THEORIES OF METALLIC CONDUCTION OE 
 ELECTRICITY WITH EXPERIMENTS 
 ON THE CONDUCTIVITY OF A 
 RO FATING COIL 
 
 ; BY 
 
 CLARENCE CARL SCHMIDT 
 B. A. Cornell College i q 1 7 
 
 THESIS 
 
 Submitted in Partial Fulfillment of the Requirements for the 
 
 Degree of 
 
 MASTER OF ARTS 
 
 IN PHYSICS 
 IN 
 
 THE GRADUATE SCHOOL 
 OF THE 
 
 UNIVERSITY OE ILLINOIS 
 
 1921 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Digitized by the Internet Archive 
 in 2016 
 
 https://archive.org/details/theoriesofmetallOOschm 
 
[ 
 
 TABLE OF CONTENTS 
 
 I INTRODUCTION* • • 1 
 
 II EARLY THEORIES OF METALLIC CONDUCTION 1 
 
 III ELECTRONIC THEORIES 4 
 
 The Free Gas Theory 
 Thomson's Doublet Theory 
 
 Pressure Exp er im ent s and "Gap” Theory of Bridgeman 
 IV EXPERIMENTS WITH MECHANICAL FORCES ON ELECTRONS. • . .13 
 
 V EXPERIMENTS BY THE WRITER 15 
 
 VI SUMMARY AND CONCLUSIONS 36 
 
 VII BIBLIOGRAPHY 28 
 

 
 
 
 
I INTRODUCTION 
 
 how does an electric current pass through a metallic wire? 
 
 This most common of electrical occurrences presents an unsolved 
 fundamental problem in electrical science. In tne passage of the 
 current through electrolytic and gaseous conductors, the theories 
 seem fairly complete and satisfactory. Experiments nave shown that 
 the electric current carried through electrolytes ana gases is accom- 
 panied by the transfer of material particles or ions and it has been 
 proved by varied and ingenous experiments that the electric transfer 
 is a convection process. 
 
 For metallic conduction, there is as yet no explanation which 
 may be regarded as at all adequate and complete. It is the purpose 
 of this paper to present and discuss the various theories of metallic 
 conduction ana to describe a new experiment which has given results 
 that seem to be of importance in tne problem. 
 
 II EARLY THEORIES 
 
 Probably the first theory of metallic conduction was formulated 
 
 by Wilhelm weoer (d4) who assumed that two equal streams of positive 
 
 . 
 
 and negative electricity flowed in opposite directions. The elec- 
 tricity would continually bp accelerated by the electromotive force, 
 if there were no resistance. This resistance according to Y/eber is 
 caused by the attraction of the opposing electrical masses. If now 
 a positive and a negative particle approach each other there will be 
 an attraction, between them, similar to that between two bodies at- 
 tracted oy gravity. The particles will under such action describe 
 spiral paths about a common center. The applied electromotive force 
 causes the paths of the rotating system to be drawn out longitudinal- 
 ly sc that the particles come into the sphere of influence of 
 
3 
 
 neighboring systems to which tney escape. New systems are thus formec 
 and by a continuous process the electric current is carried through 
 the length of the conductor. 
 
 Herwig (bb) conceives the stream of electricity as a movement 
 of the electrical particles from one molecule to another and. the re- 
 sistance as the force of opposition encountered in moving from one 
 atom to another. This involves a Change in the amplitude of vibra- 
 tion of the particles as well as a movement in the direction of the 
 electromotive force. Both of these phenomena require a certain 
 amount of work;since the movement of the particles carrying the cur- 
 rent requires that the amplitude change also, and since heating of 
 the conductor increases the energy of vibration, then heating should 
 increase the resistance of a conductor. 
 
 Another theory of metallic conduction of electricity was that 
 the charge was carried by molecules, out in the case of dense solids, 
 such as the metals, such a process is hardly conceivable, since noth- 
 ing as large as a molecule could pass through the metal with such 
 ease as must be the case, because the forces required to send a cur- 
 rent through a wire are small. 
 
 From this it is seen that some other carrier must be found to 
 account for the passage of a current. Giese (10) was one of the 
 first to form a different theory. He argued against the molecular 
 transfer theory, since if molecules were acting as the carriers there 
 must be a large amount of matter carried with the current. it had 
 been determined previously that the passage of electricity through 
 flames was by means of charged atoms, hence it was argued that atoms 
 might perform tne same office of carriers in the case of metals. In 
 
 order to produce the same effect by means of a charged molecule 
 

 
 
 
 
 
 | PJ.H 3ai^ . ■ 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 . 
 
 
 
 . 
 
 ' 
 
 . 
 
 
 . 
 
 
 
 
 
 
 
 
 
3 
 
 several thousand times (at least 1700 times in the case of hydrogen' 
 as much more matter would have to be transferred as in the case 
 where tne ions are given this property. 
 
 Furthermore, he snows that the molecule cannot be electrified. 
 Water is an example which readily shows this. Distilled water has 
 very high insulating properties, but wnen used as a solvent its ions 
 according to Giese become highly conducting. 
 
 According to his tnecry, a given substance or metal may nave 
 atoms wnich are exactly alike except for tneir cnarges, for in- 
 stance, Cu + and Cu_. The property of being a conductor or a non- 
 conductor depends upon tne relative number of each kind. The mole- 
 cules may excnange ions and thus the excess charges will find their 
 way to the surface. 
 
 When two metals are placed in contact, the excess ions on 
 each tend to neutralize each other. An e.m.f. would then be pro- 
 duced, the magnitude of whicn would depend upon tne kind of metals 
 that were placed in contact. These excess ions were termed "free 
 ions" on account of being free to move about and thus carry the 
 current. The difference between metallic and electrolytic conduc- 
 tion is that in the latter, tne ions bodily carry the charge from 
 place to place, while in the former, the charge is passed along by 
 the int ercnanging of ions between tne molecules. The metals being 
 of denser material, the molecules are bound to remain in one place 
 and are not free to move as in tne case of tne electrolyte; hence, 
 there is merely a shift of ions from one to the other. 
 
