ALTERNATE-CUEREp, ; ; /, 
 
 MACHINERY. 
 
 GISBERT KAPP, Assoo: M. INST. C. E. 
 
 Reprinted from the Minnies of Proceedings of the Insti- 
 tution of Civil Engineers, London. 
 
 NEW YORK: 
 
 D. VAN NOSTRAND COMPANY, 
 28 MURRAY STREET AND 87 WARREN STREET. 
 
 1889. 
 
: -.-.:: 
 
PREFACE. 
 
 The subject comprised under the title 
 of this monograph naturally divides 
 itself into the following sub-sections : 
 1. Alternators ; 2. Transformers ; 3. Mo- 
 tors ; 4. Meters ; 5. Mains ; and 6. 
 Accessory apparatus for use in central 
 stations and on the premises of the per- 
 sons supplied with current from such 
 stations. The question of lamps does 
 not properly belong to the subject under 
 consideration, because glow lamps are 
 equally suitable to be fed by alternating 
 and direct currents, and the alterations 
 required to make arc lamps work with 
 alternating currents are easily applied, 
 and present no special interest;. In the 
 present volume, the author deals more 
 especially with the first three subjects, 
 namely, alternators, transformers, and 
 motors. 
 
x*^ 
 
 nriiTiftsiTin 
 
 \4^v&$r 
 
 ALTERNATE-CURRENT MACHINERY. 
 
 BY GISBERT 'K&PP, Assoc. M. INST. C. E. 
 
 ALTERNATORS. 
 
 The theory of alternate-current ma- 
 chines, as given in modern text books, is 
 based upon the assumption that the cur- 
 rents are generated in wire coils, the 
 magnetic induction through which under- 
 goes periodical changes according to a 
 simple sine-function. A machine of this 
 character would be represented, in its 
 most elementary form, by a coil of in- 
 sulated wire revolving round an axis in 
 its own plane, with uniform velocity in a 
 uniform magnetic field, the axis of rota- 
 tion being at right angles to the lines of 
 the field. Fig. 1 represents such a ma- 
 chine ; a rectangular coil A B is placed 
 in the field produced by the polar sur- 
 faces N S, and if these surfaces extend 
 
6 
 
 beyond the space occupied by the coil, 
 the field within that space may be con- 
 sidered to be uniform. If the axis of 
 rotation passes midway between the sides 
 A and B of the rectangle, the electro- 
 motive forces in both branches of the 
 coil are added, and the total electro- 
 motive force at any moment is according 
 to a well-known formula 
 
 r 
 
 e = 2 7t n z - sin a (1), 
 
 2 
 
 where n is the frequency (number of 
 complete periods per second), T is the 
 number of wires in both branches A and 
 B of the coil collectively, z the total 
 induction through the coil when the lat- 
 ter stands at right angles to the lines of 
 force, and a the angle between this posi- 
 tion and the position at which the elec- 
 tromotive force e is generated. The re- 
 lation between a and , the time which 
 was required to rotate the coil through 
 this angle, is a = 2 7t n t. The total 
 electromotive force changes periodically 
 from a positive maximum through zero 
 
to a negative maximum and back again, 
 and the numerical value of these maxima 
 
 r 
 is 2 n n z - (2). For practical purposes 
 
 2 
 
 this value is, however, of less importance 
 than the mean value of the electromotive 
 
force ; for on the latter depends the 
 amount of work obtainable from the 
 apparatus. By mean value of the elec- 
 tromotive force of an alternating current 
 is meant the electromotive force of a 
 direct constant current, which vvill, in a 
 given resistance free from self-induction 
 (say for instance, the wire of a Cardew 
 voltmeter), produce the same amount of 
 heating. It is necessary that the resist- 
 ance should be comparatively large, in 
 order to limit the current to such a 
 small value that the disturbing effect of 
 self-induction may be neglected. Let T 
 be the periodic time, then e, the mean 
 value of the electromotive force, is found 
 from the equation 
 
 T 
 
 2" 2 
 
 f -/2 nn z r sin 2 n n t\dt = _ ~e\ 
 
 when R denotes the resistance of the 
 Cardew voltmeter. 
 
 The solution of this equation is a 
 simple mathematical operation, and need, 
 therefore, not be given at length. The 
 result is 
 
e = _ 11 ZT (3) 
 
 ^2 
 
 It is_convenient to compare this value 
 with that representing the electromotive 
 force of a direct- current machine having 
 the same field and external number of 
 wires r (but equally distributed) on the 
 armature. The electromotive force of 
 such a machine is given by the product 
 
 N 
 
 z r - - , where N, the number of revolu- 
 60 
 
 tions per minute, may obviously be re- 
 placed by 6(M when the machine has 
 two poles, because, in this case, fre- 
 quency and number of revolutions per 
 second are identical. If the direct-cur- 
 rent machine has more than one pair of 
 poles, say, for instance, p pairs, then it 
 is possible to wind the armature in such 
 a way that the electromotive forces due 
 to the different pairs of poles are added, 
 so that the total electromotive force is p 
 times the above value, and the same may 
 obviously be done with the similar ma- 
 chine wound for alternating currents so 
 
10 
 
 that in all cases the ratio between the 
 electromotive forces of the two machines 
 
 remains the same, namely, (4\ Of 
 
 V 2 
 
 two machines containing the same 
 amount of material, the one which is 
 wound to give alternating currents will 
 therefore produce about two and a quar- 
 ter times the electromotive force of that 
 wound for continuous current. Since 
 the halves of the armature-conductor in 
 a direct-current machine are grouped in 
 parallel, the strength of current, which 
 such a machine gives when the current 
 density is the same as in the armature of 
 an alternator of equal dimensions, is 
 twice that of the latter, so that the out- 
 put of the alternator will only be about 
 12 per cent, greater than that of a direct- 
 current dynamo of equal weight. This 
 result, it should be remembered, has 
 been obtained on the basis of the theory 
 of alternators as found in text-books, 
 whilst neglecting self-inductions, dis- 
 tortions of field, and other disturbing 
 elements ; and, before it can be accepted 
 
11 
 
 for actual practical work, it is necessary 
 to inquire whether the basis of this 
 theory is in accord with the constructive 
 data to be found in modern machines of 
 this class. Alternators as now con- 
 structed are, with a single exception, 
 multipolar machines. This exception is 
 the high-speed dynamo of the Hon. 
 Charles A. Parsons, Assoc. M. Inst. C. E., 
 which can be converted into an alterna- 
 tor by some slight changes in its arma- 
 ture and commutator ; but all the other 
 machines now in practical use have more 
 than one pair of poles, which are set in a 
 circle, and are so arranged that the coils 
 of the armature sweep past them. As 
 the interpolar space is always very small, 
 the density of lines, emerging from or 
 entering the polar surfaces, may be taken 
 as fairly uniform over the extent of these 
 surfaces. It thus becomes possible to 
 determine for every shape of coil the 
 configuration of polar surface, which will 
 produce that variation in the induction 
 corresponding to the simple sine-func- 
 tion, taken as the basis of the text-book 
 
12 
 
 theory of alternators. In the most 
 simple and most frequently occurring 
 case of a coil with straight sides, it is 
 obvious that the dimension of the pole- 
 piece, which is parallel to the active part 
 of the coil, would have to vary according 
 to a sine-function as shown in Fig. 2, 
 where ABCD represents the coil, and PP 
 the mean pitch line through the poles 
 straightened out. The shaded areas N 
 S are the poles, which are shown on 
 opposite sides of the pitch-line to make 
 their sinoidal contour at once apparent. 
 Poles of this shape are, however, not 
 used in practice ; the poles are either 
 trapezoidal, circular, or rectangular. In 
 the Mordey alternator the poles have a 
 trapezoidal face, and those succeeding 
 each other on the same side of the 
 armature are of the same sign, and are 
 placed in line with those of opposite sign 
 on the other side of the armature. Thus 
 along the mean pitch line, uniform fields 
 alternate with spaces in which there is 
 no induction; this arrangement when 
 straightened out may be represented by 
 
13 
 
 a succession of rectangles, as in Fig. 3, 
 placed above and below the line PP, the 
 height of the rectangles representing 
 half the induction between the poles. 
 The analogue arrangement of poles in a 
 direct-current machine is shown in Fig. 4, 
 where a cylindrical armature, provided 
 with an iron core, is completely sur- 
 rounded by its pole-pieces. In the latter 
 class of machines the exact shape of the 
 poles is only in so far of importance as 
 it affects the leakage of magnetism, but 
 for a given total induction through the 
 armature the shape has no influence on 
 the electromotive force. Hence for a 
 direct- current machine, apart from the 
 question of leakage, Fig. 3 is as good an 
 arrangement as Fig. 1, but for an alter- 
 nator it is not so good. A simple calcu- 
 lation, which need not be reproduced 
 here, shows that the electromotive force 
 of the arrangement represented by Fig. 3 
 is 
 
 e=Znzr (5). 
 
 In this case the electromotive force is 
 therefore only twice that of the direct- 
 
14 
 
 current machine, and as, owing to the 
 concentration of armature windings into 
 one coil, the limit of current as deter- 
 mined by heating will be reached sooner, 
 the current will be somewhat smaller 
 than half that of the direct-current ma- 
 chine, so that the output will be slightly 
 less. It must also be borne in mind that 
 a rectangular field with sharp corners 
 cannot be obtained. In consequence of 
 leakage and spreading of lines, the cor- 
 ners of the field will be more or less 
 rounded, as shown by dotted lines, and 
 this circumstance must lower the elec- 
 tromotive force beyond its theoretical 
 value. On these grounds it would there- 
 fore appear that this configuration of 
 field-magnets is not very advantageous, 
 but there are certain other considerations 
 which tend to modify this conclusion. 
 In actual machines it is not a single coil 
 which has to be dealt with, the active 
 wires in which occupy only a very nar- 
 row space, but wide coils covering a con- 
 siderable extent of surface on the arma- 
 ture ; there is also the question of self- 
 
15 
 
 induction to be taken into account, and 
 the more or less economical way of pro- 
 ducing the field. It is especially with 
 
 regard to the latter consideration that 
 alternators of this type (of which the 
 Kennedy and Mordey machines are 
 examples) have a considerable advantage, 
 
16 
 
 as their fields can be made on the " iron- 
 clad principle," so that no magnetism is 
 lost in external leakage, and only one 
 coil of exciting wire is required. 
 
 The type of field most commonly used 
 in modern alternators is, however, one 
 in which N and S poles succeed each 
 other around the armature, and in its 
 simplest form such a machine is repre- 
 sented by Fig. 5, where an iron-cored 
 cylindrical armature revolves between 
 poles which do not completely surround 
 it. The portion of the armature not 
 embraced by the poles may be taken as 
 equal to that embraced, and Fig. 6 shows 
 this disposition in a multipolar machine. 
 PP is the mean pitch-line of the poles 
 straightened out, and the poles are rep- 
 resented by the shaded rectangles which 
 have a width equal to half the pitch. 
 The mean electromotive force of this 
 arrangement is 
 
 e = JL# z T (6), 
 
 V* 
 or 183 per cent, greater than that of a 
 
17 
 
 direct-current machine containing the 
 same amount of material, whilst the cur- 
 rent, owing to the concentration of wires, 
 and consequent greater liability to heat, 
 is something less than one-half. The 
 output of this machine will, therefore, be 
 something less than 40 per cent, greater 
 than that of a direct-current machine 
 having an equal weight. It must not be 
 forgotten that the results here obtained 
 refer to machines in which the active 
 wires on the armature are concentrated 
 into single lines. This is of course 
 impossible in practice, since the wires 
 must be of appreciable thickness, and 
 must therefore occupy a considerable 
 space. An approximation can be made 
 to the theoretical condition by employing 
 an iron-cored armature with projecting 
 teeth, the coils being laid into the re- 
 cesses. As, however, armatures of this 
 type tend to produce heating of the pole- 
 pieces, if the latter be solid, the arrange- 
 ment is not generally used ; and when 
 the armature is provided with an iron 
 core, the surface of this is in most ma- 
 
18 
 
 chines perfectly smooth. The wire coils 
 on the armature, whether this contain 
 iron or not, must therefore occupy a con- 
 siderable space, and this circumstance 
 will, generally speaking, cause the elec- 
 tromotive force obtainable with the dif- 
 ferent forms of fields now considered to 
 be lower than the values given. The 
 reason for this reduction is that, with a 
 wide coil not all the turns are equally 
 and simultaneously influenced by the 
 field, and in certain positions the action 
 is differential. The calculation of the 
 effective mean electromotive force for any 
 configuration of field and armature coils 
 presents no special difficulty, being in 
 fact only an extension of the methods 
 already indicated, though for certain 
 complicated forms the operation is 
 laborious. The author only proposes to 
 deal with some simple cases, which may 
 be regarded as forming the limiting con- 
 ditions of practical designs. As regards 
 the field these cases are : 
 
 (1) Eectangular field, width of poles 
 equal to the pitch, as shown in Fig. 4. 
 
19 
 
 (2) Rectangular field, width of poles 
 equal to one-half the pitch, as shown in 
 Fig. 6. 
 
 (3) Rectangular field, width of poles 
 equal to one-third of the pitch. 
 
 As regards the armature, either the 
 whole available space may be covered 
 with coils, in which case the width of 
 each group of conductors forming a coil, 
 or one side of a coil, will be equal to the 
 pitch, and the armature will contain the 
 greatest possible amount of wire ; or, 
 only portions of the available space on 
 the armature may be covered with the 
 active conductors, which plan is generally 
 adopted in modern alternators, because 
 by it the differential action alluded to is 
 diminished. The cases selected for con- 
 sideration are the following : 
 
 (1) Width of coils equal to the pitch. 
 
 (2) Width of coils equal to one-half 
 the pitch. 
 
 (3) Width of coils equal to one-third 
 of the pitch. 
 
 The calculation of the mean electro- 
 motive force, for any combination of field 
 
20 
 
 and armature of the kind here indicated, 
 is a very simple mathematical operation, 
 and it is not necessary to give it in 
 detail. It will suffice to give the result 
 for those combinations which most 
 nearly resemble the conditions found in 
 actual practice. As has been shown, the 
 electromotive force of an alternator can 
 be expressed as the product of the elec- 
 tromotive force of a direct- current ma- 
 chine, having the same dimensions and 
 weight, and a certain coefficient depend- 
 ing on the configuration of the field- 
 magnets and armature-coils. The wind- 
 ing and other constructive data of an 
 alternator being known, it is only neces- 
 sary to find this coefficient, in order to 
 determine the electromotive force, by 
 reference to the electromotive force 
 which would be produced by the same 
 machine if the armature-coils were con- 
 nected in such a way as to give a contin- 
 uous current. In a paper by the 
 author,* read before this Institution on 
 
 * Minutes of Proceedings Inst. C. E., vol. Ixxxiii. 
 p. 123. 
 
21 
 
 the 24th of November, 1885, it was 
 shown that the electromotive force in 
 volts of an ordinary dynamo is given by 
 the formula : 
 
 e = z r N 1CT 6 (7), 
 
 where z represents the total induction in 
 English lines of force (each equal to 
 6,000 C. G. S. lines), r the total number 
 of active wires counted all round the 
 armature, and N the speed in revolutions 
 per minute. If the machine is multi- 
 polar, and if the different armature-cir- 
 cuits are arranged in series, the expres- 
 sion for e must be multiplied by p, the 
 number of pairs of poles in the field. 
 Let k be the coefficient which expresses 
 the ratio between the electromotive 
 forces of the alternate and continuous- 
 current machines, then the formula for 
 the former is 
 
 e = kpzTNW~~ G (8) 
 
 if the total induction be given in English 
 lines of force. If in C. G. S. units the 
 electromotive force is 
 
 e = kpz r?- 10~ 8 . 
 60 
 
22 
 
 The following table gives the value for 
 k for different cases, all referring to poles 
 of rectangular shape and coils in which 
 the active wires are straight : 
 
 1. Width of poles equal to pitch, ) 
 toothed armature and winding > k = 2.000 
 concentrated in the recesses . . ) 
 
 2. Width of poles equal to pitch, \ 
 
 smooth armature and winding v k = 1. 160 
 spread over the whole surface . ) 
 
 3. Width of poles equal to pitch, ^ 
 
 smooth armature and winding I k = 1.635 
 covering only one-half the sur- ( 
 face J 
 
 4. Width of poles equal to half the ) 
 
 pitch, smooth armature and ! k = 1.635 
 winding spread over the whole ( 
 surface J 
 
 5. Width of poles equal to half the") 
 
 pitch, smooth armature and ! k 2.300 
 winding covering only one-half ( 
 the surface J 
 
 6. Width of poles equal to one-third ] 
 
 the pitch, smooth armature and \k = 2.830 
 
 winding covering only one-third 
 
 I- 
 
 of the surface J 
 
 According to the ordinary sine-formula 
 the coefficient is k = 2.220; and this 
 agrees fairly well with case 5, which is 
 the most frequently met with in actual 
 practice. The formula for e presupposes 
 
23 
 
 the knowledge of the total induction z, 
 which depends upon the shape and size 
 of the field magnets, their arrangement, 
 the amount of iron in the armature, the 
 interpolar space, and the exciting power. 
 The time has long passed when a con- 
 sideration of the relation between excit- 
 ing power and total induction would 
 have been of interest. This subject has 
 been fully treated by Drs. J. and E. 
 Hopkinson before the Royal Society,* 
 and the author has also dealt with it in a 
 paper on " The Predetermination of the 
 Characteristics of Dynamos," before the 
 Society of Telegraph-Engineers and Elec- 
 tricians, f These methods have subse- 
 quently been improved, by the introduc- 
 tion of certain terms due to Professor 
 Forbes, by means of which the magnetic 
 leakage can be more accurately deter- 
 mined. With the knowledge of the total 
 induction 2, and the coefficient k, it is 
 thus possible to determine the relation 
 
 * " Dynamo-Electric Machinery," by Drs. J. and 
 E. Hopkinson Phil. Trans. Royal Society, 1888, p. 331. 
 t Journal, vol xv. 1886, p. 518. 
 
