 
 
 Uranium, the parent .jf radium. 94 Wilson, H. A., measurement of atomic 
 
 charge, 70 
 V.jitex theory of atom. 33 Wilson, C. T. R., theory of atmo- 
 
 spheric electricity, 147 
 ^^■ave theory of light, 39 Wireless teletjraphy, 47 
 
 Weber's laboratory, i; 
 
 Wiechert, discharge through gases. 6S Zeeman, discovery of magneto-optic 
 Wien's investigations on the law of efiect, 23 
 
 radiation, 22 Zeeman-eft'ect, 113 
 
 CAMBRIDGE; PRINTED BY JOHN CLAY, M..\. AT THE UNIVERSITY PRESS. 
 
.."'■f?: ■ 'H^i