 The difference, according to this theory, between conductors 
 and insulators is, that the latter have no free icns, or tnat each 
 ion is bound to a particular molecule and cannot Change to another. 
 

 
 
 
 
 
 
 . 
 
 - 
 . 
 
 
 
 
 
 . . . 
 
 . 
 
 
 
 
 
 
 
 . 
 
 
 . 
 
 ■ ■ 
 
 . 
 
 
 
 
 
 
 
 
 
4 
 
 A cnange in the resistance indicates a molecular cnange such tnat 
 the metal eitner becomes more conductive, or offers more resistance 
 according as tne number of free ions becomes greater or less. 
 
 Such a theory tho explaining some things very well was far 
 from satisfactory. Numerous experiments were made to determine 
 whether there was any transfer of matter with the passage of a cur- 
 rent. Two metals were placed in contact and a current passed 
 through them for a long time, in one case for tne space of a whole 
 year; but no trace of either metal could be found in the other. 
 
 These earlier theories have, of course, been antiquated by 
 the discovery of the electron. They are , however, important in 
 being the direct forerunners of our present electron theories of 
 conduction. Indeed, when these earlier theories are translated in- 
 to tne language of tne electronic theory, they are very suggestive 
 and in some cases show anticipation of our more recent ideas. 
 
 III. ELECTRONIC THEORIES OF CONDUCTION 
 
 The study of the conduction of electricity tnrougn gases whic. i 
 was extended about this time threw new light upon the subject, es- 
 pecially when the electron was discovered by J.J. Thomson and oth- 
 ers about 189?. It was shown that this was the basic particle of 
 electric lty whicn might move and carry the current through the 
 metal. Its small size, nign velocity, and the fact that electrons 
 
 are the same in every metal made it tne logical carrier of elec- 
 tricity. 
 
 Riecke (24) carries out an iaea expressed. 
 
 tnough not develop i 
 
 by Weber (34) who stated years before that the molecules of a metal 
 
 were surrounded by moving molecules or particles of electricity. 
 

 
 . 
 
 - 
 
 - 
 
 ■ 
 
 . 
 
 . 
 
 
 
 
 . 
 
 
 
 
 . . 
 
 . 
 
 ' 
 
 . 
 
5 
 
 These particles Riecke identifies as tne electrons or corpuscles 
 discovered tnrough the work of J.J. Thomson and others. 
 
 Drude (8) added considerable to tne ideas of Riecke and 
 showed a marked similarity between these ideas and the ideas ex- 
 pressed by Giese. The properties of carrying tne current ascribed 
 by tne latter to the ion, Drude points out belong to the electron, 
 tne cnarge unattached to any matter. These electrons seem to ocey 
 the kinetic theory of gases and hence Drude' s theory is known as the 
 "free gas" theory. 
 
 Since metals are the best conductors of both electricity and 
 heat, the electron must play a part in the conduction of both heat 
 and electricity. The fact that electrons or one metal do not diriei 
 from those of another strengthens this viewpoint. When the coef- 
 ficients of thermal and electrical conductivities are compared 
 there is found to be a constant ratio between them. This had been 
 noted many years before and is known as the Wiedemann - Franz' (38) 
 ratio; questions have been raised about the significance of this 
 ratio, but it has been shown to be a fact both theoretically and 
 experimentally. There is also a marked parallelism between the 
 change of thermal and electrical conductivity with temperature. 
 
 A gas is considered as having molecules moving in ail direc- 
 tions ana if the number is taken as N, then in each of the six di- 
 rections parallel to the three axes there are M/6 moving with a 
 velocity v. Each molecule has a momentum +mv on striking a surface 
 and -mv on rebounding from it. The pressure on the surface is 
 Nv/6 x 2mv or P = 1/3 Nmv 2 . Mm is the mass per unit volume of gas 
 or the density. 
 
 But, pressure x volume = RT 
 

 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
6 
 
 or 1/5 Nmv i 2 = RT. 
 
 This snows that the kinetic energy is proportional to the absolute 
 temperature or 1 Jd mv 2 = °c T where c <= 5/2 R/N. Thus when electrons 
 in a ooncluctor are in temperature equilibrium the velocity may be 
 found from the kinetic theory. 
 
 £e 
 
 The acceleration of an electron after collision is f = — wfter e 
 E is the electrical intensity. The average distance traveled by an 
 electron, or the mean free path is A, hence the distance traveled due 
 to the acceleration, f, is 
 
 1 Ee Af_ 
 
 2 m v a 
 
 where ^ is the time taken to traverse the mean free path in the di- 
 rection of tne field. The average velocity m this direction is 
 
 1 Ee A 2 v _1 EeA 
 
 2 m v 2 A 2 mv 
 
 but mv 2 = <=* T or m = ~~ 
 
 2 v 2 
 
 Therefore the average velocity in the direction of the field 
 E e A v 
 
 is 
 
 4 °c T 
 
 Tne corresponding current, is i\ie times the velocity where J5J is 
 the number of electrons per unit volume, or 
 
 current = = i 
 
 4°< T 
 
 i N e 2 A v 
 
 and conductivity, <r = g- = — ^- 5 ^ * 
 
 It follows that Onm's law is ooeyea, for the conductivity is inde- 
 pendent of the current and further the conductivity is inversely pro- 
 portional to the. absolute temperature. 
 
 In a very similar way the thermal conductivity k, is found to 
 be hv<* A 
 
7 
 
 The ratio of k/j- is tnen given as 
 
 Nvoc A 
 
 5 
 
 he 2 Av 
 4<*T 
 
 = Cx ) 2 |t 
 
 Experimental results show tnat this ratio witn very few exceptions is 
 constant, ana its value is given as 6.5 x IQ 10 . 
 
 Wnen an electric force is appliea tnere is a arnt oi electrons 
 in tne airection of the force. If the parts of the metal are at dif- 
 ferent temperatures the heat will flow from the hotter parts to the 
 cold. If the greater part of the heat conductivity is due to elec- 
 trons the thermal conductivity should hear a constant ratio to elec- 
 trical conductivity. 
 