24 
 
 between exciting-power and electromotive 
 force of any given alternator, provided 
 the current flowing through its armature 
 is not so great as to produce a sensible 
 self-inductive effect ; in other words, the 
 formula renders it possible to determine 
 the static, but not the dynamic, charac- 
 teristic. The effect of self-induction is 
 to produce an electromotive force, the 
 phase of which is at right angles to that 
 of the current ; and the electromotive 
 force available for doing work in the cir- 
 cuit is the resultant of the induced elec- 
 tromotive force and that due to self- 
 induction. The mathematical treatment 
 of the problem is of great difficulty, not 
 only because, with the shape of poles and 
 armature coils occurring in modern alter- 
 nators, the electromotive force is a very 
 complicated function of the time, but also 
 because the coefficient of self-induction 
 is not a constant, but varies with the 
 relative position of the coils and poles. 
 Fortunately the coefficient is compara- 
 tively small. In machines containing no 
 iron in the armature, such as the Mordey 
 
25 
 
 and Ferranti, it is almost negligible ; in 
 machines containing iron in the arma- 
 ture, which is not encircled by the coils, 
 such as the Westinghouse and Lowrie- 
 Parker, it is still very small ; and even in 
 machines in which the coils are wound 
 round the armature-core, as in Kennedy's 
 machine and that designed by the author, 
 the coefficient, although appreciable, is 
 yet not so large as to cause 'a sensible 
 error being introduced by the assump- 
 tion that it is a constant. 
 
 The method originally devised by Mr. 
 Joubert, for taking the self-induction of 
 an alternator into account, requires the 
 solution of a differential equation, but 
 the problem can be treated graphically 
 in a much more simple manner. Both 
 methods have this in common, that it is 
 assumed that the current and all elec- 
 tromotive forces are sine-functions of the 
 time. Strictly speaking, this is not cor- 
 rect, but the error is probably not very 
 great, and it is in a certain sense un- 
 avoidable ; because if this assumption 
 were discarded, and each case treated on 
 
26 
 
 the basis of the exact induction-curve as 
 mapped out from the shape of the poles 
 and armature- coils, it would lead to such 
 difficulties as to make the investigation 
 useless for practical purposes, for which 
 easily understood and simple methods of 
 calculation or construction are wanted. 
 The sine-function has, therefore, as far as 
 the author is aware, always been taken 
 by previous investigators as the basis of 
 their calculations, and in this he pro- 
 poses to follow 7 their example. 
 
 Let, in Fig. 7, the length of the line O 
 I represent the maximum value of the 
 current (crest of the current wave), and 
 let this line revolve in the direction of 
 the arrow around O as the center, with 
 an angular velocity w == 2 n n, then the 
 projection of O I upon the vertical gives 
 the current at any moment. Let the 
 line O L represent the maximum electro- 
 motive force of self-induction due to the 
 current O I, then the projection of O L 
 will similarly give the electromotive force 
 of self-induction at any moment, the two 
 lines preserving during rotation their 
 
27 
 
 rectangular position. The effective elec- 
 tromotive force which produces the cur- 
 rent O I, through a resistance having no 
 self-induction, may be represented by a 
 length O E drawn to the same scale as 
 O L, and since the effective electro- 
 motive force must be the resultant of the 
 induced electromotive force and the 
 
 electromotive force of self-induction, the 
 former can be found by constructing the 
 parallelogram OLE E t , the line O E t - 
 giving the induced electromotive force 
 both as regards magnitude and position. 
 If the length O E represents the elec- 
 tromotive force lost on account of the 
 resistance of the armature, the remaining 
 
28 
 
 portion R E represents the electromotive 
 force actually available for doing work 
 through the resistance of the external 
 circuit ; and this will the more nearly 
 approach the induced electromotive force 
 the smaller the armature resistance, and 
 the smaller the self-induction of the 
 whole circuit. In order to obtain a 
 maximum of work in the external circuit 
 with an alternate-current plant, it is, 
 therefore, advantageous that the armature 
 of the alternator should have a minimum 
 of resistance, and that the total self- 
 induction of the circuit should be small. 
 The commercial value of a given type of 
 alternator depends, amongst other things, 
 upon the amount of work which can be 
 obtained from a given weight of materials 
 employed in the construction of that par- 
 ticular type of alternator. This value 
 will, therefore, be roughly indicated by 
 the ratio between the length of the lines 
 R E and O E,- ; and this ratio the author 
 proposes to call "plant-efficiency.'' It 
 must be borne in mind that the plant- 
 efficiency is not an indication of the more 
 
29 
 
 or less perfect way in which the alterna- 
 tor transforms mechanical into electrical 
 energy, but merely an indication of the 
 commercial value of any given type for 
 producing a maximum amount of exter- 
 nal work with a minimum weight and 
 cost of the materials employed. The 
 electrical efficiency of the machine can 
 also be found from the diagram. Accord- 
 ing to a well-known law, the internal 
 energy of the machine is equal to half 
 the product of maximum current and 
 maximum induced electromotive force 
 multiplied with the cosine of the angle 
 of lag < between the two. 
 
 Internal energy =: ^ E, I cos $, 
 when E, is the maximum value of the 
 induced electromotive force and I the 
 maximum value of the current. The ex- 
 ternal energy available for doing work is 
 the product of the mean current and 
 mean terminal electromotive force, which 
 is equal to half the product of their 
 maximum values. 
 
 External energy = E I, 
 when E is the maximum value of the 
 
30 
 
 terminal electromotive force. The elec- 
 trical efficiency is therefore given by the 
 ratio of the lines E E and O E, which is 
 obvious considering the electromotive 
 force O K, absorbed in overcoming the 
 resistance of the armature, is the only 
 purely electrical loss occurring in the 
 armature. 
 
 In alternators used for parallel distri- 
 bution, the object aimed at is generally 
 to .keep the terminal pressure constant 
 for all currents, and this can be attained 
 by working with a constant field and 
 suitably varying the speed, or by main- 
 taining a constant speed and adjusting 
 the exciting current so as to suitably 
 vary the electromotive force. The latter 
 plan must be adopted if several machines 
 are worked in parallel; and in this case 
 the diagram, Fig. 7, may be used to 
 determine the limits between which the 
 exciting current must be varied, to keep 
 the terminal electromotive force constant 
 for all loads from no current to the 
 greatest current the armature is intended 
 to pass. The mean induced electromo-. 
 
31 
 
 tive force e is found by Formula 1 and 
 its maximum value is obviouly , 
 
 E; = V"2 e (9). 
 
 The maximum value of the electromotive 
 
 force of self-induction is given by 
 
 Es = 2 n n L I . . . ." . . . (10), 
 
 where L is the coefficient of self-induc- 
 tion, which can be determined by various 
 well-known methods ; or by running the 
 armature in a constant field, and measur- 
 ing the terminal electromotive force with 
 and without a current passing through 
 the armature. The electromotive force 
 lost in resistance is simply the product 
 of the current I and the resistance K of 
 the armature. With the knowledge of 
 these quantities it is now easy to deter- 
 mine the strength of field required for 
 any current, since, according to Formula 
 1, the field is proportional to the induced 
 electromotive force. For this purpose 
 the diagram, Fig. 7, may be used, or the 
 modified form shown in Fig. 8. Here O 
 A represents the terminal electromotive 
 force, which is to be kept constant. A 
 
32 
 
 B represents to the same scale the loss 
 of electromotive force in the armature 
 when the machine is giving its full cur- 
 rent, and O D the corresponding electro 
 motive force of self -induction. Make B 
 C parallel with and equal to O D, and 
 describe circles over the diameters O A, 
 O B, O D, and O C ; then a straight line 
 
 revolving round O as a center, in the 
 direction of the arrow, is cut by these 
 various circles in such a way that the 
 segments represent the respective values 
 of the different electromotive forces at 
 any moment. Thus, at the time when 
 the angular position a has been reached, 
 the induced electromotive force is Oc, the 
 effective electromotive force is Ob and 
 
33 
 
 the terminal electromotive force is Oa, 
 all three being positive, that is, in the 
 direction of the current, whilst the elec- 
 tromotive force of self-induction is 
 negative, and is given by the length Od. 
 As the load decreases, the circle O D 
 becomes smaller, until, when no current 
 is allowed to flow, it has shurnk into a 
 point, whilst there is no loss due to re- 
 sistance, and the points B and C coincide 
 with A. In this case the terminal and 
 induced electromotive forces are identical, 
 and the strength of the field is a mini- 
 mum and proportional to the length O A. 
 At full output the strength of the field 
 is a maximum and proportional to the 
 length O C. Having thus found the 
 limits of the strength of field required in 
 these two cases, it is an easy matter to 
 determine, by the usual formulas con- 
 necting strength of field and exciting 
 power, the strength of the exciting cur- 
 rent for each. It will be immediately 
 apparent from the diagram, that of two 
 machines, the one having a high and the 
 other a low plafftPefficteaox, the latter 
 
34 
 
 will at full output require more exciting 
 current, and if the resistance of its field 
 be the same, this circumstance will in- 
 fluence the electrical efficiency unfavora- 
 bly. On the other hand, a sensible 
 amount of self-induction, although it 
 does reduce the plant-efficiency, is indis- 
 pensable for the safe working of alterna- 
 tors in parallel, and must, therefore, be 
 regarded as rather an advantage than 
 otherwise. 
 
 The problem of parallel working is of 
 the greatest importance for central 
 stations, as only by such an arrangement 
 of machines can absolute continuity of 
 the service and the greatest economy at 
 all times be ensured. Alternate current 
 distribution should, if possible, be carried 
 on in the same simple manner as direct- 
 current distribution, that .is to say, the 
 Dumber of machines at work should corre- 
 spond as nearly as may be to the output at 
 any time ; the addition or withdrawal of 
 machines should not even momentarily 
 interrupt the supply of current to any 
 part of the system of distribution. These 
 
35 
 
 conditions, it will easily be seen, can only 
 be fulfilled if the machines are capable of 
 working in parallel, and it is, therefore, 
 of considerable practical importance to 
 investigate the conditions under which 
 alternators may be expected to work 
 safely in parallel. This subject is closely 
 allied with that of working alternators as 
 motors, and it is, therefore, convenient 
 to consider the two problems jointly. 
 The analytical treatment of an alternate- 
 current motor is even more difficult than 
 that of an alternate- current generator; 
 but by an extension of the graphic 
 method, the question can be treated in a 
 very simple and easily understood man- 
 ner. It is well known that an alternator 
 will not start without mechanical assist- 
 ance, but that, having been started at a 
 speed approximately corresponding to 
 the frequency of the supply current, it 
 is kept in motion when this current is 
 switched on to its terminals, and it may 
 under certain conditions even give off 
 mechanical energy. The problem in- 
 volved may therefore be stated somewhat 
 
36 
 
 in this fashion: Given an alternator, with 
 excited field and running at a proper 
 speed, and a pair of terminals from which 
 any desired strength of current can be 
 obtained at a constant pressure, and hav- 
 ing a frequency approximately corre- 
 sponding to the speed of the alternator, 
 what will be the condition under which 
 the alternator will fall into step with the 
 supply-current, and what will be the 
 relation between the current passing 
 and the mechanical energy given off? 
 Also, how will this relation be affected by 
 variations in the strength of the field ? 
 Let, in Fig. 9, the circle E f represent the 
 terminal electromotive force, and the cir- 
 cle E the electromotive force induced in 
 the armature ; and assume a certain 
 current under which the electromotive 
 force of self-induction is O L and that 
 lost in resistance O R. Since both 
 quantities are proportional to the cur- 
 rent, their resultant O A will for all 
 currents form the same angle with 
 the axes, the point A being simply 
 shifted further out on the line Oa if the 
 
37 
 
 current increases. It is obvious that the 
 line O A. can also be considered as the 
 resultant of E and E$, and the position 
 of these two electromotive forces becomes 
 thus at once defined. In order that the 
 machine may give off mechanical work, 
 the induced electromotive force must 
 
 evidently be opposed to the current, that 
 is to say, it must lie to the right of the 
 vertical, whilst the terminal electromotive 
 force must lie to the left. Under these 
 conditions only one parallelogram of 
 forces is possible, namely, that de- 
 termined by the points O B A C where 
 
38 
 
 O B = E and A B = E f . The mechan- 
 ical energy given off is proportional to 
 the product O A and O D, the latter 
 quantity being the horizontal projection 
 of O B, and obviously equal to B F. 
 Now consider what will happen if 
 some of the load is taken off the ma- 
 chine. The immediate result will 
 evidently be a slight acceleration of the 
 armature, by virtue of which the maximum 
 electromotive force E is induced sooner 
 than previously. In the diagram this 
 effect is represented by the advance of 
 the point B towards B t and a correspond- 
 ing displacement of C towards C,. The 
 length O A will also be slightly altered, 
 viz., decreased until B has come into line 
 with line Oa', and further on increased 
 again. If the whole of the load betaken 
 off, B will almost coincide with B t and 
 the angle of lag & will be a maximum ; 
 but the length O A, that is to say, the 
 current, will not be materially differ- 
 ent from its previous value, as can easily 
 be seen from the diagram. Hence it 
 follows that, even when running light, 
 
39 
 
 the motor will allow a considerable cur- 
 rent to pass ; but, as the phase of the 
 current lags behind that of the terminal 
 electromotive force by nearly a quarter 
 period, the energy supplied is very small; 
 in fact, only sufficient to overcome the 
 resistance of the armature-circuit and 
 mechanical friction. With an increase of 
 load the armature will be slightly re- 
 tarded, which will bring the point B 
 further away from B x in the diagram. 
 Simultaneously the lag <P will decrease, 
 and the current which is proportional to 
 the length O A will increase, the increase 
 being slow until the point B 2 is reached, 
 and afterwards more and more rapid. 
 Since the mechanical energy given off is 
 proportional to the product of current, 
 and induced electromotive force multi- 
 plied by the cosine of the angle the latter 
 forms with the current, it may be repre- 
 sented by the area of the shaded rectangle, 
 and a series of such rectangles for 
 different positions of the point B on the 
 circle, which represents the induced 
 electromotive force, can be determined. 
 
40 
 
 In this way it is found that the mechani- 
 cal energy continues to grow as B is 
 shifted higher up on its circle, until a point 
 in the neighborhood of B 3 is reached, 
 after passing which the energy decreases 
 again. Any position of B between Bj 
 and B 3 is stable, an increase or decrease 
 of load resulting in an automatic adjust- 
 ment of the current and of the phases of 
 the electromotive forces ; but if the 
 machine be overloaded beyond the point 
 B 3 , its condition is unstable, as an in- 
 crease of load, tending to further retard 
 the induced electromotive force, must 
 necessarily reduce the capacity of the 
 machine for doing work, and thus bring 
 it out of step with the current. Here is 
 a danger from which direct-current 
 machines are exempt. The power of a 
 direct-current machine, when used as a 
 motor, increases indefinitely with the 
 current, and an overload, if it does not 
 last long enough to burn up the arma- 
 ture, does no harm; but with an alterna- 
 tor, an overload, if it lasts only a few 
 seconds, may bring the armature to a 
 
41 
 
 stand-still, and if the self-induction is 
 not sufficient, this may eventually cause 
 it to burn up. 
 
 It has been shown by Mr. H. Wilde* 
 and by Dr. J. Hopkinson,| that two similar 
 alternators can be coupled in parallel to 
 the same circuit; and Professor Adams 
 described, in a paper read in 1884 before 
 the Society of Telegraph -Engineers and 
 Electricians, J some experiments in which 
 he succeeded in driving an alternator as 
 a motor,current being supplied by another 
 alternator of similar construction. Mr. 
 "Westinghouse, in some central stations 
 in America, has also worked alternators 
 in parallel. There is thus sufficient ex- 
 perimental proof that such a method of 
 working is possible, and the only question 
 which remains to be investigated is 
 whether, for safe working, any very 
 
 * Proceedings Lit. and Philosophical Society of Man- 
 chester. Vol. viii. p. 62: and "Philosophical Magazine," 
 January, 1869. Fourth Ssries. Vol. xxxvii. p. 54. 
 
 t The Inst. C. E. Lectures on "The Practical Appli- 
 cations of Electricity." Session 1882-83. "Some Points 
 in Electric Lighting." By Dr. J. Hopkinson. 
 