 If an electric current passes through an unequally heated metal 
 electrons will drift from places of high potential to low potential 
 and carry heat with them; thus the passage of a current through a 
 conductor will alter the flow of heat. This is known as the Thomson 
 effect discovered by Sir Wm. Thomson, Lord Kelvin. 
 
 J.J. Thomson (30) in noting the effect of temperature on the 
 resistance of metals finds that the free gas theory is inadequate. 
 
 If the electrical conductivity is due to free electrons, changes in 
 conductivity are due to two factors, n, the number of free electrons 
 and,k, their mobility. The product of n and k, if the conductivity 
 
 is inversely proportional to the absolute temperature, must be greater 
 
 ~ w 
 
 '*"* TTTrr 
 
 at low temperatures than at high. Since n is proportional to e 
 where T is the absolute t emperature temperature and H the gas con- 
 stant, and, since this factor decreases at low temperatures, then 
 the increase in nk must be due to k, or tne mobility of the system. 
 But this he finds must also decrease witn lowering temperature. 
 

 
 
 
 
 
 
 
 
 , 
 
 . 
 
 . 
 
 . 
 
 
 
 
 
 . 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 . 
 
 • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' " ' 
 
 
 
 
 
 
 
 
 
 
 ^ A . . O - * - 
 
 
 Y-fifriilS '-iX j.1 
 
 -x ;-s ? • I J... 
 
 
 
 
 
 
 
 , 
 
 
8 
 
 Experiments by Kammerlingh Onnes (29) with metals at low tem- 
 peratures seern to contradict the free electron theory. He found that 
 the resistance of pure metala, sucn as lead or mercury, at the tem- 
 perature of liquid helium becomes so small that it cannot be measured, 
 or less than one thousandth-millionth of its value at 0°C. In the 
 case of a lead ring a current started by moving a magnet up to the 
 coil will continue for some time without any very noticeable decrease 
 if the low temperature is maintained. In fact, it was estimated that 
 the current would continue for four days before it would fall to one 
 half its value. 
 
 In view of the difficulties thus involved in tne free gas theo- 
 ry, Thomson (29,31) introduced another, analogous to the molecular 
 theory of magnetism. He shows in this that the presence of a nega- 
 tive charge means that there is also an equal positive charge. Each 
 molecule of the metal has an electric dipole consisting of equal and 
 opposite electric charges a short distance apart. When an electric 
 force is applied these dipoles align themselves in chains extending 
 in the direction of the electric force from one end of the conductor 
 to the other. These doublets will produce strong electric forces 
 which will tend to pull the electron from one atom to another. The 
 difference between a conductor and an insulator lies in the power to 
 resist the electric force. In the case of metals this is much less 
 than in the case of insulators. 
 
 The superconductivity noticed by Onnes (30) in which the re- 
 sistance drops down very suddenly to practically zero may be ex- 
 plained in that the electric force polarizes the dipoles in the metal 
 which then remain in that state wnen cooled, even though the electric 
 force is removed; hence, the current will still persist. 
 
9 
 
 In measuring the effect of increased pressure in twenty two 
 metals Bridgeman (2) found that with the exception of two, namely, 
 antimony and bismuth, the increase in pressure tends to decrease the 
 resistance. He found, however, that, though under pressure of, say, 
 12,000 Kg and at a temperature of 0°C the volume is smaller than at 
 0° absolute. Nevertheless, the decrease in resistance at that pres- 
 sure is a very small percent of the change in resistance brought 
 about when tne metal is cooled to 0° absolute, a point toward which 
 the resistance tends to approach zero. In the case of the two abnor- 
 mal metals mentioned the change in volume due to pressure of 12,000 
 is almost three times tne change in volume due to cooling to 0° 
 absolut e. 
 
 In regard to the theory of Brude, Riecke and Lcrentz, which ex- 
 presses the conductivity in the form 
 
 conductivity = constant g— 
 or 
 
 conductivity = constant — 
 
 N 
 
 wnere N is the number of electrons per cubic centimeter, 1, the mean 
 
 free path, u, the velocity, ana T, the absolute temperature, 
 
 Bridgeman (6) says it fails to account satisfactorily for specific 
 
 heat, tne variation of resistance with temperature, and the negative 
 
 pressure coefficient at constant temperature. The values of N, 1, 
 
 and u, cannot under any circumstances be conceived as changing enough 
 
 to account for the change in resistance. According to J.J. Thomson's 
 
 doublet theory the conductivity is given by 
 
 n _ 1 N epdM 
 C ~ 3 KT 
 
 where N is the number of doublets per cubic centimeter, p, the numbei 
 of electrons emitted by an atom per second, M, the moment of the 
 
. 
 
 . 
 
 . 
 
 
 
10 
 
 doublet, K, tne gas constant, ana d, the distance between tne centers 
 
 of adjacent atoms. Part of the increase might be accounted for a 
 
 change in N. But, according to Richardson, N for tungsten at 2000° 
 
 16 
 
 is 1.7 x 10 which tends to be the minimum value, ana tnis number is 
 much too large to justify tne equation; hence, he concludes that this 
 theory is inadequate. 
 
 The equation developed by Grdneisen (11) is the only one wnich 
 in any way provides for pressure effects. It is given as 
 
 u^TF^T = u ( “F) s * s + Ws ~ c^H^P \} +C * T T J- 
 
 in which u is tne velocity of the electrons and — ( i ^) a is a relation 
 
 U h? ° 
 
 between pressure ana velocity which is unknown. The electrical re- 
 sistance of a pure metal is proportional to a universal function of 
 T/ v m where v m is determined by tne atomic temperature. Grdneisen 
 found tnat the experimental results of Lissell, Williams, LaFay, 
 Beckman, Chwolson, Tomlinson and others agree with this equation. 
 
 According to this tneory of Grdneisen the movement of the elec- 
 trons increases witn the compression of the atoms wnile the amplitude 
 
 of vibration decreases. This is also the viewpoint of Bridgeman who 
 
 * 
 
 states also that the cnange in resistance may be due to a change in 
 the amplitude of vibration of the atoms. 
 
 Be proposes a different tneory for tne resistance offered to 
 the passage of electrons. In the free gas theory this is explained 
 as due to the friction of the electrons, or the collision of the 
 electrons with the molecules, wisn thought of it as caused by the 
 collision of the electrons with the positive neucleous. 
 
 Bridgeman (6), on tne otner hand, departs radically from this 
 and says that no resistance is offered to the electron as if passes 
 through tne molecule, bux tnat the resistance is explained as due to 
 

 
 
 
 
 
 
11 
 
 a force which is required to move the electron from one atom to an- 
 other. He calls this tne "gap" theory, he snows that the electron 
 encounters difficulty in getting away from the atom, as may oe seen 
 from the fact that there is a definite ionizing potential in the case 
 of a gas. This force which tends to act against the passage of an 
 electron from atom to atom is less, the smaller the mean free path, and 
 since this depends upon tne amplitude of vibration, tnen tne amount 
 of resistance can be traced directly to this amplitude of vibration. 
 
 The classical expression for the resistance R = —f — will hold 
 
 e 3 nl 
 
 where, e is the cnarge, ana v the velocity of the electron. Ohm’s 
 law can also be explained by this theory. 
 
 In support of this theory Bridgeman shows that where the metal 
 changes tne state of aggregation as in melting, the state wnicn has 
 the smaller volume has the least resistance. This is snown to be 
 tirue not only in the case of metals which act normally upon melting, 
 but also for those in which abnormalities occur as in Bi, Sb, and Ga, 
 
 in which tne liquid has tne lesser volume ana tne smaller resistance. 
 
 He holds that this is more easily explained from tne "gap” point of 
 view tnan from Wien’s in which collision with the atomic center forms 
 the basis of explanation. 
 
 He points out tnat in the classical theory 1, ana v were de- 
 termined, but n was made to fit as the case demanded without snowing 
 
 sufficient reason for the change in n. In the gap theory variations 
 in v and 1 are accounted for; hence, there is no necessity for chang- 
 ing the value of n, wnich is therefore assumed to remain constant. 
 
 Bridgeman applies the "gap" theory to a number of different ef- 
 fects to snow that it is not inconsistent with a number of facts 
 which nave been explained by various, and in most cases, well 
 
12 
 
 established, tneories. 
 
 This theory also gives a consistent explanation of the resist- 
 ance of alloys. Of these there are two kinds, those wnicn for m mixed 
 crystals ana those which do not. In the latter case separate crystal 
 are formed of the two metals and the resistance may be compared from 
 the rule of mixtures. In the other case, however, mixed crystals 
 are formed which combine best according to certain proportions of the 
 two components. That in some cases an imperfect fitting of tnese dif li- 
 ferent snaped crystals would tend to make the gaps larger and tnu3 
 increase the resistance can easily be deduced. in such a case the 
 resistance of an alloy might be greater than either of its component 
 parts. That the alloy is not affected so much by changes in tempera- 
 ture is explained by assuming that the imperfect fitting of atoms 
 persists as the temperature is decreased. 
 
 hard drawing increases the resistance of metals and is explained 
 as due to tne breaking up of the crystal grains and thus making the 
 atoms fit more imperfectly. This explanation seems adequate even 
 though the density decreases when the metal is hard drawn. 
 
 Hardness is caused by a staggering of atoms, hence the increase; 
 resistance is due to a greater separation of atoms, which is in ac- 
 cord with the facts. 
 
 In the case of solid bismuth and antimony wnicn are abnormal, 
 tne increase of resistance per degree rise in temperature is normal. 
 This is because tne effect of temperature overshadows any pure volume 
 effect so that the temperature coefficients are nearly the same for 
 all metals irrespective of the pressure coefficient. That liquid 
 bismuth behaves normally is probably due to the fact that tne crystal 
 structure is broken down and then the metal acts like any other. 
 
13 
 
 In none of tne work done by Bridgeman was anything found that 
 would seem inconsistent witn tne "gap” theory. This, however, seems 
 to nave wcnueri'ul elastic possibilities so tnat any case may be ex- 
 plained without actually contradicting tne tneory itself. 
 
 Though not referred to by Bridgeman, F. Credner’s (7) work on 
 tne effect on resistance of twisting, bending ana drawing would fur- 
 ther lend weignt to tms view. From a large amount of data he found 
 that eacn of these operations greatly increases the resistance. 
 Tamrramwho did similar work gives a three fold explanation: formation 
 of space between individual crystalline units, tne alteration of the 
 structure in tne case of stretching, ana formation of spaces on tne 
 gliding surfaces in tne case of twisting. 
 
 IV. EXPERIMENTS WITH MECHANICAL FORCES ON ELECTRONS 
 
 Various attempts have been made to throw additional light upon 
 the subject of metallic conduction of electricity by means of experi- 
 ments in wmcn mechanical forces are exerted upon the electrons. The 
 earliest of these were made when the ionic theories of conduction 
 prevailed. I. Bosi (quoted by nail (12)) in 1877 used electrolytes 
 of sulphate of zinc ana sulpnate of copper wnicn he caused to flow 
 with a velocity of 10 to 12 cm. per second, ne reported that there 
 was considerable difference in tne resistance when the current flowed 
 in the direction of motion and when it flowed in the opposite direc- 
 tion. 
 
 J.B. Haywood (12) of the Harvard Graduate school, repeating 
 the experiment under Professor E.H. nails’ supervision, found tnat 
 no such larg£ effects could be detected and indeed if any effect ex- 
 isted, it was exceedingly small. 
 
14 
 
 A similar experiment with electrolytes of different kinds was 
 performed by J. Nabl (19). His results show that there was practical- 
 ly no change in the resistance aue to changing the direction of the 
 current with respect tc the movement of the electrolyte. 
 
 Tolman, Osgerby, and Stewart (32) used an acceleration method. 
 
 A tube was filled with potassium iodide and suddenly accelerated 
 which set up a difference of potential between the two ends. When 
 such a tube was placed around the rim of a bicycle wheel and quickly 
 started a current was set up. This is explained as due tc the fact 
 that tne negative iodide atoms are heavier tnan the positive potas- 
 sium ions and thus set up an e.m.f. opposite tc the direction of ro- 
 tat ion. 
 