 $ Journal, vol. xiii. p. 515. 
 
42 
 
 delicate adjustment of the two machines 
 is necessary. If such were to be the case, 
 it is obvious that the system would not 
 be suitable for use in central stations, 
 where the service must be carried on day 
 and night, and perferably, by a staff not 
 composed of skilled electricians. If a 
 battery of alternators be already running 
 in parallel, and it be desired to add 
 another machine, the obvious precaution 
 which an ordinary attendant would take 
 is to excite its field, and to run the 
 machine up to the proper speed before 
 joining it to the circuit. Some devices 
 have been introduced for indicating the 
 coincidence of phases by means of glow- 
 lamps, but as the use of these requires 
 careful attention and quickness of per- 
 ception on the part of the operator, it 
 would be preferable to so construct the 
 machines that absolute coincidence of 
 phase should not be required. If each 
 alternator is driven by its own engine, 
 and if the latter is provided with a good 
 governor, there will be no difficulty in 
 obtaining very approximately the right 
 
43 
 
 speed, and the only adjustment which 
 must be left to the attendant is that of 
 the exciting current. The question now 
 arises, within what limits an error in this 
 adjustment may be safe. To this ques- 
 tion the diagram, Fig. 10, gives the 
 answer. 
 
 In this diagram the circle E f repre- 
 sents as before the terminal electro- 
 motive force, and the line Oa the direc- 
 tion of the resultant of the electromotive 
 force of self-induction, and that lost in 
 overcoming the resistance of the arma- 
 ture. The coefficient of self-induction 
 depends upon the density of lines within 
 the armature core, being obviously the 
 smaller, the more nearly the core is 
 magnetized to saturation. For a reason 
 which will be explained presently, the 
 iron in the armature of alternators is 
 never magnetized to any considerable ex- 
 tent, the induction seldom exceeding 
 about 7,003 C. G. S. lines, and as at this 
 low figure the permeability is not much, 
 if at all, affected, it can be assumed, 
 without any great error, that the self- 
 
44 
 
 induction is a constant for all field in- 
 tensities which occur in practice. The 
 angle which the line Oa forms with the 
 horizontal may therefore be taken as 
 unaffected by the exciting current, and 
 the length O A as proportional to the 
 current passing through the armature. 
 
 Fir. 10 
 
 In fact, by employing a suitable scale 
 the armature current may be read 
 directly off this line on the diagram. 
 Suppose then that O A represents the 
 greatest current which may be safely 
 allowed to flow. The determination of 
 this current will generally not depend so 
 
45 
 
 much on the capacity of the machine 
 which is being switched on, as on the 
 consideration as to how much current 
 may be temporarily taken from the other 
 machines at work. A new machine is 
 only added, if the limit of output of 
 these machines is in danger of being* 
 overtaken by the demand for current in 
 the whole system ; and since the avail- 
 able margin will therefore necessarily be 
 small, the machine about to be coupled 
 up must not be allowed to act as a heavy 
 drain upon the other machines. This 
 limits the amount of current at disposal 
 for bringing the new machine into step. 
 Let this maximum current be represented 
 by the line O A. Before switching on, 
 the engine is started and run up to the 
 proper speed, in which case the only 
 work done by the engine is that of over- 
 coming its own friction, and the fric- 
 tional resistance of the machine. If, at 
 the moment of switching on, the in- 
 duced electromotive force occupies a 
 position to the left of the line Oy, there 
 will be a momentary increase of current r 
 
46 
 
 and a considerable retarding action of 
 the armature, causing it to lag so that 
 the induced electromotive force will 
 immediately be shifted back to a position 
 on the right of Oy, and, finally, as the 
 engine is supposed to give all the power 
 required for running light, equilibrium 
 will be attained by the induced electro- 
 motive force occupying a vertical posi- 
 tion, when no work is being done by the 
 current. The question now is, what 
 must be the induced electromotive force 
 to limit the current to the value of OA.? 
 This is easily found by describing round 
 A as center with radius E ( , a circle, and 
 marking its point of intersection E' with 
 the vertical Ot/. That strength of field 
 which will give the induced electromotive 
 force OE' is the least which may safely 
 be permitted. The terminal electro- 
 motive force will then occupy the posi- 
 tion OE'j, which line is parallel with and 
 equal to E'A. It should be observed 
 that there is between the terminal elec- 
 tromotive force and the current a differ- 
 ence of phase, amounting to nearly a 
 
47 
 
 quarter period. This circumstance tends 
 greatly to minimize the disturbing in- 
 fluence of the machine, which absorbs 
 current upon the other machines supply- 
 ing current to the mains. Thus the idle 
 machine takes most current at a time 
 when the mains take hardly any, and 
 when the mains take maximum current 
 that passing through the machine is 
 nearly zero. This automatic process of 
 compensation is valuable, as it allows the 
 permissible current, for bringing the new 
 machine into step, to be fixed consider- 
 ably higher than would otherwise be safe. 
 If, without increasing the field, more 
 steam be given to the engine, causing it 
 to push the armature into a more 
 advanced phase, the point E' will move 
 to the left of the vertical, and E', will 
 move to the right, whilst the current will 
 increase. As soon as E' t has passed the 
 vertical it opposes the current, and the 
 machine ceases to act as a motor, but 
 acts as a generator supplying current to 
 the mains. It is thus possible to run the 
 machine as a generator in parallel with 
 
48 
 
 the other machines, notwith standing the 
 fact that its field is considerably weaker ; 
 but it will readily be understood that the 
 working under these conditions cannot 
 be economical, though it may be per- 
 fectly safe. Now suppose that, instead 
 of giving more steam to the engine, the 
 strength of the field is increased. The 
 point E' will then be shifted further 
 down on the vertical line Oy, and the 
 point A will be similarly shifted further 
 towards O, that is to say, the current 
 will be decreased. "When the field has 
 been so far strengthened as to make the 
 induced electromotive force equal to the 
 terminal electromotive force, the point 
 E' will occupy the position E", and the 
 terminal electromotive force the position 
 OE" f , whilst no current will be passing. 
 If now more steam is given, the machine 
 will begin to work as a generator, and 
 the current will reappear and attain its 
 former value, when the power given off 
 by the engine corresponds to the posi- 
 tion E"' of the induced electromotive 
 force. The terminal electromotive force 
 
49 
 
 will then occupy the position E/". 
 Under these conditions the machine will 
 work as a generator, and much more 
 advantageously than formerly, though 
 not yet with the best plant efficiency. 
 To obtain the best result the field must 
 further be streDgthened, until the in- 
 duced electromotive force occupies the 
 position E, and the terminal electromo- 
 tive force the position E f , when the maxi- 
 mum output is reached, and the machine 
 is in exactly the same condition as the 
 other machines. The safe limits, within 
 which the strength of the field may be 
 varied, are given by the length of the 
 lines OE and OE', and the ratio of the 
 corresponding exciting currents must 
 obviously be still greater than that be- 
 tween these two lines. It is instructive 
 to observe what part armature resistance 
 and self-induction take in determining 
 this ratio. As regards the armature re- 
 sistance, the lower this can be made the 
 better ; in modern machines the resist- 
 ance is generally so small that the loss of 
 pressure is limited to from 2 to 5 per 
 
50 
 
 cent, of the terminal electromotive force. 
 The length of the line OK in the diagram 
 will therefore be, at most, one -twentieth 
 part of OE. When the self-induction is 
 negligible, the point A will almost con- 
 cide with O, and E' will almost coincide 
 with E", which shows that the safe mar- 
 gin between minimum and maximum 
 field-strength, and, consequently, also 
 between the smallest and largest per- 
 missible exciting current, is extremely 
 narrow. Machines of this type can only 
 be run in parallel if the strength of their 
 fields is adjusted with almost mathe- 
 matical precision, and as this would 
 require more skill and attention than is 
 available with the ordinary staff of a cen- 
 tral station, such machines are practically 
 unfit for parallel working. To make 
 them fit for this method of working 
 either the armature resistance or the self- 
 induction must be increased. An in- 
 crease of resistance, in order to be 
 effective, would have to be so consider- 
 able (as is easily seen from the diagram), 
 as to seriously prejudice the electrical 
 
51 
 
 efficiency of the machine, and this expe- 
 dient may, therefore, be dismissed as 
 impracticable. The other plan of in- 
 creasing the self-induction is not open to 
 the same objection. It has the effect of 
 lowering the plant efficiency, but its 
 influence upon the electrical efficiency is 
 only indirect, and so small that it may be 
 neglected. From the foregoing, it will 
 be readily seen that the only and suffi- 
 cient condition, for successful parallel 
 working, is a sensible amount of self- 
 induction in the armature circuit. If the 
 armature itself does not possess this 
 quality in a sufficient degree, a choking 
 coil of suitable self-induction must be 
 inserted into the circuit of each machine. 
 The results here arrived at, by a mere 
 theoretical investigation, are entirely 
 borne out in practice. It is well known 
 that alternators having no iron in their 
 armatures cannot be run in parallel, 
 except by the adoption of some such 
 expedient as choking coils ; also that 
 parallel running is feasible with those 
 alternators which have iron- cored arma- 
 
52 
 
 tures, and then with different degrees of 
 security. To make this point clear, 
 compare two machines of different types 
 such, for instance, as the Westinghouse 
 or the Lowrie-Parker on the one hand, 
 and the Kennedy or the author's machine 
 on the other hand. In the two first- 
 named machines the coils are simply laid 
 on to the surface of the armature-core 
 on one side, whilst in the other two 
 machines they completely surround the 
 <?ore. The coefficient of self-induction 
 may, in the latter case, be roughly taken 
 as four times that of the former case, 
 other conditions being approximately the 
 same ; and it may therefore be expected 
 that parallel working will be safer and 
 more certain when the armature con- 
 ductor surrounds the core, than when it 
 lies only on one side of the core, though 
 in this case it will still be possible. The 
 experience gained in America with West- 
 inghouse machines is in this respect very 
 instructive. It has been found that the 
 experiment of parallel coupling always 
 succeeds when the machines are loaded 
 
53 
 
 to about half their maximum output or 
 over, but that machines working under a 
 less load cannot, with certainty, be so 
 coupled. A glance at the diagram, Fig. 
 10, shows the reason for this. The 
 margin of field strength is represented 
 by the difference in the length of the 
 lines OE' and OE. This difference de- 
 pends, to a great extent, upon the length 
 of the line OA, which is proportional to 
 the current. If the current through the 
 machines already at work is small, it will 
 be further reduced as soon as the new 
 machine is switched on, and under these 
 conditions the margin of field strength 
 may become so small, that a slight 
 reduction of the exciting current of one 
 machine may cause this to be over- 
 powered by the other machines, and 
 hence the experiment may fail. If, how- 
 ever, the current through all the machines 
 is sufficiently great, then the margin will 
 also be sufficient to cover such irregu- 
 larities in the exciting currents, and in 
 the torque of the engines, as may be 
 expected to exist when the plant at a 
 
54 
 
 station is in the charge of ordinary 
 attendants, who are not skilled electri- 
 cians or engineers, and hence the experi- 
 ment will succeed. 
 
 It has now been shown how, by the 
 aid of certain diagrams, the behavior of 
 any type of alternator, as regards its 
 ability to run in parallel, can be ascer- 
 tained ; but, as the use of these diagrams 
 involves a certain amount of study, it 
 would be desirable if they could be 
 superseded at least for ordinary prac- 
 tical work, where an approximate solu- 
 tion of such problems is quite sufficient 
 by some more simple method. Such 
 a method is afforded by the use of the 
 characteristics of alternators. If a ma- 
 chine of that class be run at constant 
 speed on open circuit, and with different 
 strengths of exciting current, different 
 terminal electromotive forces are ob- 
 tained. A characteristic can then be 
 plotted, the abscissas of which represent 
 exciting power, whilst the ordinates 
 represent terminal electromotive force, 
 which in this case is the same as the 
 
55 
 
 induced electromotive force. Having' 
 obtained this curve, which the author 
 suggests should be called the static 
 characteristic, the dynamic characteristic 
 of the machine is determined for a cer- 
 tain current through the armature, by so 
 adjusting the resistance in the external 
 circuit, that for all exciting powers the 
 current remains the same. It is import- 
 ant that the external circuit should not 
 contain any appreciable amount of self- 
 induction a condition easily fulfilled if, 
 during the experiment, the machine be 
 set to work a bank of transformers which 
 are feeding glow-lamps. The curve so 
 obtained must obviously lie wholly under 
 the first curve, and the difference be- 
 tween the ordinates of the two curves 
 depends upon the choice of current for 
 which the second test has been made, 
 and upon armature resistance and self- 
 induction. Besides these two curves 
 there is a third,, namely, that obtained 
 when the machine is driven as a motor 
 running light. To find this curve it 
 would be necessary to have at disposal 
 
56 
 
 at least two similar machines, using the 
 one as generator and the other as motor, 
 and during the test the field strength of 
 both machines would have to be regu- 
 lated, so as to keep the current at its 
 predetermined value. Thus, a " motor 
 characteristic " would be obtained, lying 
 wholly above the %t static characteristic." 
 As the last experiment is, however, only 
 possible if two machines can be tested at 
 the same time, and is also somewhat 
 delicate, another method of investigation 
 is desirable. This can be done by using 
 diagram Fig. 10, to deduct the motor 
 characteristic from the two other curves, 
 the experimental determination of which 
 presents no difficulties whatever, and 
 would, indeed, be made as a matter of 
 course in the ordinary test of the ma- 
 chine before it is sent out from the 
 manufacturer. The static characteristic 
 gives the strength of field for every ex- 
 citing power. From the measured re- 
 sistance of the armature it can easily be 
 found how much of the effective electro- 
 motive force has been absorbed in over- 
 
57 
 
 coming this resistance, whilst the dynamic 
 characteristic shows the terminal electro- 
 motive force. Thus all the necessary 
 elements exist for determining, by the 
 aid of a diagram similar to Fig. 10, the 
 length of the line O A and its angular 
 position. Having found the point A, the 
 position of the point E' is determined, 
 that is to say, the strength of field O E', 
 which in a motor will allow the prede- 
 termined current to pass. The corre- 
 sponding exciting power is found from 
 the static characteristic, and this gives 
 one point of the motor characteristic. 
 In the same manner other points corre- 
 sponding to different terminal electro- 
 motive forces can be ascertained. The 
 construction is a mere geometrical opera- 
 tion, and need, therefore, not be 
 explained in detail. Fig. 11 shows the 
 relative position of these curves for one 
 of the author's machines, constructed for 
 an output of 60 kilowatts. For the 
 figures from which the diagram has been 
 constructed the author is indebted to Mr. 
 C. E. L. Brown, who has kindly con- 
 
58 
 
 sented to carry out an in dependent series 
 of experiments, at the works of the 
 Maschinenfabrik Oerlikon, in Switzer- 
 land, the electrical department of which 
 is under his charge. The experiments 
 were made as follows : The alternator 
 was driven at its normal speed of 600 
 revolutions per minute by a steam engine, 
 and its terminals were connected with a 
 
 Fig. II 
 
 transformer wound to reduce the press- 
 ure in the ratio of 20 to 1. The press- 
 ure on the secondary terminals of the 
 transformer was measured by one of Sir 
 William Thomson's balances, specially 
 made for alternate-current work. The 
 
59 
 
 alternator was connected with a bank of 
 transformers, and the secondaries of the 
 latter with lamps and other resistances. 
 A Siemens dynamometer was inserted 
 into the armature circuit, and the second, 
 ary currents from the transformers were 
 occasionally measured by a Sir William 
 Thomson's ampere balance. The primary 
 and secondary currents were found to be 
 very nearly in the ratio of 1 to 20. The 
 exciting current was adjusted by a rheo- 
 stat, and measured by a spring -solenoid 
 instrument. The first run was made with- 
 out external current in the armature cir- 
 cuit, except the small amount necessary to 
 record the pressure by the stepdown 
 transformer. In this manner, data were 
 obtained for plotting the static charac- 
 teristic OS. In the diagram the excit- 
 ing power is plotted on the horizontal 
 and the pressure on the vertical. The 
 exciting power was found by multiplying 
 the exciting current by 280, there being 
 one hundred and forty turns of wire on 
 each magnet-coil. The resistance of the 
 armature when warm is 1.74, and that of 
 
60 
 
 the field 1.73 ohm. The bank of trans- 
 formers, with its load of resistances, was- 
 next switched on, and the latter so reg- 
 ulated as to give an armature current of 
 10 amperes at different degrees of excita- 
 tion. The dynamic characteristic marked 
 " 10 amperes " in the diagram was thus 
 found. Similar experiments were made 
 to obtain the characteristics at 20 and 30 
 amperes, which are also shown. Either 
 of the three dynamic characteristics can 
 now be used to determine, by the aid of 
 diagram Fig. 7, the coefficient of self- 
 induction of the armature ; and, if it 
 were not for certain disturbing causes r 
 the three coefficients thus obtained 
 should be identical. In reality, this is, 
 however, not the case, chiefly because 
 the impressed electromotive force is not 
 a true sine function of the time. There 
 is also another disturbing element,, 
 namely, the demagnetizing influence of 
 the armature current upon the field. 
 When the machine is working as a gene- 
 rator, the poles developed in the arma- 
 ture by the current itself are of the same 
 
61 
 
 sign as the field-poles which they ap- 
 proach, and the armature has therefore a 
 tendency to push back some of the lines 
 emanating from the field-poles. To over- 
 come this tendency the exciting current 
 must be increased, the effect being 
 similar to, though not identical with, that 
 produced by self-induction. When the 
 machine is working as a motor, the effect 
 of the armature current is generally, but 
 not always, to strengthen the field-poles. 
 The author has selected the dynamic 
 characteristic at 30 amperes, for the de- 
 termination of the coefficient of self- 
 induction ; because the great distance 
 between this curve and the static charac- 
 teristic made it probable that the result 
 would be more accurate than with one of 
 the nearer curves. The coefficient of 
 self-induction was found to be L = 0.955 
 in C. G. S. measure, and 0.0955 in prac- 
 tical units. Taking this as a basis, it is 
 now an easy matter to find, by means of 
 the graphic construction shown in Fig. 
 10, the motor characteristic for any given 
 current. A current of 10 amperes was 
 
62 
 
 chosen, as sufficiently small to avoid a 
 sensible disturbance of the working ma- 
 chines when a new machine is switched 
 on, and this gives the dotted curve, 
 marked a 10 amperes," in Fig. 11. The 
 interpretation of this diagram is obvious. 
 The normal terminal pressure of this 
 machine is 2,000 volts, and by drawing 
 the corresponding horizontal line, its 
 points of intersection with the curves 
 m s g fix the exciting power in each 
 case. In order that the machine may 
 work as a generator, giving its maximum 
 current of 30 amperes, the exciting power 
 must be O G. If the machine is merely 
 to run without producing or absorbing 
 current, it must be excited to an amount 
 represented by O S, and if running idle 
 as a motor, and allowing not more than 
 a 10 ampere current to pass, the excita- 
 tion must be O M. The ratio between 
 the length of the lines O M and O G 
 gives, therefore, in a rough-and-ready 
 way, an indication of the degree of safety 
 when working in parallel. 
 