 E.F. Nichols (20) brought forth a new aspect by rotating a 
 disk of aluminum on which contacts had been made at tne center and 
 the outer edge. The object of this experiment was to determine from 
 the e.m.f. developed and the change in resistance, whether the posi- 
 tive or the negative charge was bound in the case of metal. Since 
 the ratio of e/m is different for positive and negative particles, 
 the mass must be different, which would then give different centrifu- 
 gal forces to each. His apparatus did not prove sensitive enough 
 to enable him to detect results which would verify his theory that 
 negative ions were tne carriers of electricity. Various heating ef- 
 fects also proved to be complications, obscuring any positive con- 
 clusions. 
 
 The acceleration method was used by Tolman and Stewart (32) 
 with metals. Their former apparatus was modified by replacing the 
 tube filled with an electrolyte by a coil of wire. When this was 
 rotated and suddenly stopped there was a "kick" in the connected 
 

 
 
 
 
 
 
 
 
 
 
 
15 
 
 galvanometer, explained as aue to the momentum of the electrons. The 
 experimental results seem to warrant the acceptance of the free elec- 
 tron theory ratner than the doublet theory, for it is more probable 
 that free electrons could be given a momentum than that such accele- 
 ration would in any way produce action on electric doublets. 
 
 Two years previous to the worx just described an experimental 
 investigation was carried on in the Physics Laboratory of the Univer- 
 sity of Illinois by Paul L. Bayley (1). In this work the coil was 
 rotated uniformly ana the change of resistance caused by rotation 
 noted. The results were quite small and could not be determined very 
 accurately; though in general the resistance was larger when the coil 
 was rotating. ' Experimental difficulties in the form of poor contact 
 of the measuring instruments with the rapidly moving parts were not 
 encouraging for further experiments, until better methods of connec- 
 tion could be devised. 
 
 V. EXPERIMENTS BY THE WRITER 
 
 It was the purpose of the present experiment to remove the dif- 
 ficulty involved in tne contact and to extend, if possible, the re- 
 sults obtained by Bayley. It was hoped thus to obtain additional 
 data which would disclose more concerning the mechanism of metallic 
 conduction. 
 
 The coil consisted of a large number of turns of insulated 
 copper wire. No. 25 B.& S., which was wound in a groove in the edge 
 of a disk made from one-inch compressed fibre. Tne average diameter 
 of the coil was 28.5 cm. Each layer of wire was covered with shellac 
 and allowed to dry before the succeeding layer was wound, thus insur- 
 ing perfect insulation with all the wires rigidly bound together. 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 ■ 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 - 
 
 
 
 
 
 
 
 ✓ 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 . I - 
 
 
 
 
16 
 
 The ends of the wire were attached to plates on eitner side of the 
 disk which in turn were rigidly attached to tne snaft of tne wheel 
 or disk. The shaft was cut at the middle, sc tnat the two ends were 
 insular ed from each otner. The arrangement of tne various parts are 
 shown in Fig. 1. 
 
 In order that there might be no possible magnetic effects, all 
 parts of tne disk and bearings were made of non-magnet ic substances, 
 brass was found best adapted for the metal parts. The axle itself 
 turned in babbitt bearings which were set on dry wood pillars and 
 completely insulated from each other. On the outer end of tne snaft 
 a pulley was attached to wnich a quarter inch leather belt trans- 
 ferred the power from a two-phase, 60-cycle, 440-volt induction motor 
 which was used in order to prevent any possible electromagnetic in- 
 duction effects. 
 
 In the apparatus used by Bayley the contact with the revolving 
 parts was made by means of a mercury fountain attached to a hollow 
 tube all parts of which were amalgamated, this tube acted as a brush 
 to collect the current. This contact was found to be very unsatis- 
 factory by Bayley particularly at high speeds, because thermoelectric 
 forces were produced and it was also variable in action. Another 
 method was therefore used in the present experiments. A tapering 
 hole was carefully drilled along the axis of the shaft at each end 
 and into this a brass plug which could easily be removed was fitted. 
 At the center of the plug a copper wire. No. 25 B.&S. , was soldered. 
 This wire extended outward in line with tne axle for a distance of 
 fifty feet and twisted upon itself wnen tne coil was set in motion. 
 This is tne method used by Tolman and Stewart. In this experiment 
 where change in resistance was to be noted instead of an e.m.f. or an 
 

 
 . 
 
 . 
 
 
 
 
 . 
 
 . 
 
 . 
 
 ■ 
 
 ' 
 
 . 
 
 
 
 
 
 
 
 . 
 
 , . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 . ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
w 
 
 C=i 
 
 lr\ 
 
 Seta.!'! n_9 Wc£x~ 
 
 Rubber 
 Insu lotion 
 
 Pu/ ley 
 
 Rotating Coil 
 C.C. Schmidt 
 
 May/ 92 ! 
 
 n 
 
 Coll 
 
 W / re 
 
 / 
 
 Bea rings 
 — — ^ 
 
 
 s 
 
 
 
 niMi 
 
 s 
 
 
 
 
 111 
 
 
 
 
 
 - ^ / o 
 
 ’Short circuit 
 
 Sere w 
 
 /V> J 
 
 -Hickory Him 
 
18 
 
 instantaneous current sucn an arrangement proved very unsatisfactory. 
 It was found that there was a large change in tne resistance of the 
 wire when it was twisted; this as is shown also in the work of 
 Tammam and Credner (7) wnicn has been mentioned above. 
 
 In order to eliminate this effect tne wire was made oniy tnirty 
 centimeters long and passed through a tube containing mercury. The 
 ends of the tube were stopped and the wire passed through the stop- 
 per at each end and then fastened to a small swivel and kept taut by 
 a spring. In this way tne wire was made to spin in the mercury bath 
 when the disk was rotated. This device proved a very satisfactory 
 means of connecting tne fixed part of the circuit witn the revolving 
 coil as tne results snow. The cnange in resistance was quite constats 
 and very small in magnitude. Tne actual change due to the rotation 
 of tne connections could be determined by inserting a screw which 
 connected the brass plates on tne two faces of the disk; this made a 
 
 snort circuit around the coil. The effect of rotation on the resist- 
 % 
 
 ance of the contact could thus be determined and allowance made for 
 the same. 
 