 The formulas and diagrams given above 
 
63 
 
 will be found sufficient for the purpose of 
 designing alternators, or for the deter- 
 mination of the behavior of any given 
 machine of this class, except in one par- 
 ticular, viz., its greater or lesser liability 
 to heating. In any dynamo machine 
 there are two causes for beating, in 
 addition to purely mechanical friction. 
 These are the resistance of the conduct- 
 ors, and the change of magnetization of 
 the iron parts of the machine. As re- 
 gards the first cause, the result of which 
 may be comprised under the term "cop- 
 per heat," it is so well understood that 
 it need not in this place be specially con- 
 sidered. One circumstance in connec- 
 tion with this subject should, however, 
 be mentioned. The liability to the gen- 
 erations of eddy currents in the armature 
 conductor is much greater with alter- 
 nators than with continuous-current dyna- 
 mos, and in alternators it varies with 
 the construction of the armature to a 
 greater degree. When the armature 
 contains no iron, the heating from 
 eddy currents is much greater than 
 
64 
 
 when such a core is used, and the arm- 
 ature conductor must therefore be made 
 in the form of a very narrow strip ; 
 so narrow, in some cases, that two or 
 more conductors must be placed in par- 
 allel to obtain the required cross-section- 
 al area. The author is unable to suggest 
 an entirely satisfactory explanation for 
 the effect of the iron core in reducing 
 eddy currents in the copper, and merely 
 mentions the fact as deserving attention, 
 In direct current machines he has ob- 
 served a somewhat similar phenomenon. 
 When the pole-pieces are provided with 
 extensions, the heating of the armature 
 bars is much less than when these exten- 
 sions are absent, and consequently the 
 trouble of heating can to a certain de- 
 gree be overcome by the simple expe- 
 dient of fitting so-called u horns " to the 
 pole-pieces. Another cause of the gen- 
 eration of heat, in machines of all types, 
 is the change of magnetization in the 
 iron parts. This can be minimized by 
 lamination, but never entirely avoided. 
 In direct-current machines the heating 
 
65 
 
 from this cause is, however, generally so 
 small that no account is taken of it, ex- 
 cept when the induction in the armature 
 core is pushed beyond 18,000 or 20,000. 
 In alternators a much lower induction 
 already produces sensible heating, and it 
 is for this reason that the maximum in- 
 duction must be fixed at a very low figure. 
 With a perfectly laminated core, the 
 heating is attributed to an effect which 
 Professor J. A. Ewing calls 4 ' hysteresis,'' 
 and which may be described as molecular 
 friction. This scientist has experiment- 
 ally determined the energy required to 
 carry certain samples of iron through a 
 complete cycle of magnetization of vary- 
 ing degree, and the results have been pub- 
 lished in the Philosophical Transactions 
 of the Royal Society.* They are given 
 in ergs per cubic centimeter ; but as 
 these units are rather inconvenient for 
 application in the workshop or drawing 
 office, the author has, in the following 
 table, which refers to annealed wrought- 
 iron, translated them into watts per ton. 
 
 * Phil. Trans. Royal Society, 1885, p. 523. 
 
66 
 
 Induction. 
 
 Watts per ton 
 n = 100. 
 
 HP. wasted 
 in Heat. 
 
 2,000 
 
 650 
 
 0.87 
 
 3,000 
 
 1,100 
 
 1.48 
 
 4,000 
 
 1.650 
 
 2.21 
 
 5,000 
 
 2,250 
 
 3.02 
 
 6,000 
 
 2,900 
 
 3.89 
 
 7,000 
 
 3,750 
 
 5.03 
 
 8.000 
 
 4,450 
 
 597 
 
 9,000 
 
 5,550 
 
 7.43 
 
 10,000 
 
 6,650 
 
 8.90 
 
 A third column has been added, show- 
 ing the HP. transformed into heat, and 
 therefore wasted per ton of armature- 
 core, if the frequency is 100 complete 
 periods per second. 
 
 In the paper referred to above, Pro- 
 fessor Ewing says that mechanical vibra- 
 tions tend to decrease the amount of 
 energy dissipated per cycle ; and that, as 
 the armatures of dynamo machines are, 
 whilst working, in a continual state of 
 vibration, it will be very much less than 
 indicated by the figures he found when 
 experimenting on pieces of iron which 
 were at rest. The author, however, 
 doubts whether this view can safely be 
 accepted. The vibrations of an armature 
 
67 
 
 are not of the sharp character compara- 
 ble to a series of blows, which would 
 tend to shake the magnetism out of the 
 core, and practical experience shows that 
 with high inductions there is a very con- 
 siderable amount of heating. In an ex- 
 perimental machine which the author 
 constructed, the induction through the 
 armature core was about 16,000, and with 
 a frequency of 80 this core became 
 heated far beyond the point which might 
 be considered safe in practical work. In 
 subsequent machines the induction was 
 therefore reduced by degrees, until a safe 
 limit was reached at about 7,000 lines 
 per square centimeter. 
 
 Professor Ewing found that the energy 
 dissipated by hysteresis increased with 
 the speed at which the cyclic change of 
 induction was performed. This circum- 
 stance has not been taken into account 
 in the preparation of the table, as no 
 exact data are at present available, by 
 which the increase of heat due to what 
 Professor Ewing calls "viscous hys 
 teresis" can be determined. There is, 
 
68 
 
 however, good reason to believe that the 
 speed, at which the cyclic change of 
 magnetism is performed, has a consider- 
 able influence on the amount of energy- 
 dissipated in heat per cycle. It is prob- 
 ably due to the disadvantage of too great 
 a speed that most makers of alternators 
 find it advisable to work at a moderate 
 frequency. Another reason telling 
 against high frequency is that either the 
 speed of the machine, or the number 
 of field-poles, or both, must be incon- 
 veniently great. A very high speed is 
 mechanically objectionable, and the adop- 
 tion of a large number of field-poles 
 has the disadvantage of reducing the 
 pitch, and therefore of increasing the 
 magnetic leakage, whilst at the same 
 time lowering the coefficient k, all of 
 which tend to an increase in the weight, 
 bulk, and cost of the machine. On the 
 other hand, too low a frequency would 
 also tend to increase the cost of the 
 machine, and there must therefore be, for 
 each type of alternator, a particular 
 frequency at which the best results are 
 
obtained. Existing knowledge of the 
 subject is not sufficient for the determi- 
 nation of this frequency by a mere 
 theoretical investigation, but the experi- 
 ence obtained by several makers of these 
 machines, during the last few years, tends 
 to show that the best frequency for 
 nearly all types is below 100. Something 
 more will be said on this point later on. 
 
 TRANSFORMERS. 
 
 The theory of transformers is so sim- 
 ple, and so generally understood, that it 
 would be waste of time to enter into it at 
 length in this place. Modern transform- 
 ers are provided with a core of laminated 
 iron, which is symmetrically placed in 
 respect to the two windings, so that the 
 same induction passes through both the 
 primary and the secondary coils. Under 
 these conditions, the electromotive force 
 induced in the two circuits is simply pro- 
 portional to the number of convolutions 
 they respectively contain, and the ter- 
 minal electromotive forces are found 
 by adding in the case of the primary, 
 
70 
 
 and subtracting * in the case of the 
 secondary, the loss of pressure due to 
 the resistance of the coils. In a properly 
 designed transformer, this loss of pressure 
 should not exceed about 1 per cent, of 
 the terminal pressure, so that the maxi- 
 mum difference in the pressure main- 
 tained in the secondary circuit between 
 the lamps, when the apparatus is used 
 for parallel distribution, does not exceed 
 2 per cent, between full load and no load, 
 provided the pressure in the primary 
 circuit is kept constant. Such a small 
 variation in the pressure is well within 
 the limits attainable in direct lighting, 
 when a compound- wound dynamo is driven 
 from an engine with a fairly good gov- 
 ernor, and the corresponding variation in 
 the illuminating power of the lamps is 
 sufficiently small to be tolerated. A 
 greater variation, especially if it occur 
 suddenly, would, however, not be per- 
 missible ; and the question, therefore, 
 arises as to what rules should be followed 
 in designing transformers so that they 
 may as nearly as possible be self-regulat- 
 
71 
 
 ing. The obvious answer is: make the 
 resistance of the circuits as small as 
 possible, and this may be accomplished 
 in three ways : 1. By using wire of 
 large area. This expedient naturally 
 increases the weight, bulk, and cost of 
 the transformer, and must, on account of 
 commercial reasons, not be pushed too far. 
 2. By employing a core approximating to 
 a circular or quadratic section, so as to 
 decrease the perimeter of the coils- 3. 
 By working with a high induction or 
 with a high frequency. 
 
 The mean electromotive force in volts, 
 induced in the coil of a transformer, can 
 be expressed by the following formula 
 
 e = (q) 2 n n z r lO' 8 ..... (11), 
 
 in which (q) is a coefficient denoting the 
 ratio between the maximum and the 
 mean electromotive force (when the cur- 
 rent-wave may be taken to follow a sine- 
 
 function, this coefficient is _.. V n is 
 
 _.. V 
 '-t/ 
 
 the frequency, z the total induction in 
 absolute units, and r the number of turns 
 
72 
 
 contained in the coil. The larger 3, and 
 the smaller n, may be the length of wire, 
 and therefore its resistance. On purely 
 theoretical grounds it would thus seem 
 that a small core strongly magnetized, 
 and as high a frequency as the alternator 
 can be constructed to give, would pro- 
 duce the best results. In practice, how- 
 ever, this conclusion is wrong, and the 
 heating of the iron core in consequence 
 of hysteresis must be taken into consider- 
 ation. Since a transformer is an 
 apparatus at rest, and is frequently 
 placed in a position where the cooling 
 effect of the air is small, the ratio 
 between heat generated and available 
 cooling surface must be very different 
 from that permissible in a dynamo, and 
 it is this ratio which is the true limit to 
 the output of a transformer. The arma- 
 ture of a dynamo is not subjected to 
 excessive heating if a cooling surface of 
 about 1 square inch per watt be provided. 
 In a transformer, the amount of cooling 
 surface should be at least three times, 
 and if possible, four or five times as great, 
 
73 
 
 especially as it is somewhat difficult to 
 determine, in a body of such complicated 
 shape, what is and what is not effective 
 cooling surface. From the table, giving 
 the loss by hysteresis, it is easy to calcu- 
 late for any induction, frequency, and 
 weight of the active iron in a transformer, 
 what the iron heat will be ; and, adding to 
 this the copper heat, the total cooling 
 surface required to keep the apparatus at 
 a safe temperature can at least approxi- 
 mately be determined, or the limit of 
 output for any given transformer can be 
 approximately found. In this connection, 
 it is instructive to compare two similar 
 transformers, one twice the linear dimen- 
 sions of the other. For the purpose of 
 the comparison, it may be assumed that 
 in the small transformer the iron heat is 
 the same as the copper heat, though it is 
 generally greater. If the large transformer 
 be worked at the same induction per unit 
 area of core, its iron heat will be eight 
 times that of the small transformer, or 
 four times its total heat; but as the 
 cooling surface is only four times as 
 
74 
 
 great, the anomalous result follows, that 
 the output of the large transformer is 
 zero, although its electromotive force 
 would be four times that of the other. 
 The comparison would have been still 
 more unfavorable to the large trans- 
 former, had it been assumed that the 
 iron heat in the small transformer was 
 greater than its copper heat. Therefore 
 in order to be able to work the large 
 transformer at all, the induction must be 
 reduced ; that is to say, the transformer 
 used on a circuit, the electromotive force 
 of which is less than four times that of 
 the small transformer. The resistance 
 of the large coil is half that of the small 
 one, and if the large transformer is 
 worked so as to have equal iron and 
 copper heat, the latter may obviously be 
 four times that of the small transformer; 
 this gives a current 2.82 times the value 
 of that in the small transformer. The 
 weight of iron in the large transformer 
 being eight times that of the small one, 
 the loss by hysteresis per unit of weight 
 must be half. The corresponding indue- 
 
75 
 
 tion can be ascertained from the table. 
 If, for instance, the induction in the 
 small transformer is 8,000, that in the 
 large one may be taken as 5,000, and the 
 
 electromotive force will be ' AA X 4 = 
 
 o,UUU 
 
 2.5 times that of the small transformer, the 
 output being increased in the ratio of 1 
 to 2.82 x 2.5 = 7 about. It should 
 also be observed that the larger appa- 
 ratus will, in so far, be more advantageous, 
 as a smaller proportion of space will be 
 occupied by insulation, especially if both 
 are required for the same circuit. The 
 ratio of output may thus be increased 
 from seven to eight, or ten, or even more, 
 according to the size of transformer 
 taken as the basis of the comparison. 
 For larger sizes, the advantage resulting 
 from an increase of linear dimensions 
 will be less apparent, and the output may 
 roughly be taken as proportional to the 
 weight. 
 
 The question of frequency, alluded to 
 in connection with alternators, is also of 
 considerable practical importance with 
 
76 
 
 regard to the transformers, and it is 
 much to be desired that electrical engi- 
 neers should adopt a common standard. 
 Comparing two equal transformers, one 
 working with a frequency of, say, 65, and 
 the other with one of, say, 130, it is 
 evident that with the same pressure the 
 induction in the latter need only be half 
 that in the former ; and if the effect of 
 static hysteresis only had to be con- 
 sidered, the heat generated per cycle 
 would be something less than half ; and 
 the total heat generated in a given time 
 would, with the high frequency, be some- 
 what less than with the low frequency. 
 But as there is every reason to believe in 
 the existence of viscous hysteresis, by 
 virtue of which the heat generated per 
 cycle increases with the speed at which 
 the cycle is performed, there must be a 
 limit to the frequency beyond which a 
 further increase becomes disadvantage- 
 ous. With the present imperfect knowl- 
 edge of this subject, it is not possible 
 to determine this limit on theoretical 
 grounds ; but it is justifiable to look to 
 
77 
 
 common practice for an indication of 
 what the limit probably is. The present 
 types of alternators and transformers, it 
 must be remembered, are not mere 
 experimental machines, but the survival 
 of the fittest, and it would be futile to 
 ignore the lessons taught by practical 
 experience extending over several years. 
 The following is a list giving the average 
 frequencies adopted by several designers : 
 Ferranti, 67 ; Lowrie-Parker, 88 ; Mor- 
 dey, 100 ; Zipernowsky, 42 ; Kennedy, 
 60 ; Kapp, 80 ; Westinghouse, 133. 
 