 When tne first measurements were taken with the coil at rest, 
 it was found that very slight changes in temperature caused a very 
 noticeable change in the resistance as was determined with a Carey 
 Foster bridge. Considerable time was then spent in finding a room 
 where the temperature change would be as small as possible. A 
 double walled room ordinarily used for sound determinations was 
 finally used. To show how sensitive the coil was it may be noted 
 that the presence of one person cr the lighting of an ordinary elec- 
 tric light could be detected by tne cnange in resistance . 
 
 A representative curve of the determination of tne resistance 
 of this coil are shown in Fig. 2. The resistance is plotted as the 
 

 
 
 
 
 
 . 
 
 •« 
 
 . 
 
 
 I 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
20 
 
 ordinate ana tne time as tne abscissae. it shows a graaual cnange 
 in tne resistance while the coil is being rotated, wnich was inter- 
 preted as due to temperature changes since copper has a large temper- 
 ature coefficient of resistance. A second coil was made using manga- 
 nin wire. The results were of similar character, but the changes were 
 much less. 
 
 Later the manganin was replaced by 85.0b meters of "Advance” 
 wire obtained from the Driver-tiarris Company. This was not shel- 
 lacked during the winding because the ordinary insulation was suffi- 
 cient for the small electromotive forces used, in order to make the 
 wire coil rigid ana to guard against any possible stretching, due to 
 centrifugal force, which would of course tend to increase the resist- 
 ance, n elted sealing wax was poured into the groove over tne wire. 
 After tnis was smootniy turned down in a latne a hickory rim was 
 bent into a circle and bound closely upon tne rim of the disk by brass 
 bands. In this way the wire was held rigidly in place. 
 
 When tnis new coil was rotated a relatively large increase of 
 resistance was noted. Since "Advance" wire nas a very small tempera- 
 ture coefficient, and that negative (-.U000056), such an effect could 
 not be explained as a temperature cnange in tne resistance of the 
 wire, hence, another source of the effect was sougnt. Since it con- 
 tinued after the rotation ceased it was suspected to be due to 
 thermo-electric forces. To test this, a sensitive galvanometer was 
 placed in series with the coil; then a heated block of brass was 
 placed in contact with the bearing and also with the pulley. Each 
 of these gave a deflection. By placing the heated block on the con- 
 tacts between the wire and the brass piares similar results were ob- 
 tained. It was evident that though the disk was apparently symmetri- 
 cal, one side was heated more tnan the ocher, either by the friction 
 

 . 
 
 
 
 
 
 
 . 
 
 . 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 . 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 .. 
 
 
 . 
 
21 
 
 in the bearings or else by friction of the belt on the pulley. This 
 unsymmetrical heating ana the consequent thermo-electromotive force 
 was the cause of this irregularity. Hence, the disk was remodeled. 
 
 Fig. 3 shows the remodeled form of the disk. Two radical 
 changes were made:- first the babbitt bearings were replaced by high 
 speed ball bearings of the S.K.F. type, thus reducing the friction 
 to practically nothing; furthermore, the brass was taken out as a 
 metal of the circuit by drilling a hole along the axis of the shaft 
 and another through the brass plate perpendicular to the shaft. 
 
 Through this passage an insulated copper was drawn. The brass plug 
 was removed and replaced by one of copper which was insulated from 
 the brass shaft by a hard rubber bushing. In this way the thermo- 
 junctions between brass and the other metals were removed. The 
 junctions between the copper and "Advance” wire were placed on eithei 
 side of the disk as near the outer edge of the disk as possible. It 
 would be very unlikely that under these conaitions the two junctions 
 would be at temperatures noticeably different. In fact, no thermo- 
 electric forces existed at these points after these changes were mads. 
 In order that the coil could be short circuited a fine copper wire 
 was soldered to the plugs, which could be connected to small screws 
 near the ends of the shaft. 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

23 
 
 TABLE I 
 
 SHOWING EFFECT OF ROTATION ON RESISTANCE OF THE COIL 
 
 Resistance of Coil 
 at Rest 
 
 235.87S04 ohms 
 235. 88446 
 235.92873 
 Mean 
 235.8970 
 
 Resistance of Coil 
 Rotating Speed 7188 RPM 
 
 Simultaneous Observation 
 of e. m. f . in volts 
 
 None 
 
 less than -.000005 volts 
 less tnan .000005 
 
 Observat ions 
 of e.m.f. 
 
 +.000015 volts 
 . 000015 
 .000007 
 . 000006 
 . 000015 
 
 236.218345 ohms 
 236.216986 
 236.317893 
 236.21699 0 
 236.221055 
 Mean 
 
 236.318073 
 
 Increase in resistance of coil when rotating and at rest 
 
 .3210 ohms. 
 
 Table I includes some of the results showing the resistance 
 while at rest and while rotating at a speed of over 7000 RPM. From 
 the data it may be seen that the total change in resistance is .321 
 ohm. The e.m.f. which was produced while the coil was rotating is 
 shown also. This was measured by means of a Wolff potentiometer. 
 
 The resistance and the potential could be found almost simultaneous!; 
 by throwing a switch, making connections first with the W r clff poten- 
 tiometer and then with the Carey Foster bridge. The reason for nctilg 
 

 
 
 * 
 
 , 
 
 . 
 
 . 
 
 . • . 
 
 . 
 
 
 
 
 
 
 
24 
 
 the e.m.f. was to have a cneck on any thermo-electric forces which 
 might he present. 
 
 There seems no direct relation between this change of resist- 
 ance and the small e.m.f. which was detected. It is thought that 
 these were largely due to contacts outside of the disk for similar 
 effects were noted wnen the coil was short circuited. In order to 
 determine wnat influence these electromotive forces might have on 
 the resistance, the current in the Carey Foster bridge was reversed; 
 the effect of this is shown in Table II. 
 
 TABLE II 
 
 SHOWING EFFECT OF REVERSING BATTERY OF CAREY FOSTER BRIDGE 
 
 Resistance of Coil at Re3t 
 
 Current Direct Current Reversed Difference e.m.f. Observed 
 235,92873 ohms 235.92873 ohms 0 0 
 
 Resistance of Coil Rotating 7188 R. P.M. 
 