 Taking the average of European prac- 
 tice, the frequency is 73, or 8,750 reversals 
 per minute ; whilst the American practice 
 is 16,000 reversals per minute. This 
 large difference may, to a certain extent, 
 be accounted for by the general pre- 
 ference of European engineers for 
 machines running at moderate speeds: 
 but even when making due allowance for 
 this circumstance, it can hardly be as- 
 sumed that, had actual practice shown 
 the value of a very high frequency, elec- 
 trical engineers would not have found 
 
78 
 
 means to obtain it. As a matter of fact, 
 most makers of alternators have started 
 with a very much higher frequency than 
 they have finally adopted. It should 
 also be remembered that the higher the 
 frequency, the greater is the loss of con- 
 ductivity in the mains, owing to the 
 unequal distribution of current through- 
 out their cross-section. 
 
 MOTORS. 
 
 Machines, for the conversion of the 
 electrical energy of alternating currents 
 into mechanical energy, have not yet 
 been brought to such a state of perfec- 
 tion that the problem may be considered 
 as solved and, in fact, very little is at 
 present known as to the details of such 
 machines and the results achieved. Mo- 
 tive power may be obtained from an 
 alternating current in either of three dis- 
 tinct methods : 
 
 1. By the employment of an alternator 
 with separately excited field, and some 
 means to first run it up to speed before 
 switching the current on. Various de- 
 
79 
 
 vices have been patented for this method 
 of working ; but the author is not aware 
 that any one of them has proved success- 
 ful. 
 
 2. By the employment of a direct-cur- 
 rent motor with laminated field-magnets. 
 Such motors have been made and tested 
 by various engineers ; and about two 
 years ago the author also experimented 
 with such a motor, but the results were 
 discouraging. The work obtainable from 
 this motor, with a given current and 
 electromotive force, was only about one- 
 fifth part that which might be obtained 
 from the same motor if driven by a direct- 
 current of the same measured strength, 
 and flowing under the same pressure. 
 This particular type of motor was, there- 
 fore, commercially, quite unfit for use 
 with an alternating-current supply. After- 
 wards, when investigating the reason of 
 this failure, the author found that the 
 electromotive force of self-induction was 
 far too great, in comparison to the counter 
 electromotive force of the armature, pro- 
 ducing thus a very large lag, and conse- 
 
80 
 
 quently reducing the plant-efficiency too 
 much. The investigation also showed 
 that, under the most favorable conditions, 
 the plant- efficiency of such a motor could 
 not be more than 70 per cent. This 
 figure would be obtained if the electro- 
 motive force of self-induction of the whole 
 machine were by some means reduced, so 
 much as to be only equal to the counter 
 electromotive force developed in the 
 armature, a condition which is extremely 
 difficult, if not impossible, to fulfill in 
 practice. For this reason, it would seem 
 that a motor of this type must always be 
 very much larger, more costly, and heavier 
 than a direct-current motor of the same 
 power. It has the advantage of being 
 self -starting, and not requiring any ac- 
 cessory apparatus for its excitation or 
 regulation. On the other hand, there is 
 the liability of the armature burning up, 
 if the motor should fail to overcome its 
 load almost at once. When the armature 
 is at rest, and an alternating current 
 passes through the machine, those arma- 
 ture coils which at the time happen to be 
 
81 
 
 short- circuited by the brushes, are in the 
 same condition as the secondary of a 
 transformer short-circuited upon itself. 
 They will therefore be liable to burn up 
 if the armature cannot start at once. 
 
 3. The third method of producing 
 motive power from alternating currents 
 has hitherto received most attention, as 
 being the most promising. It is due to 
 a discovery made about a year and a half 
 ago by Professor Galileo Ferraris, of 
 Turin. This scientist found that a cop- 
 per cylinder, suspended between two 
 coils, was set into rotation if alternating 
 currents of the same period, but of differ- 
 ent phase, were sent through these coils 
 which were placed at right angles to each 
 other. The explanation, given by Pro- 
 fessor Ferraris, was that the resultant 
 field of the two currents revolved round 
 the common center-line of the coils, and, 
 by means of eddy currents created in the 
 copper cylinder, dragged the latter after 
 it. This principle has been practically 
 developed by Mr. Nicola Tesla and others, 
 and motors have been actually built in 
 
82 
 
 which a revolving field causes, by a kind 
 of electro-magnetic drag, an armature to 
 revolve and give off mechanical work. 
 This action may be represented by the 
 diagram Fig. 12, where A A and B B are 
 two coils placed at right angles, and in 
 
 the common center, O, of which is placed 
 an armature consisting of a series of coils 
 C, each forming a loop closed on itself. 
 Let there be a difference of phase of 90 
 between the currents in the two field- 
 
83 
 
 coils, and let F a and O F b represent 
 the fields produced at a given moment. ^ 
 The resultant field is O F, and this 
 revolves round O with a speed corre- 
 sponding to the frequency. In the 
 diagram a cross placed in the circle, 
 representing a wire or coil, signifies an 
 ascending, and a dot a descending cur- 
 rent. When the rotation of the resultant 
 field takes place in the direction of the 
 arrow, the current in B must be approach- 
 ing zero, and that in A its maximum 
 value. 
 
 Suppose, now, the armature is held at 
 rest. The field in sweeping through the 
 conductors C induces in them currents in 
 the direction indicated, and the latter 
 therefore exert a torque. Imagine the 
 field stationary, and the armature re- 
 volved by a belt. In this case, the cur- 
 rent created in the short-circuited coils 
 will produce a torque tending to resist 
 rotation. The lower the resistance of 
 these coils and the greater the speed of 
 rotation, the more power will be required 
 to rotate the armature. It is thus evident 
 
84 
 
 that with a revolving field the greatest 
 force will be exerted upon the armature 
 < whilst this is at rest, and that the power 
 will decrease as the armature begins to 
 revolve and gather speed, just as in an 
 ordinary direct-current motor. When 
 the speed of the armature coincides with 
 that of the resultant field, no torque is 
 produced, whilst at starting, the torque 
 is a maximum. According to the load 
 put upon the motor, the armature will 
 therefore automatically assume that speed 
 which corresponds to the load, and in 
 this respect the condition of working is 
 comparable to that of a continuous- cur- 
 rent motor. The current through the 
 armature coils C produces a field, O/> 
 which is also revolving, and in sweep- 
 ing past the wires in the fixed coils A A, 
 B B, creates therein an electromotive 
 force which is always opposing the flow of 
 current, as will easily be seen from the 
 diagram. It will also be observed that 
 the wave of counter electromotive force 
 in the fixed coils, thus produced, coin- 
 cides with the current wave therein, and 
 
85 
 
 the current must therefore do work in 
 flowing against this counter electromotive 
 force. It is this work thus absorbed by 
 the fixed coils which partly reappears as 
 mechanical work of rotation given out by 
 the armature, the rest being dissipated 
 in heat as in the case of an ordinary 
 motor. The action here explained is in 
 reality not quite so simple, because self- 
 induction and the reaction of the arma- 
 ture field have to be taken into account ; 
 but, in the absence of all experimental 
 data, it is not worth while to attempt a 
 closer investigation. The author has 
 thought it expedient to explain the funda- 
 mental principle of this type of motor, as 
 the investigation might prove useful to 
 those who are about to experiment with 
 similar machines. He regrets that he is 
 unable to place before the institution 
 details of the Tesla motor, as constructed 
 by the Westinghouse company ; but as 
 these details are only partially worked 
 out, and secured by patents, this com- 
 pany naturally prefer to withhold them 
 for the present. 
 
86 
 
 DESCRIPTION or MODERN ALTERNATORS. 
 
 By the courtesy of the designers and 
 makers of alternators, the author is able 
 to submit particulars of the various 
 machines most in use, and he takes this 
 opportunity of tendering his thanks to 
 those who have supplied him with infor- 
 mation. Modern alternators may be 
 divided into two classes, disk machines 
 and drum machines. In the former type 
 the armature-coils are arranged in disk 
 form, and the magnet- poles are presented 
 to them from two. sides, whilst in the 
 latter type the coils are arranged to form 
 a cylindrical surface, and the magnet- 
 poles are presented from one side only. 
 
 DISK MACHINES. 
 
 THE FERRANTI ALTERNATOR. 
 
 The magnet-cores are of trapezoidal 
 section (Figs 13 and 14), and supported 
 in cast-iron yoke- rings. The armature- 
 conductor is a thin corrugated copper 
 strip, wound with a strip of vulcanized 
 fiber of equal width upon a brass core. 
 
87 
 
 To avoid eddy currents in the latter, it is 
 subdivided into corrugated narrow strips, 
 separated from each other by asbestos. 
 
 The shape of the coils is shown in Fig. 
 13. At the inner and narrower end, the 
 core is provided with a projection by 
 which it is secured to a bobbin-carrier. 
 
88 
 
 This in turn is supported by, but insu- 
 lated from, the central armature- ring. 
 There are two bobbins to each carrier, 
 which forms the metallic connection be- 
 tween their inner ends. There is thus a 
 double insulation for the coil, first that 
 of the conductor against the core and 
 bobbin-carrier ; and secondly, that of the 
 latter against the body of the machine. 
 The winding of the coils is such, that the 
 current passes through the armature in 
 two parallel circuits, by which arrange- 
 ment the terminals of the armature are 
 brought to opposite points of a diameter ; 
 and the connection between adjacent 
 coils is made automatically, by the act of 
 inserting and bolting up the coil so that 
 it is impossible for a workman, when 
 replacing a coil, to make a wrong con- 
 nection. The mean width of poles is 
 approximately equal to half the pitch, and 
 the width of coils is also equal to half the 
 pitch, so that, according to the table on 
 p. 22, the coefficient is 2.300. The follow- 
 ing are the principal data for a 150-HP. 
 machine: 500 revolutions, 46.5 amperes, 
 
89 
 
 2,400 volts, 20 poles. Armature-conductor 
 490 by 11.8 mils. Current density 4,000 
 amperes per square inch. Eesistance of 
 armature 1.2 ohm. Total number of con- 
 ductors on armature 3,440, divided into 
 two parallels of 1,720 each. Field-coils of 
 320 turns, exciting current 13.5 amperes. 
 Exciting power 17,200 ampere turns. 
 The interpolar space is 0.75 inch, and the 
 area of pole-pieces 14.5 square inches. 
 From these data it is possible to calculate 
 the total flow of the lines. It is in 
 English measure z 130, and the elec- 
 tromotive force of an equivalent con- 
 tinuous current machine would be 1,150 
 volts. The ratio between this figure and 
 the electromotive force of the alternator is 
 2.15 which agrees fairly well with the theo- 
 retical value of 2.30 mentioned above. 
 The resistance of the field is 12.8 ohms, 
 and the exciting energy 2,340 watts, or 
 only 2.3 per cent, of the total output. 
 
 THE MORDEY ALTEKNATOR. 
 
 This is also a disk machine; but with 
 revolving magnets and Jixed armature 
 
90 
 
 (Figs. 15 and 16). The cores of the arma- 
 ture-coils are of porcelain, and the con- 
 ductor consists, as in the Ferranti ma- 
 chine, of a thin copper strip insulated in 
 a similar manner. The distinguishing 
 feature of this machine is the field. It 
 is of the iron-clad description, and the 
 poles on one side are all of the same 
 sign, whilst those on the other side of the 
 
 Fig. 15 Fig. 16 
 
 armature are of the opposite sign. There 
 is only one coil of exciting wire, C, placed 
 as shown in the diagram. The field may 
 be diagrammatically represented by Fig. 
 . Each group of conductors is fairly 
 concentrated, and the coefficient k for 
 ihis machine should, therefore, lie be- 
 iween the values given for cases 1 and 3 
 on p. 22. If the conductors were con 
 
91 
 
 centrated into lines, k would be equal to 
 2 000 ; if they occupied half the available 
 space, it would be 1.635. As will be 
 seen from the diagram, they occupy 
 rather less than half the available space, 
 and k should, therefore, be greater than 
 1.635, though smaller than 2. The 
 author has not been able to obtain the 
 constructive data of this machine, and 
 can, therefore, not give the exact value 
 of the coefficient k for it. 
 
 THE KENNEDY ALTERNATOE. 
 This, too, is a disk machine of the iron- 
 clad type, the field being stationary and 
 the armature revolving (Figs. 17, 18, and 
 19). The latter has an iron core, A, 
 round which the armature-conductor is 
 wound, forming radial coils with parallel 
 sides. The field-magnet consists of an 
 external yoke-ring y bounded by flat 
 annular disks, which on their inner faces 
 are provided with polar projections N S, 
 occupying relatively to each other inter- 
 mediate angular positions. There is only 
 one coil of exciting wire, which is, how- 
 
92 
 
 ever, for greater convenience of manufact- 
 ure, arranged in two halves, C C. Fig. 
 19 shows diagrammatically the relative 
 position of armature-coils and field-poles 
 straightened out. It will be seen that 
 the lines of force, which are shown dotted, 
 have an oblique course, and that only 
 one-half of the wire in the armature- coil 
 is at any time in the position of greatest 
 activity. The width of the field, when 
 making due allowance for the fringe of 
 lines which surrounds the actual contour 
 of the pole-piece, is somewhat greater 
 than the pitch ; whilst the proportion of 
 space on the armature, occupied by the 
 winding, varies with the radial distance 
 at which it is measured. On the inside 
 of the armature nearly the whole space 
 is so filled, whilst on the outside a little 
 less than half the space is occupied by 
 the coils. For the outside of the arma- 
 ture the coefficient k will, therefore, be 
 nearly that corresponding to case 3 on 
 p. 22, and for the inside it will be rather 
 less than that of case 2. The actual 
 mean value of the coefficient should 
 
93 
 
 therefore be a little greater than 1.000 ; 
 and this is, indeed, the case. According 
 to a calculation made by Mr. Kennedy, 
 the number of English lines emanating 
 from one pole is, for the machine illus- 
 trated in the diagram, 148 ; the armature 
 contains twelve coils of 18 turns each, or 
 216 turns in all ; the exciting-coils con- 
 Fig. 17 Fig; 18 
 
 tain 570 turns of wire, the exciting-current 
 is 25 amperes, and the exciting-power is, 
 therefore, 14,250 ampere turns. At a 
 speed of 800 revolutions per minute, the 
 terminal pressure is 150 volts, and the 
 current 70 amperes, the output being 
 thus 10.5 kilowatts. No experiments 
 have been made to determine the induced 
 electromotive power, which naturally is- 
 
94 
 
 effected by the self-induction of the 
 armature; but Mr. Kennedy states that 
 the difference in exciting-power for full 
 output and running idle is not great. 
 From analogy, with his own machines, 
 the author estimates that the induced 
 electromotive force of this machine, when 
 giving 150 volts at the terminals, would 
 be about 165 volts. If the armature were 
 coupled to give a continuous current, the 
 electromotive force produced by a field 
 of 148 lines would be 154, so that k in 
 this machine must be about 1.07. This 
 yalue, although rather smaller than 
 theory would indicate, is yet within the 
 limits given above. It may here be 
 mentioned that Mr. Kennedy advocates 
 the use of low-pressure alternators in 
 connection with step-up transformers if 
 a high-pressure current is required. 
 
 THE KAPP ALTERNATOR. 
 
 The author's machine is illustrated in 
 Figs. 20 and 21. The armature-core A 
 consists of thin band-iron coiled upon 
 a cast-iron supporting-ring, and sur- 
 
95 
 
 rounded by radial coils c. The yoke- 
 rings y are of cast-iron and the magnet- 
 cores of wrought or cast-iron with pole- 
 shoes of rectangular shape. This arrange- 
 ment has been adopted with a view to 
 obtain, on the outside of the armature, as 
 large a ratio as possible between the 
 pitch and length of field. For the same 
 purpose, and to reduce leakage, the pole- 
 shoes are tapered on the inside. The 
 importance of securely supporting the 
 armature-wires is well known, and there 
 is hardly any good continuous-current 
 dynamo now made, in which this 
 mechanical detail does not receive careful 
 attention. In alternators, a strong 
 mechanical fastening of the wires on the 
 armature is, however, of even greater 
 importance. The mechanical strain to 
 which the wire is subjected depends, not 
 on the mean current and electromotive 
 force at which the machine is rated, but 
 on their maximum values. Now the 
 maximum value of the periodic current 
 is about 40 per cent, greater than the 
 mean value, and the same is the case with 
 
96 
 
 the electromotive force. It should also 
 be observed that the strain depends upon 
 the induced electromotive force, whereas 
 the output is computed on the basis of 
 the terminal electromotive force, which is 
 somewhat smaller. As the crest of 
 the current- wave occurs in ordinary 
 work only a very short time after that 
 of induced electromotive force, the 
 mechanical strain to which each coil is 
 subjected twice in each period is that 
 due to the maxima, and not to the 
 mean values of current and electromotive 
 force, and is therefore, roughly speaking, 
 at least twice as great as in a continuous- 
 current machine of the same dimensions 
 and output. To provide against these 
 strains, the author inserts between the 
 coils driving-horns of insulated metal or 
 fiber, which pass through the core of the 
 armature, and are secured in their posi- 
 tion by radial bolts. The machine 
 illustrated in Figs. 20 and 21 is one of five 
 120-kilowatt alternators now in course 
 of construction for a central station ; 
 but as it has not yet been tested, it will 
 
 
97 
 
 be more satisfactory to give the construct- 
 ive data and results of a smaller machine 
 which has been at work. For this pur- 
 pose may be selected the machine to 
 which the characteristic curves, Fig. 11, 
 refer, and which has been exhaustively 
 tested by Mr. Brown, as already men- 
 tioned. The supporting-ring of this 
 machine is 31 inches in diameter and 3 
 inches wide. The core is of equal width 
 
 Fig. 21 
 
 and 8 1 inches deep, and is wound with 
 1,120 turns of wire arranged in fourteen 
 coils of 80 turns each. The resistance of 
 the armature when warm is 1.74 ohm. 
 The field-magnet cores are of cast iron, 3 
 inches by 7 inches, with rounded corners, 
 and there are no pole-shoes. Each mag- 
 net-coil contains 140 turns of wire, and 
 the total resistance of the twenty-eight 
 coils is 1.73 ohm. The calculated 
 
98 
 
 strength of field at 6,000 ampere turns is 
 142 English lines, and the observed 
 terminal pressure when running idle is 
 1,560 volts. This gives a coefficient of k 
 = 2.370. When working at the full out- 
 put of 30 amperes and 2,000 volts, the 
 mean electromotive force of self induction 
 is 1,250 volts, and the induced electromo- 
 tive force 2,400 volts. The loss of press- 
 ure through armature-resistance is in that 
 case 52 volts, or about 2.6 per cent, of 
 the terminal electromotive force. The 
 energy lost in exciting the field, which 
 requires 12,000 ampere turns, is 3.2 
 kilowatts, or 5.3 per cent, of the output. 
 This somewhat large percentage is due to 
 the employment of cast iron for the mag- 
 net-cores, and will, in the machine illus- 
 trated, which has cylindrical magnets of 
 wrought-iron, be reduced to about 3.5 
 per cent. The data, obtained by Mr. 
 Brown in testing this machine, suffice for 
 the determination of the coefficient of 
 self-induction by means of a diagram 
 similar to Fig. 7, which also gives the 
 angle of lag. Knowing the latter, it is 
 
99 
 
 possible to find the counter-magnetomo- 
 tive force due to the armature current at 
 the moment when the armature-coils are 
 in mid position between two poles. The 
 maximum current is 30 ^/ 2 = 42.4 am- 
 peres, and the corresponding exciting 
 power is 3,400 ampere turns. This, 
 however, occurs at a time when the coils 
 are well within the polar surfaces of the 
 field-magnets and the poles produced in 
 the armature are between neighboring 
 field-poles. In this position the demag- 
 netizing effect must be naturally small. 
 It will be a maximum when the armature- 
 poles are opposite the field-poles, and at 
 that moment the exciting power of the 
 armature-current is about 1,700 ampere 
 turns. This value must, therefore, be 
 deducted from the exciting power applied 
 to the field, to obtain the true effective 
 exciting power which is causing the flow 
 of lines. Thus the value 10,300 is ob- 
 tained, to which corresponds a flow of 
 215 lines through the armature, giving 
 the induced electromotive force of 2,400 
 volts. 
 
100 
 
 DRUM MACHINES. 
 
 The distinctive feature of the machines 
 comprised under this title is that the 
 armature-coils do not surround the core, 
 but are arranged on one side of it. As 
 examples, may be mentioned the Westing- 
 house, the Lowrie-Parker, and the 
 Zipernowsky machines. 
 
 THE WESTINGHOUSE ALTERNATOR. 
 
 The armature-core of this machine is 
 of cylindrical shape, and consists of thin 
 iron plates, as in an ordinary drum machine 
 for continuous currents. To provide 
 internal cooling surfaces, the plates have 
 large holes cut out of their surface, form- 
 ing, when put together, tunnels, going 
 from end to end, through which the air 
 
101 
 
 may circulate. The armature-coils are 
 of flat-link shape with rounded ends, 
 which are laid on to the surface of the 
 core with the active wires parallel to the 
 spindle. The ends of the coils are 
 brought over the end plates of the core, 
 and bent inwards, as shown at C C C in 
 Fig. 22, and secured in that position. 
 The straight portions of the coils are 
 fastened to the surface of the core by 
 insulated filling-pieces and binding-hoops. 
 The field magnets N S are set radially 
 outside the armature, and their outer ends 
 are connected by a cylindrical yoke y. 
 
 THE LOWRIE-PARKER ALTERNATOR. 
 
 This machine is the inverse of that 
 just described. The armature is outside 
 of the field, and is stationary, whilst the 
 magnets revolve. The armature coils 
 C 0, Fig. 23, are also of flat-link shape, 
 the conductor consisting of a cotton- 
 covered strip, wound upon a core of wood 
 with rounded ends ; they are laid with 
 their long side parallel to the spindle on 
 to the inner surface of the armature-core 
 
102 
 
 A, which is built up of segmental plates 
 to form a cylinder. The following par- 
 ticulars refer to a 100-kilowatt machine, 
 giving 50 amperes at 2,000 volts terminal 
 pressure, the speed being 380 revolutions 
 per minute. The internal diameter of the 
 armature core is 5 feet 1 inch ; the radi- 
 al depth 5 inches, and the length 14 
 
 Fig. 23 
 
 inches. There are twenty-eight coils, 
 each containing 21 turns, so that the 
 total number of active wires, counted all 
 round, is eleven hundred and seventy- six. 
 The field-magnet cores are of wrought- 
 iron, 3 inches by 13^ inches, and are 
 bolted to a wrought-iron yoke-ring y. 
 Each magnet-coil contains 130 turns of 
 
103 
 
 exciting wire, and the exciting current is 
 28 amperes. This gives an exciting 
 power of 7,300 ampere turns, and the 
 calculated strength of field is 144 English 
 lines. Allowing for self-induction and 
 resistance, the induced electromotive 
 force may be taken as 2,050 volts. A 
 continuous-current machine of the same 
 dimensions and winding would give 900 
 volts ; so that for this machine, as 
 arranged to produce an alternating car- 
 rent, the coefficient is k = 2.280. The 
 width of the field, allowing for the fringe 
 of lines round the pole pieces, is 0.55 of 
 the pitch, and the surface covered by the 
 active wires in the armature-coils is 0.55 
 of the whole surface. Theory would there- 
 fore indicate that the coefficient must be 
 very nearly that given for case 5, p. 22, 
 namely, 2.300, which is indeed the case. 
 In a 150-kilowatt machine of the same 
 type, the output is 75 amperes at 2,000 
 volts pressure ; the speed being 500 
 revolutions per minute, and the frequency 
 100. There are twenty-four poles, and 
 the number of active wires is six hun- 
 
104 
 
 dred .and twenty- four. Resistance of 
 armature, 0.32 ohm, and that of the 
 magnets 25 ohms ; exciting current, 15 
 amperes. 
 
 THE PARSONS ALTERNATOR. 
 This machine, of which Fig. 24 is a 
 
105 
 
 section, and Fig. 25 an elevation, is a 
 modification of the well-known turbo- 
 electric generator of the Hon. Charles 
 Parsons. It is the only example of a 
 bi-polar alternator with drum arma- 
 ture, the arrangement being inter- 
 mediate between that shown in Fig. 4 and 
 Fig. 5, that is to say the polar surfaces, 
 although not completely surrounding the 
 armature, yet embrace more than half 
 the circumference. The armature core 
 consists of thin iron plates insulated 
 with paper and mica, and threaded upon 
 the spindle. The wire is laid on in a 
 single-coil, the ends being brought to 
 two collecting rings in the usual way. 
 In a 75-kilowatt alternator, giving 75 
 amperes at 1,000 volts terminal pressure, 
 the armature core is 7 inches in diameter 
 by 30 inches long, and the core of the 
 field-magnets, which are of cast-iron, is 5 
 inches by 28 inches. The armature coil 
 contains one hundred and ten active 
 wires, 127 millimeters in diameter, giving 
 a current-density of 6,000 amperes per 
 square inch. The armature resistance is 
 
106 
 
 0.22 ohm, and the loss 1.65 per cent, of 
 the output. The speed of this machine 
 is 6,000 revolutions a minute, giving a 
 frequency of n = 100. The armature 
 weighs 360 Ibs., and the weight of the 
 complete machine, including the turbo- 
 motor, is about 2 tons. 
 
 THE ZIPERNOWSKY ALTERNATOR. 
 
 This machine is very similar to the one 
 just described, the principal difference 
 being that, instead of a smooth core, the 
 armature is provided with a core having 
 Pacinotti projections on the inside. The 
 magnet cores N S, Fig. 26, are composed 
 of U-shaped iron plates, to avoid the heat- 
 ing which would otherwise result from 
 the employment of an armature with pro- 
 jections. The core of the armature is 
 formed of thin sheet-iron segments A, 
 each resembling a very short or shallow 
 T, the armature- coil C being placed over 
 the central stem of the T. Each arma- 
 ture-segment with its coil forms a sepa- 
 rate part, which can be inserted or with- 
 drawn without disturbing the rest of the 
 
107 
 
 machine. This is a great practical ad- 
 vantage, because facilitating repairs. 
 There is a slight increase in magnetic 
 resistance due to the want of continuity 
 of armature-core , but, as the interpolar 
 space can be made very small, the 
 increase of magnetic resistance in the 
 armature-core is permissible. 
 
 Fig. 26 
 
 ZIPERNOWSKY 
 
 The following particulars refer to a 
 machine giving 40 amperes at 2,000 volts 
 terminal pressure, or 80 kilowatts output. 
 There are fourteen poles in the field, and 
 fourteen coils in the armature. The 
 speed is 360 revolutions per minute, and 
 the frequency is n = 42. The weight of 
 laminated iron in field and armature col- 
 
108 
 
 lectively is 1 ton 7 cwt., and that of 
 copper in the armature and field- coils 
 collectively is 930 Ibs. The armature 
 resistance is 1.038 ohm, and that of the 
 field coils is 3.24 ohms ; exciting current, 
 28.7 amperes. The loss of energy in the 
 armature is thus 2.08 per cent., and that 
 in the field 3.33 per cent, of the output. 
 Some experiments were made with this 
 machine to determine the loss due to 
 windage, mechanical friction, and mag- 
 netic friction. For this purpose the 
 machine was driven by a belt from an 
 electromotor, the efficiency of which had 
 previously been determined. For the 
 first experiment, the field of the alterna- 
 tor was not excited, and the power 
 required to drive it, at its normal speed 
 of 360 revolutions per minute, was found 
 to be 4.07 HP. In the next experiment 
 the field was excited so as to produce a 
 terminal pressure of 2,000 volts ; but no 
 current was allowed to fiow through the 
 armature, and the power required was 
 found to be 9.81 HP. From these 
 experiments it appears that the loss 
 
109 
 
 through magnetic friction was 5.74 HP. 
 The total commercial efficiency at full 
 load, when the alternator is driven by a 
 belt, and inclusive of the power required 
 for exciting, is given by the makers as 
 87 per cent. When driven direct by a 
 high-speed engine, the efficiency is 
 slightly greater, as there is no increase of 
 journal friction by reason of the strain in 
 the belt. 
 
 , DISCUSSION. 
 
 Sir George B. Bruce, President, said 
 the paper was evidently the work of a 
 perfect master of the subject of which he 
 treated, and he was sure that the mem- 
 bers would award the author a vote of 
 thanks for so valuable a contribution; 
 nor did he think that he was sanguine 
 in anticipating that it would give rise to 
 a very important discussion. 
 
 Mr. Gisbert Kapp was afraid the mem- 
 bers had found the paper somewhat dry ; 
 but he had asked several friends to 
 enable him to render the proceedings a 
 little less dry by lending him some appa- 
 
110 
 
 ratus. Most of the exhibits were new, 
 and were then shown for the first time. 
 Amongst the transformers was Mr. Mor- 
 dey's, consisting of rectangular iron 
 plates and bridge-pieces, the long rect- 
 angular coils being threaded through 
 the cavities thus formed. The construc- 
 tion was very simple. The metal stamped 
 out of the rectangle formed a bridge- 
 piece, thus completing the magnetic cir- 
 cuit. The transformers were built up in 
 such lengths as would give the required 
 pressure, and the same plates were used 
 for various sizes. If half the pressure 
 was wanted, half the number of plates 
 were used. There were other trans- 
 formers exhibited, two by Messrs. Lowrie 
 and Hall, which were used in Eastbourne, 
 and were about to be used in London, 
 and also two of the author's type. The 
 core of the Lowrie-Hall transformer was 
 composed of iron plates insulated with 
 varnished gauze, and brought together 
 over the ends, where they were protected 
 by iron caps. The coils were wound on 
 both limbs. There was a 20-HP. trans- 
 
Ill 
 
 former of his own of last year's pattern, 
 made by Messrs. Goolden and Co., and 
 another, the latest pattern, made by 
 Messrs. Johnson and Phillips. The 
 difference between them was that one 
 was adapted for indoor and the other for 
 outdoor use. They could act under 
 water, and had been so worked for trial, 
 but of course no one would think of 
 working them in that way permanently. 
 In the new pattern transformer the wind- 
 ing was split up, so that the difference of 
 potential of any particular coil would 
 only be a portion of the total pressure. 
 The arrangement had another advantage, 
 If transformers were wanted for two- 
 different voltages, say 50 or 100, the 
 same instrument could be used, as the 
 two secondary coils could be coupled in 
 parallel for the lower and in series for 
 the higher pressure. Some iron tubes 
 and steel boxes were shown, as used by 
 Messrs. Lowrie and Hall for distribution 
 by underground mains, drawn through 
 these tubes. There were also samples of 
 lead cables, the joints of which were 
 
112 
 
 made in an extremely ingenious way. 
 The lead was first stripped off, and the 
 insulation removed over a distance suffi- 
 cient for making the joint; when that 
 had been done the insulation was re- 
 placed, a plate of lead was put over it, 
 and a tool like a pair of nippers took it 
 round and squeezed it into the shape 
 shown, making a very strong joint. The 
 connections were made in the boxes, 
 which were of three different types. 
 Among the exhibits was one of Mr. 
 Lowrie's meters, containing a secondary 
 battery in series, with a set of copper 
 depositing plates and the consumers' 
 main leads. It did not measure currents, 
 but conductivity-hours. If the lamps 
 were switched on, but no alternate cur- 
 rent was sent through, and the circuit 
 was simply completed, the meter regis- 
 tered conductivity-hours ; but as the 
 pressure was always on the mains, that 
 was equivalent to registering energy. It 
 had been in use for two years with good 
 results. There was also an ampere 
 meter for alternate currents. The cur- 
 
113 
 
 rent was measured by the heating effect 
 upon compound strips of metal. The 
 heat caused the strips to deflect, and the 
 motion was recorded by a pointer. There 
 was also shown a safety fuse which 
 would be melted, and thus interrupt the 
 current when it exceeded a predeter- 
 mined limit. With a pressure of 2,000 
 volts it was necessary to have a long 
 fuse-wire, because the arc set up upon 
 the melting of the fuse would remain, 
 unless its length were considerable. As 
 a further precaution to make the arc 
 cease, the wire itself was within a slot in 
 the slate base, and the arc was thus sub- 
 jected to the chilling effect of the cold 
 slate. He was able to exhibit a portion 
 of a switch of Mr. Lowrie's installation 
 in Kensington, by which the connection 
 between the different dynamos and cir- 
 cuits could be effected. Also an ex- 
 tremely neat arrangement for showing 
 the insulation of the cables. It was 
 nothing more than a Geissler tube. One 
 pole was put to earth, and the other was 
 put to any one of the circuits which were 
 
114 
 
 to be tested. If the insulation was good 
 on that particular circuit, the tube 
 showed a bluish light, which was ob- 
 served through the eye-hole. If it was 
 faulty, this static effect did not take 
 place; the tube remained dark, and the 
 attendant knew that the particular main 
 was faulty, and must be repaired. If 
 complicated tests were used, requiring 
 elaborate apparatus which could not be 
 easily carried about, the men would be 
 sure to neglect it. With a little instru- 
 ment like that exhibited, two or three 
 circuits could be tested in half a minute. 
 The author also exhibited an armature- 
 coil of one of Mr. Mordey's dynamos 
 shown in Fig. 15 ; also various measur- 
 ing instruments, a safety apparatus in- 
 vented by Captain Cardew, and various 
 photographs of the large dynamos made 
 by Mr. Ferranti. 
 