 236.21834 ohms 236.21608 ohms .00226 .000015 volt 
 
 236.21789 ohms 236.21699 ohms .00090 .000015 volt 
 
 Mean .00155 ohm 
 
 Since the e.m.f. was always in the same direction this snows 
 that the maximum change in resistance due to this source is only 
 .00226 ohm and that this is actually twice as great as tne true ef- 
 fect, for here a difference was taken between the resistance plus 
 the effect ana the resistance minus the effect of the e.m.f. The 
 
 true value for the maximum value observed is only .00113. 
 
 To make further correction the resistance due to the mercury 
 contact must be considered. The actual resistance of the system 
 when the coil is cut out by tne short circuiting device is .20371 
 ohm wnen at rest, and .213099 ohm when tne wneel is rotating. 
 

 
 
 
 
 . . . 
 
 
 
 
 
 
 
 
 
 
25 
 
 showing that a change of .010828 ohm is due to the contact alone. 
 
 Deducting tnese two measurable effects from the observed change 
 in resistance tne data may be summed up as follows: 
 
 Gross change in resistance .3210 
 
 Change due to e.m.f. .00155 
 
 Change due to contacts . 010828 
 
 .012378 .0 12378 
 
 Corrected change in resistance .309622 chm 
 
 which is a change of 1.38$ of the total resistance of the coil. 
 
 While these observations were being made with tne Advance wire 
 coil, the original copper coil was Changed and the connections made 
 independent of the brass parts. It was found, however, that this wan 
 so sensitive to temperature changes due to the high temperature 
 coefficient of copper, that no reliable results of the effect of ro- 
 tation could be obtained. When it is considered that the linear ve- 
 locity of the coil is about 225 miles per hour, or 105 meters per 
 second, the tremendous amount of air friction may be realized and it 
 is thought that the greater part of the change in resistance (over 
 4 ohms was observed) was largely due to a heating effect. That ther< 
 was no thermo-electric effect was shown from the fact that the elec- 
 tromotive force measured with the potentiometer was very slight. 
 
 The elimination of this temperature effect is very desirable 
 and it may be possible, though at present it offers the chief diffi- 
 culty in the determination of the effect of rotation on the resist- 
 ance of pure metals. 
 
 One ppssible source of error whose weight could not be deter- 
 mined was, that though all precautions were taken to make the wire 
 coil rigid, there is a possibility that the disk as a whole expanded 
 

 
 . . . 
 
 . 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 I 
 
 . 
 
 
 
 
 
 
36 
 
 thus causing a stretching of the wire which would undoubtedly in- 
 crease the resistance. But with the disk, made up as it is with hea\|Jr 
 brass plates on each side for a quarter of the diameter, and finally 
 bound with outer rims of sealing wax and hickory, it hardly seems 
 possible that there is any appreciable stretch of the disk. Tests 
 must, however, be made for any possible stretching effect. 
 
 VI. SUMMARY AND CONCLUSIuNS 
 
 1. A method has been devised by which practically constant 
 electric connections can be made with a coil revolving at a speed of 
 over 7000 RPM. The change in resistance from rest to full speed 
 due to this contact is a very small amount,- in this case only .0108 
 ohm. 
 
 2. The thermo-electric forces produced in the system have 
 been reduced to a very small, practically negligible quantity. In 
 the final results 1 /3$> of the total change of resistance could be 
 traced as possibly due to thermo-electric forces. 
 
 3. It has been shown that the resistance of a coil of Advance 
 wire is increased by rotation at a speed of 7000 RPM. The change of 
 resistance When allowance is made for other known effects is .309 
 ohm, or 1.38% of the total resistance. The centrifugal acceleration 
 in this case is 763,000 cm. per sec. per sec. 
 
 4. Experiments, similar to those made with " Advance" wire, 
 with a copper wire showed temperature effects too great to allow any 
 deductions as to the effect of rotation alone. The elimination of 
 this temperature effect is the next step, as it is very desirable 
 
 to get results with a pure metal. 
 
 5. It is not easy to analyze the results of this experiment 
 

 
 
 
 
 
 
 
 
 
 
 * 
 
 . 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 . 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 ; 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 . 
 
 
 
 . 
 
 . 
 
 
 
 
 
 
 
 
27 
 
 as to its bearing upon the theory of metallic conduction, because it 
 does not point conclusively to any one of the tneories advanced. The 
 results are, however, more easily explained on the "free gas" or the 
 "doublet" theory than on the "gap” theory. 
 
 a. If we are to accept the free electron tneory, tnen tne 
 change of resistance might be explained by assuming that the elec- 
 trons are thrown to the outer portion of the conductor, thus produc- 
 ing a crowding or condensation of tne electrons, which would cause 
 an increase in tne number of collisions or the friction between the 
 electrons and tne molecules. 
 
 b. According to tne doublet theory suen a force might tend to 
 depolarize the dipoles or to throw tne chain of doublets out of line 
 which would cause tne resistance to become greater. 
 
 c. If a stretching effect of the wire could be detected then 
 such a change in resistance might be explained from the "gap» tneory 
 whicn seems to cover quite extensively tne cnanges of resistance due 
 to cnanges in the structure of a metal. But the possioility of a 
 stretching of tne coil, as already noticed, seems to be very small. 
 
 If the increase is due to a pure rotation effect then it is hard to 
 see how it can be explained on any "gapn theory. 
 
 The fact, that tne results obtained are entirely from "Advance* 
 wire which is an alloy and nence involves factors not present in 
 pure metals, prevents definite conclusions as to tne effect of a 
 similar experiment upon a pure metal. 
 
 The writer wisnes to thank Professor A.P. Carman for his int- 
 erest and encouragement in the experimental work ana the valuable 
 assistance given in the preparation of this manuscript. 
 

 
 
 
 
 . 
 
 . 
 
 
 
 . 
 
 . 
 
 . 
 
 . 
 
 
 
 
 
 ' 
 
 . 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 . . 
 
 . 
 