 Mr. W. H. Preece said that it might 
 not be out of place if he were to give a 
 brief account of the steps, taken in re- 
 cent years, to solve the very difficult 
 question of the economical distribution 
 
115 
 
 of electricity over large areas. The 
 paper was, as the author had himself 
 said, dry; but it really recounted the 
 gigantic advances that had been made 
 in the means required for this economical 
 distribution of electricity. It was only 
 ten years since Edison solved the ques- 
 tion of incandescent lamps, and only 
 seven years ago it was found scarcely 
 possible to distribute currents over areas 
 so as to bring electric lighting to com- 
 pare in any way with gas. At that 
 time a distinguished French electrician, 
 Mr. Gaulard, showed how, by using alter- 
 nate currents of high electromotive force, 
 to distribute electrical energy to a distance 
 by stepping down, as it were, from high 
 pressure to low pressure. The process 
 was similar to that now carried out in 
 London, to distribute power by means of 
 water under very high pressure ; also to 
 that adopted by the Gas Light and Coke 
 Company at Beckton, to send gas under 
 high pressure to holders in different 
 parts of London, and there distribute it 
 under low pressure. Such progress had 
 
116 
 
 been made in America, on the Continent, 
 and in England, that ere very long he 
 believed there would be electric lighting 
 all over the country. There was hardly 
 a town of any size that was not consider- 
 ing the subject, and that, in twelve 
 months' time, would not be on the road 
 to having electric lighting distributed 
 throughout large areas. At Deptford, 
 there was the extensive enterprise of the 
 London Electrical Supply Corporation, 
 and the enormous machines of Ferranti 
 were being constructed, intended to pro- 
 duce an electric pressure equal to 10,000 
 volts, representing an electromotive force 
 something like that of a small lightning 
 flash. It would be driven through a 
 very small conductor up to the neighbor- 
 hood of the Monument, and would there 
 step down to a lower pressure, and so be 
 distributed throughout the city of Lon- 
 don over a very large area. One was the 
 proper rate a"t which the alternating cur- 
 rents should be transmitted. The prac- 
 tice had hitherto been to commence at a 
 very high rate. Mr. Ferranti began with 
 
117 
 
 about 150 currents (300 positive and 
 negative) per second, but by practice he 
 had come down to the figure given by 
 the author, 78. He had omitted to men- 
 tion that throughout Europe the alter- 
 nate-current machines were being con- 
 structed principally by the firm of Ganz, 
 of Budapest. The Zipernowsky system 
 had been introduced, first in Italy, and 
 there worked out thoroughly by a dis- 
 tinguished electrician, Professor Ferraris. 
 In the Zipernowsky system the frequency 
 had been brought down from 150 cur- 
 rents to 42. Mr. Parker had brought it 
 down to 100. That was one of the most 
 important questions to be decided, and it 
 could only be done by actual practice. 
 Sir William Thomson had shown, by 
 mathematical reasoning, that if currents 
 were sent with high frequency, they had 
 not time to penetrate into the interior of 
 a conductor. With a solid conductor, 
 say between Deptford and London, one 
 inch thick, at a high frequency, the cur- 
 rent would not enter more than about 3 
 millimeters inside the surface ; thus the 
 
118 
 
 only part of the copper conductor con- 
 veying electricity was a thin shell or tube 
 on the outside. If that theory was true, 
 the lower the frequency, the greater 
 the efficiency of the conductor. Another 
 reason for less frequency was that meters 
 could work with greater accuracy. 
 
 Professor George Forbes congratula- 
 ted the author on the many points of 
 interest he had brought forward, and 
 especially on the happy way in which he 
 had taken in hand the theoretical inves- 
 tigation of alternate-current machines. 
 He had been one of the first to discuss 
 the theory of continuous-current machines 
 in the proper fashion, and since he read 
 his paper on the subject before the 
 Institution two years ago, the progress 
 had been marked, so that now the design 
 of continuous- current machines was as 
 straightforward work as any mechanical 
 designs. He had taken up the subject of 
 alternate-current machines in the same 
 spirit. On the occasion when the author 
 read a paper on continuous -current 
 machines. Professor Forbes had occasion 
 
119 
 
 to complain of the combination of differ- 
 ent kinds of units, C. G. S. and English. 
 He was sorry to find the same practice in 
 the present paper, and he wished to 
 enter his protest against it. The author 
 spoke of the mean value of the electro- 
 motive force of an alternating current 
 as the electromotive force of a direct 
 current, which will, in a given re- 
 sistance free from self-induction (say, 
 for instance, the wire of a Cardew 
 voltmeter), produce the same amount of 
 heating. That was not a definition of 
 the mean value of the electromotive 
 force of the alternating current. He 
 would beg the author to withdraw the 
 expression " mean value," and substitute 
 " virtual or equivalent value." Until a 
 few years ago, electricians thought that 
 a continuous current was the proper one 
 to use for distribution, and that the 
 alternate current was objectionable for 
 many reasons. But now they were quite 
 agreed that the alternate current was in 
 many of its adaptations far more beauti- 
 ful and more readily adapted than the 
 
120 
 
 continuous current. It was due to 
 Messrs. Gaulard and Gibbs that a system 
 of transformation from high to low pres- 
 sure was economical, and could be 
 adopted in central-station distribution. 
 If any one doubted the value of the work 
 done by those gentlemen, he would ask 
 what would be the present position of 
 electric lighting if Messrs. Gaulard and 
 Gibbs had not, during two or three years 
 of persistent opposition on the part of 
 electrical engineers, forced upon electri- 
 cians the conclusion that the use of their 
 secondary generators was an economical 
 mode of distributing electricity? While 
 great credit was due to the author for 
 having introduced his theoretical views 
 on the question of alternators, it was 
 unfortunate that he had not given more 
 numerical facts as to alternating 
 dynamos, as to the efficiency of the 
 transformers, and so. on. Such data 
 were very desirable. The author had 
 raised the question whether dynamos 
 ought to be worked in parallel or 
 separately. On the ordinary contin- 
 
121 
 
 uous system, as first generally adopted 
 by Edison and now universally employed, 
 it was customary to have all the mains in 
 the district connected together into a 
 large network, with feeding cables going 
 to different points. Those feeders, all 
 issuing from the central station, were 
 connected to one set of mains, and all the 
 dynamos were connected in parallel to 
 those mains, all working together. There 
 was a difficulty in alternate currents work- 
 ing in parallel, and it was customary 
 not to make a single network over the 
 whole of a district, but to subdivide it 
 into small districts, each with its separate 
 feeder, and each feeder might be fed by 
 a separate machine or a number of the 
 feeders might be grouped on to one 
 machine. At the beginning of the 
 evening's work all the feeders would be 
 upon one machine. As the consumption 
 of electricity increased, some of the 
 feeders would be passed on to another 
 machine, and so on. The author con- 
 sidered it impossible to switch the feeders 
 on to a new dynamo without making 
 
122 
 
 some flickering in the light. In the 
 course of last year Professor Forbes had 
 examined a very large number of central 
 stations, and the alleged difficulty did 
 not exist. The facility of switching in 
 dynamos without producing a flicker 
 depended simply on the kind of switch 
 employed. With a rapid switch there 
 was no difficulty. He had lately examined 
 the central station at Borne, set up by 
 Messrs. Ganzand Company, and there was 
 there a switch-board of very great 
 ingenuity, designed by Mr. Blathy. As 
 to the possibility of working conveniently 
 in parallel he might say that experience 
 in America had been completely against 
 it. It was there found that it was possi- 
 ble to work in parallel, but that it 
 enormously increased the amount of 
 skilled attention required in a central 
 station ; and for that and other reasons 
 it was far better to divide the district 
 into a number of sub-districts, feeding by 
 separate feeders from the station. That 
 method had also another advantage, 
 especially in a country like America, 
 
123 
 
 where overhead wires were chiefly em- 
 ployed. There was another reason why 
 in America it had been preferred not to 
 work in parallel, and that was that the 
 alternations of current in the machines 
 were more frequent, and such machines 
 could not be worked in parallel so easily 
 as those which had a small number 
 of alternations. Dynamo machines for 
 alternate currents might be divided into 
 two classes, those which had little or no 
 self-induction, and those which had large 
 self-induction. The Siemens machine 
 was one of those with little or no self- 
 induction, and the Mordey machine was 
 another of the same type. The armature 
 of the Mordey machine looked very like 
 that of the Siemens machine ; but it was 
 a fixture, and the Mordey machine had a 
 peculiarity, in common with the Kennedy 
 machine, which made it extremely origi- 
 nal. In machines of the Siemens and 
 Mordey types, where there was no self- 
 induction, if it was desired to introduce 
 it, all that had to be done was to put 
 large self-induction into the circuit. This 
 
124 
 
 was as convenient as to put the self-in- 
 duction into the machine itself. Still, 
 there was a good deal to be said in favor 
 of machines of iron. One advantage was 
 that the clearance between the poles and 
 the armature was smaller, and there- 
 Figs. 27 
 
 SINGLE-COIL ALTERNATOR. 
 Scale T V . 
 
 fore the magnetization could be got at a 
 cheaper rate. One of the great advances 
 in alternators of ]ate years was the in- 
 vention of the Mordey machine. * Up to 
 that time alternate- current machines were 
 
125 
 
 made with a large number of pole-pieces. 
 The Mordey machine had two ; in the 
 old type each pole-piece had its magnet- 
 izing coil, and there was great waste of 
 energy. Mr. Mordey had introduced 
 the idea of using a single coil to magnet- 
 ize the whole machine. Professor Forbes 
 had been experimenting a little in that 
 direction, and had thought it might be 
 worth while to exhibit a rough diagram 
 of a design, of a somewhat similar kind, 
 in which not only was a single coil used 
 for the magnetizing circuit, but a single 
 coil was also used for the induced circuit 
 (Figs. 27). The induction through the 
 radial parts of the field magnet was con- 
 stant. To prevent heating, it might be 
 necessary to subdivide the iron at the 
 pole-pieces into circular sheets ; but that 
 would be seen when the first machine 
 was made. That type of machine was 
 certainly of interest, and might prove of 
 value. The Parsons machine seemed to 
 him to fulfill the requirements of an 
 alternator better almost than anything 
 else which had been produced. It also 
 
126 
 
 had the advantage of having a single 
 current for the field magnet, and a single 
 current for the induced armature. He 
 would only say a word on the subject of 
 the speed of alternations. Why was it 
 that in America 16,000 alternations per 
 minute were used, and in Europe at the 
 greatest number of stations 5,000 1 One 
 reason was, that in America high-speed 
 machines were in favor, and in Europe 
 low speed. Another reason was that, in 
 America, the object aimed at was to get 
 the greatest output from the plant. 
 Those reasons were sufficient for having 
 high-speed alternations. Unless it was 
 wished to work in parallel, it could not 
 be doubted that working with the high- 
 est-speed alternations was best ; but for 
 working in parallel there could be no 
 doubt that the lowest speed was the 
 most easy mode. That was one reason 
 why the Ganz machines at Rome were so 
 suitable. In this country there was a 
 general belief that converters had an 
 efficiency of 95 or 96 per cent, on full 
 load ; but the greater number of con- 
 
127 
 
 verters used in England had nothing like 
 that efficiency. The smaller converters 
 used by Messrs. Ganz and Co. certainly 
 had nothing like it ; they guaranteed an 
 efficiency of 88 per cent. The machines 
 exhibited, the Lowrie converter, the 
 
 Fig. 28 
 
 CURRENT OUTPUT. GROSVENOR GALLERY. 
 
 Kapp converter, and still more the Fer- 
 ranti, had all a high magnetic resistance, 
 which he was inclined to think would be 
 unfavorable. He knew that tests had 
 been made with a Kapp converter by 
 Professor Ayrton, but in totally differ- 
 
128 
 
 ent conditions from the machines shown, 
 and he should be surprised if they had 
 the same efficiency as that shown by 
 Professor Ayr ton. He might take one 
 example, which was very prominent, 
 namely, the converters used by the Gros- 
 venor Gallery. The diagram (Fig. 28) 
 showed the output of amperes in a day, 
 and it would be seen that between four 
 and seven hours in the morning, when 
 no lamps were on circuit, there was an 
 indication of four thousand lamps being 
 on circuit according to the ampere- 
 meters of the central station ; that meant 
 a waste of nearly four thousand lamps 
 always going on, and with higher loads 
 the loss was greater. The loss as shown 
 in the diagram was 20 per cent, of the 
 maximum with no load ; consequently 
 when there was any load on, the efficiency 
 was certainly not 80 per cent. 
 
 Mr. Llewelyn Atkinson said that the 
 constants which the author had de- 
 veloped, and which were very useful in 
 comparing the electromotive force of 
 alternating machines with machines of a 
 
129 
 
 continuous type, seemed to him might be 
 made more useful if they were more 
 separated. The constant K appeared to 
 contain three things. First, it contained 
 a multiplier of 2, introduced because 
 alternate-current machines generally had 
 only a single circuit, not two circuits, but 
 this was not always the case, and the 
 multiplier 2 should be put separately. 
 Secondly, the constant included a quan- 
 tity depending on the shape of the field 
 and the arrangement of the armature; 
 and thirdly, the quantity depending on 
 the ratio of the square root of the mean 
 square to the mean. All those three 
 quantities appeared to be mixed up in 
 one number, which might be useful in 
 the particular cases given, but he thought 
 that if they were dealt with as he sug- 
 gested they would be more useful to 
 subsequent designers. Nearly all cases 
 which the author had treated so lucidly 
 as to connecting machines in parallel, 
 had been by the graphic method, which 
 depended entirely upon the assumption 
 that the electromotive force followed a 
 
130 
 
 law of simple harmonic motion. He was 
 aware that there were great difficulties in 
 treating it in any other way. First of all 
 the law itself had to be found ; and 
 secondly, he did not know of any graphic 
 method of solving the problems when 
 the law had been found. The analytical 
 method was probably even more difficult. 
 It was generally complicated enough 
 even with a simple harmonic law, but 
 with irregular laws he did not know that 
 it could be touched at all. He had 
 therefore brought a diagram showing 
 what were likely to be the errors even in 
 the machines the author had treated of, 
 taking the harmonic law to be true. In 
 Figs. 29, the curves 1, 2, and 3 showed 
 the primary electromotive force induced 
 in a coil all wound in a single line on the 
 armature, by moving it through three 
 different forms of magnetic field, a har- 
 monic field, a field such as occurred in 
 the Mordey machines, and a field such^as 
 was given by the author's machine with 
 opposite poles alternately on each side 
 of the ring, with a space equal to the 
 
131 
 
 width of pole. 1. Gave a curve of 
 primary current in a circuit with self- 
 induction which was also'harmonic, and 
 a curve of secondary electromotive force 
 also harmonic. 2. Gave a curve of pri- 
 mary current similar in form to the 
 
 Figs. 29 
 
 Primary 
 
 curve of electromotive force, but with all 
 the corners rounded and the rises and 
 falls less abrupt ; but the curve of 
 secondary electromotive force was differ- 
 ent, and showed rapid rises and falls 
 and intervals of no electromotive force. 
 
132 
 
 3. Gave the same class of results as re- 
 garded primary electromotive force and 
 current ; but here also the curve of 
 secondary electromotive force was very 
 different. Owing to the fall in positive 
 value of electromotive force being dis- 
 continuous with the rise in negative 
 electromotive force, there were two im- 
 pulses of secondary electromotive force, 
 corresponding to each impulse of primary 
 electromotive force. The effect of this 
 was to double the loss from magnetic 
 friction, or hysteresis, in any magnetic 
 apparatus in the secondary circuit. That 
 at once showed the very important bear- 
 ing of the question of the law of elec- 
 tromotive force, and the errors that 
 might be introduced by assuming a 
 simple sine-function. He had alluded to 
 the difficulty of treating the matter ana- 
 lytically and graphically ; but there was 
 one mode of treating it which he had not 
 seen mentioned, the idea of which oc- 
 curred to him two years ago, and he had 
 designed some machines for carrying it 
 out, namely, treating the subject by means 
 
133 
 
 of the integrator (Figs. 30). To take the 
 simplest case of a circuit, of which the 
 induction and resistance were known, and 
 
 \ 
 
 \ 
 
 \ 
 
 the law of electromotive force was given 
 by means of a curve. Let a b c d be an 
 integraph in skeleton consisting of a rect- 
 
134 
 
 angular frame. In this frame slid a 
 cross-bar e' /, and within an inner slide 
 a b g A, a carrier through which was 
 pivoted, on a vertical axis, an integrating 
 roller. The angular position of this roll- 
 er was fixed by a rod k e, sliding at e 
 through a pivot 6, free to move round in 
 a horizontal plane. The whole frame 
 could only move in a direction parallel to 
 the axis of x, being constrained by guides 
 or rollers. Through the bar e' f at m 
 passed a pointer. Let x t y t be co-ordi- 
 nates of a point on the curve of primary 
 electromotive force, let x y be co-ordi- 
 nates of the curve traced by the non- 
 slipping integrator roller. Then by the 
 construction of the apparatus the follow- 
 ing relations held good. The roller was 
 always tangential to the curve x y, and 
 
 
 d x me 
 
 Let the scale of y be such that y 
 
 Tjl T 
 
 = _, and let m e = _, where L and K 
 
 II B 
 
 were coefficients of self-induction and 
 resistance. 
 