 
1. Bayley, P.K. 
 
 Electron Theories of Metallic Conduction of Electricity. 
 
 Thesis in University of Illinois Library, Filed June, lyi4. 
 
 2. Bridgeman, P.W. 
 
 Theoretical Consideration on the Nature of Metallic Resist- 
 ance with Especial Regard to the Pressure Effect. Physical 
 Review, Vol.9, p 269, April, 191?. 
 
 3. Bndgeman, P.W. 
 
 The Electrical Resistance of Metals Under Pressure. Proc. 
 Amer. Acad. Arts & Sci. , 52, 9, p 573, Feb. ,1917. 
 
 4. Bndgeman, P.W. 
 
 Thermo-Electromotive Force, Peltier Heat, and Thomson neat 
 Under Pressure. Proc. Amer. Acad. Arts and Sci. , 55,4, 
 p 269, Mar. ,1918. 
 
 5. Bridgeman, P.W. 
 
 Electrical Resistance Under Pressure, Including Certain 
 Liquid Metals. Proc. Amer. Acad. Arts & Sci., 56, 3, p 61, 
 Feb. , 1921. 
 
 6. Bridgeman, P.W. 
 
 The Electrical Resistance of Metals. Physical Review, 27, 2, 
 p 161, Feo.,1921. 
 
 ?. Credner, F. 
 
 Alteration of Electrical Resistance of Wires When Subjected 
 to Tension or Torsion. Zeitscnr. Pays. Chem. , 82, p 457,1913. 
 8, Drude, P. 
 
 The Ionic Theory of Metallic Conduction. Phys. Zeirscnr. 1, 
 p 161-165, Jan. ,1900. 
 

 
 
 
 
 
 - 
 
 
29 
 
 9. Drude, P, 
 
 Electron Tnecry of Metals. Ann. d. Phys. u. Cnem. , o, p 569, 
 Nov. ,1900. 
 
 10. Giese, W. 
 
 Theorie aer Electric lt&tsleitung, Ann. d. Phys. u. Chem. , 3, 
 p 576, 1889. 
 
 11. Gr&neisen, E. 
 
 Temperature ana Pressure. Vern. D.P. Geseil. , 15, p 186,1913. 
 
 12. hall, E.Jti, 
 
 Tne Resistance of Moving Electrolytes. Phys. Rev., 7, p 246-7, 
 1898. 
 
 13. Jaeger, V/. and Dieselhorst, H. 
 
 Conductivity for heat and Electricity. Preuss. Akad. Wiss. , 
 Berlin, 38, p 719-726, July, 1899. 
 
 14. Kensky, T. 
 
 Conduction of Electricity in Metals. Zeitschr. Electro-Chem. , 
 14, p 406-410, 1908. 
 
 15. Koenigsberger , T. 
 
 Conduction of Electricity. Zeitschr. Electro-Cnem. 15, p 97, 
 1909. 
 
 16. Kraus, C.A. 
 
 The Temperature Coefficient of Resistance of Metals at Con- 
 stant Volume and its Bearing on tne Theory of Metallic Con- 
 duction. Phys. Rev., 2, 4, 159, 1914. 
 
 17. Livens, G.H. 
 
 The Electron Theory of Metallic Conduction. Phil. Mag. , 29, 
 pp 173 and ^t25-452. 
 
 18. Lees, C.H. 
 
 The Electronic Theories of the Properties of Metals. Nature, 
 

 
 
30 
 
 95, 675-577, 1915. 
 
 19. Nabl, J. 
 
 Resistance of Electrolytes in Motion. Akad. Wise. Wien. Sitzt. 
 108, Ila, p 1538-1558, 189y. 
 
 20. Nichols, E.F. 
 
 Die Moglichkeit einer durch Zentrifugale Beschleunigung er- 
 zengten electromct onschen Kraft. Phys. Zeitschr. p.640, 1908. 
 
 21. Richardson, O.W. 
 
 Theory of Metallic Conduction. Amer. Electro-Chem. See. Trans. , 
 21, p 69, 1912. 
 
 22. Reinganum, O.W. 
 
 Theory of Metallic Conduction. Ann. d. Phys. u. Chem. , 2, 2, 
 p 398-403, 1900. 
 
 23. Riecke, E. 
 
 Thermal and Electric Conductivities. Ann. d. Phys. u. Chem. , 
 
 2, 4, p 835, 1900. 
 
 24. Riecke, E. 
 
 Electric Conduction in Metals. Phys. Zeitschr. 2, p 639, 
 
 Aug., 1901; Ann. d. Phys. u. Chem., N.F.,66, p 543, 1898. 
 
 25. Schenck, R. 
 
 Electron Theory. Studies in Metallic Solid Solutions. 
 
 Ann. d. Physik, 62, 2, p 261-290, May, 1910. 
 
 26. Skaupy, F. 
 
 Electric Conduction in Metals. Verh. d. D. P. Gesell. 16, p 
 156-167, F eh. 1914. 
 
 27. Swann, W.F.G. 
 
 Metallic Theory of Conduction. Phil. Mag. , Ser.6, 27, pp 441- 
 454, 1914. 
 

28. Starling, S. G. 
 
 Electricity and Magnetism. Chapter XVII, p 549. 
 
 29. Thomson, J.J. 
 
 Conduction of Electricity Through Metals. Phil. Mag. , 30, p 
 192, 1915. 
 
 30. Thomson, J.J. 
 
 The Conduction of Electricity Through Metals. Nature, 96, 
 pp 494-499, Dec. 1915. 
 
 31. Thomson, J.J. 
 
 Corpuscular Theory of Matter. Chapter IV. 
 
 32. Telman, R.C. and Stewart, T.D. 
 
 Electromotive Force Produced by Acceleration of Metals. 
 Phys. Rev., 43, Ser.2, 8, p 9?, Aug. 1916. 
 
 33. Tunzelman, G.W. 
 
 A Treatise on Electrical Theory. Chapter XIV. 
 
 34. Weber, W. 
 
 Weber’s Werke, p 294. 
 
 35. Wiedeman, G. 
 
 Die Lehre der Electric irat , 4, p 825.