135 
 
 - -y 
 
 Then ^ = ? -, and if the co-ordi- 
 
 dx L 
 
 5 
 
 nates x y represented current and time 
 
 --C 
 
 respectively, = 5 , or L 4- C 
 
 at L d t 
 
 II 
 
 E = E, the well-known equation of a 
 single electric circuit. Thus, by means 
 of the integrator, any curve of electro- 
 motive force being given, it was possible 
 to draw the curve of the current. He 
 had further designed machines which it 
 would take too long to describe, in which 
 with two circuits, given the coefficients 
 of self-induction, the coefficient of mutual 
 induction, and the resistances, with a 
 given curve representing the primary 
 electromotive force, the curve of primary 
 current, the curve of secondary electro- 
 motive force, and the curve of secondary 
 current could be deduced ; and all the 
 curves, their shapes, their retardation, 
 and everything about them. 
 
136 
 
 Mr. James Swinburne considered that 
 the author's papers gave rise to very 
 good discussions, partly because they 
 were of great value, partly because he 
 always selected a subject which was 
 fashionable at the moment, and partly 
 because he always gave a good deal to 
 disagree with. With regard to the 
 author's assumption of the curve of sines r 
 he could not see how that afforded any 
 help. In practical work, the only use he 
 had ever found in the curve of sines was 
 in seeing whether an instrument, which 
 had really a coefficient of self-induction r 
 would give a reading within a certain 
 percentage error. But in dealing with 
 machines the assumption seemed to him 
 to be an absurdity. It was only taken 
 because it admitted of a great many 
 mathematical calculations; but if the 
 mathematical calculations were based 
 upon empirical data he thought it was a 
 pity to make them. Of course it was 
 continually said that by Fourier's theo- 
 rem any curve could be made up of 
 curves of sines. In practical work elec- 
 
137 
 
 tricians never dealt with anything that 
 had constant coefficients of self-induction 
 or mutual induction, except measuring 
 instruments ; and in dealing with trans- 
 formers, incandescent lamps, and so on, 
 nothing of the sort came in. He thought 
 that all the comparisons in the paper 
 between direct and alternate-current 
 machines were based on an assumption 
 which made them misleading. The 
 assumption was that the direct machines 
 had the same number of armature turns 
 as the corresponding alternate machines. 
 In practice he did not think that any- 
 body ever made the air-gap otherwise 
 than equal to the breadth of the pole. 
 The output varied as the product of the 
 number of useful wires in a coil, and the 
 breadth of the pole-piece. The product 
 was greatest when they were equal, and, 
 consequently, designers always made 
 them equal. The fringe of useful field at 
 the edge of the pole-pieces must be 
 allowed for ; and in some cases the ques- 
 tion was complicated by the sides of the 
 coils being parallel, and the sides of the 
 
138 
 
 pole-pieces radial, as in the author's 
 machine. The author assumed a con- 
 stant coefficient of self-induction in the 
 armature. In the machine which had 
 originated the unwarrantable assumption 
 of the curve of sines the self - induction 
 was nearly constant. It was a Siemens 
 machine used by Mr. Joubert. There 
 was very little iron near the armature, 
 and the coefficient of self-induction was 
 sensibly constant. In the modern ma- 
 chine there was not a constant coefficient 
 of self induction in the armature. This 
 would, he was afraid, detract from the 
 usefulness of Mr. Atkinson's ingenious 
 application of the integraph. The sim- 
 plest way of calculating these things was 
 not to take the coefficients of self-induc- 
 tion, but to find out how many lines of 
 force were cutting the circuit at any 
 time, by plotting the field-magnets out, 
 and working through a graphic method. 
 It was a long and tedious process, but it 
 could be done, and done with the same 
 accuracy as, for example, Professor 
 Forbes' dynamo leakage calculations. In 
 
139 
 
 the direct-current dynamo the term self- 
 induction had been a sort of scapegoat 
 for all sorts of vague notions. Even 
 now, in the case of direct-current ma- 
 chines, people talked about the self- 
 induction of the armature. Of course 
 there was no such thing in direct-current 
 machines, and the idea was an absurdity. 
 In alternating currents, the same accusa- 
 tion could not be brought, as the whole 
 armature current varied. With regard 
 to motors, he thought the author's ideas 
 were a little elementary for the present 
 day. With a big motor, running at dif- 
 ferent loads, it was found that the 
 efficiency was extremely small if depend- 
 ence was placed simply on self-induction 
 to keep the motor and dynamo in step. 
 He did not think that motors which de- 
 pended on self-induction for keeping in 
 step would answer. Many seemed to 
 think the chief difficulty was to start the 
 motor. This was really a small matter. 
 All that had to be done was to make a 
 self -exciting motor. Take the commuta- 
 tor machine made by the Thomson- 
 
 
140 
 
 Houston Company, and put it on an 
 alternate-current circuit, it would start, 
 sparking badly at first, and eventually 
 synchronize. There was no difficulty in 
 starting the motor, but it would not 
 work efficiently. The author had asked 
 how it was that the waste power was 
 greater in the armature conductors of a 
 machine without iron in the armature. 
 It was because the change of field was 
 more abrupt in that case. A great deal 
 had been said at different times about 
 the advantage or disadvantage of iron in 
 the armature, and it had been discussed 
 both from a mechanic's and from an 
 electrician's point of view ; but when it 
 was remembered that the air-space (in 
 spite of what Professor Forbes had 
 stated) was really the same whether there 
 was iron in the armature or not, it would 
 be seen that the whole question was one 
 of mechanical construction against elec- 
 trical efficiency. If there was iron, power 
 would be lost in the reversal of mag- 
 netization. If there was none, there 
 must be, to a certain extent, flimsiness. 
 
141 
 
 There might be a more efficient machine, 
 but there was mechanical flimsiness. A 
 great deal had been said about hysteresis. 
 He considered the author's inductions 
 from Professor Swing's paper were faulty. 
 Plotting out the curve from that paper, 
 or from Dr. Hopkinson's paper, it would 
 be found that up to an induction of 
 about 1(>,000, the power spent with a 
 given alternation frequency varied sensi- 
 bly as the induction. The author's 
 argument, that if there was half the 
 induction there was half the heat per 
 cubic centimeter, was all very well, but 
 then twice the iron must be used to get 
 the same output. And it was found that 
 in transformers made according to pres- 
 ent practice with low inductions, in which 
 most of the heat was wasted in the iron, 
 as it was in transformers of any reason- 
 able size, the speed of alternation made 
 very little difference theoretically in the 
 output. He said theoretically, because 
 not enough was known about the time 
 lag, and if that was great, larger outputs 
 could be obtained with few alternations. 
 
142 
 
 He was aware that he was disagreeing 
 with Professor Forbes on that point. He 
 hoped Mr. Atkinson would be able to 
 supply some particulars as to time lag, 
 that he would run the machine at differ- 
 ent speeds and give the results. The 
 kind of time lag mentioned by Professor 
 Ewing was well known to dynamo 
 makers, and the phenomenon was fa- 
 miliarly known as " creeping up." A 
 new machine sometimes took twenty 
 minutes to reach its full excitation. 
 Messrs. Crompton & Company were 
 making experiments on the important 
 question of loss of power in iron in 
 alternate-current work, which he hoped 
 would be published. The question of 
 hysteresis concerned transformers as well 
 as dynamos. He thought that most 
 people used bad iron. In testing iron 
 for direct-current machines they only 
 thought whether, with a given excitation, 
 they could get iron highly magnetized. 
 That was not the important question in 
 alternating currents. The question of 
 hysteresis must be considered. There 
 
143 
 
 was a great difference in samples in that 
 respect. He had made transformers 
 which would work with very high induc- 
 tion without heating. One of the most 
 important questions to settle was fre- 
 quency of alternation. It had been stated 
 that in America 16,000 alternations per 
 minute had been taken, and at Home 
 5,000, and what should be made the 
 standard in England should be carefully 
 considered. In dynamo construction he 
 found that it was easier to make a ma- 
 chine giving a very few alternations. In 
 that way he could construct a larger 
 dynamo for the same amount of money, 
 and make it work more efficiently. The 
 first point to be settled was the number 
 of vibrations per second which could be 
 separated by the eye. Helmholtz, in his 
 " Sensations of Tone as a basis for the 
 Theory of Music," gave 24 per second. 
 Messrs. Crompton were going to try 
 some experiments to ascertain whether it 
 was possible to detect flickering in 100- 
 volt lamps with anything below 24 vibra- 
 tions per second. It was a pity that the 
 
144 
 
 author had not been able to say some- 
 thing about the Tesla motor. The first 
 form was hardly a practical alternating 
 motor, because it required three leads ; 
 and if a third lead was run, a direct cur- 
 rent might as well be used, with motors 
 which it was known could work efficiently. 
 But he thought that he had heard some- 
 thing about a lagging transformer, and 
 he should like to know more. The 
 author had called the Mordey alternator 
 an iron-clad machine. He considered 
 the word iron-clad a bad one to use. It 
 was employed some time ago to denote a 
 machine in which the iron was wrapped 
 round the copper ; the idea was that a 
 line integral of the magnetic force of 
 more than 0.4 n time the ampere turns 
 could somehow be obtained ; but that 
 fallacy had been dropped, and it was a 
 pity to use the word for the Mordey 
 machine, where the only object was to 
 get the whole field through one exciting 
 coil. The fault of the Kennedy machine, 
 as far as it had any, was the want of ven- 
 tilation. One of the great beauties of the 
 
145 
 
 Mordey machine was that it could be run 
 with a high-current density without fear. 
 It had occurred to him that if an iron 
 core were put into the Mordey machine, 
 and the field-poles were arranged to 
 come alternately, as in the Kennedy 
 machine, instead of opposite, the power 
 of the machine would be doubled without 
 greatly increasing the expense. He 
 agreed with Professor Forbes as to the 
 credit due to Messrs. Gaulard and Gibbs, 
 which, he thought, ought to be frankly 
 acknowledged. 
 
 The Hon. Charles Parsons said, as the 
 calculations in. the paper were based on 
 the sine-function, he would refer to a few 
 figures in connection with a bipolar 
 alternator, in which the sine-function was 
 nearly realized. With two poles, and 
 one continuous coil, or two half-coils, as 
 on the drum of the armature, Fig. 8, the 
 electromotive force was a sine-function, 
 and the self-induction was considerable. 
 In a contemplated machine, the output 
 of which was intended t to be 75,000 watts, 
 the current to excite the magnets would 
 
146 
 
 be about 1 per cent, of the output, and 
 the loss in resistance of the armature 
 about 1^ per cent. This machine (Fig. 
 31) was not yet completed, but one of 30 
 units had been made and carefully tested. 
 With regard to the question of hysteresis, 
 he had made experiments upon the loss 
 in the armature. A cylindrical armature - 
 core, 5 inches in diameter, composed of 
 one hundred iron plates to the circle, 
 insulated from each other by thin paper, 
 but not insulated from the spindle, was 
 rotated in a bi-polar field. The energy 
 required to rotate the core when sub- 
 jected to different intensities of field, and 
 at different speeds of rotation, was meas- 
 ured by independent means. In one 
 experiment the same core was rotated, 
 first at 8,500 revolutions per minute, and, 
 secondly, at 6,500, the strengths of field 
 being maintained in the two cases in- 
 versely proportional to the speeds ; or, 
 in other words, the electromotive force, 
 had the core been wound with conduct- 
 ing wire, would have been identical in the 
 two cases. It was found that the power 
 
147 
 
 lost by hysteresis at the lower speed was 
 approximately 25 per cent, greater than 
 at the higher speed. Taking the energy 
 absorbed per revolution in the two cases, 
 it followed that the loss per reversal was 
 approximately proportional to the square 
 of intensity of field. Some other ex- 
 periments had been made which went to 
 show that the viscous hysteresis, or the 
 viscous resistance of the iron to rapid 
 
 Fig, 31 
 
 TURBO ELECTRIC-GENERATOR FOR 75,000 WATTS 
 FOR ALTERNATING CURRENTS. 
 
 change of polarity, was a comparatively 
 small quantity. The intensity of field in 
 the above experiments was between 5,000 
 and 7,000. At 6,000 revolutions of the 
 apparatus the loss would be 2 HP., which 
 was about 2 per cent, of the output, mak- 
 ing a total electric efficiency of 9-> or 96 
 per cent, for the dynamo. The machine 
 
148 
 
 was arranged with an electric- control 
 governor, which was compounded so as to 
 keep the electromotive force at the 
 terminals constant at all loads. Though 
 the machine revolved at a high rate of 
 speed, the wear and tear was exceedingly 
 small, the same bearings having been 
 running for three years without any per- 
 ceptible wear. 
 
 Mr. W. B. Esson observed that most 
 of the points he might have referred to 
 had been taken up by other speakers. 
 The induction-curve in the Mordey alter- 
 nator would differ considerably from the 
 curves shown by Mr. Atkinson, for two 
 or three reasons. The first was that the 
 coil in the alternator was not merely a 
 rectangle concentrated in one wire, as 
 would be required to give the particular 
 curve in 2, but a number of wires occupy- 
 ing a considerable angular width. An- 
 other reason was that, owing to the 
 trapezoidal form of coil, the induction 
 commenced very gradually, and rounded 
 off the corners in the diagram. The 
 peculiarity of that alternator was that the 
 
149 
 
 induction-curve of individual coils was not 
 symmetrical, but as, in the machine, the 
 coils were joined in pairs, it came out that, 
 though considering any coil by itself there 
 was a want of symmetry, when the coils 
 were joined up all in series, the induction- 
 curve of the whole machine was quite 
 symmetrical ; it was almost exactly a 
 curve of sines. Mr. Mordey had meas- 
 ured the induction through a pair of coils 
 for all different positions of the field- 
 magnet in a whole period, and, as the 
 curves showed that the induction lay 
 within the curves of sines at some parts, 
 so the curve for the electromotive force 
 lay at some parts outside the curve of 
 sines. He should like to have some 
 definite information regarding the amount 
 of power required to reverse the mag- 
 netism in the ring- cores of alternating 
 dynamos for different degrees of mag- 
 netization. He had been making some 
 investigations into the subject, and had 
 found that to prevent the cores of alter- 
 nate-current dynamos from over- heating, 
 the thickness of the iron core for 100 
 
150 
 
 periods per second should not exceed 9 
 centimeters, with a saturation of 7,000 
 C.G.S. units. 
 
 Mr. A. A. Campbell Swinton said there 
 seemed to be an impression amongst 
 engineers generally, that very high-speed 
 machines like the Parsons must sacrifice 
 durability. At Sir William Armstrong's 
 works at Newcastle there were three 
 Parsons machines, two of which ran at 
 about 9,000 revolutions per minute, and 
 the other at about 17,000. They had 
 been running for three years, and neither 
 the bearings nor the spindles had 
 appreciably worn away. He thought 
 that the brushes and the commutators 
 were the only parts which had been re- 
 newed. Of course the Parsons machine 
 was a special machine, and the bearings 
 were specially designed to suit the high 
 speed. The fact that high speed, had 
 not received much attention was not the 
 fault of electricians, but of engineers. If 
 engineers had produced engines that 
 would run at these high speeds, no 
 doubt electricians would have produced 
 dynamos for them. 
 
151 
 
 Mr. F. H. Nalder observed that the 
 insulation resistance between the primary 
 and secondary coils was important in the 
 transformer. It was also of very great 
 moment to apparatus in houses, which, if 
 the insulation should break down, would 
 be liable to cause fire. That in a meas- 
 ure was looked after by most makers, 
 who provided covers for the transformers. 
 Engineers should come to the same con- 
 clusion with respect to the minimum insu- 
 lation resistance between primary and 
 secondary coils, and also between the 
 primary coils and the frame. 
 
 Professor W. E. Ayrton remarked that 
 although some points in the paper might 
 be not quite right, it was of a highly 
 suggestive character. He wished to 
 follow those who had preceded him in 
 condemning the expression "mean elec- 
 tromotive force " employed in the paper. 
 The expression had now a perfectly 
 definite signification in connection with 
 alternate-current circuits ; it was, there- 
 fore, a mistake to use it as a substitute 
 for another expression which had also a 
 
152 
 
 clear significance "the square root of 
 the mean square." The two things differ 
 by many per cent., depending on what 
 function the electromotive force was of 
 the time. The mean electromotive force 
 of an alternate-current and a direct-cur- 
 rent dynamo similarly constructed, must 
 be identical. Whether the wire was piled 
 up all at one place, or whether spread 
 uniformly round the armature, as in a 
 direct- current dynamo, the mean electro- 
 motive force must necessarily be the 
 same. The author spoke of " English 
 lines of force." There was no such thing 
 as an English line of force ; lines of force 
 were essentially cosmopolitan. One of 
 the great charms that the absolute 
 system of electrical and magnetic units 
 possessed was its perfect harmony, and 
 therefore any attempt to introduce new 
 units should be deprecated. Experi- 
 ments made at the Central Institution, 
 for finding the coefficient of self-induc- 
 tion of a transformer, showed that in no 
 case was there no lag between the im- 
 pressed electromotive force and the pri- 
 
153 
 
 mary current. Measurements had also 
 been made by a number of students to 
 ascertain the lag between the primary 
 and the secondary current. The second- 
 ary circuit was sending a current through 
 a non-inductive resistance. For all 
 speeds varying from about 1,600 semi- 
 alternations, up to 16,000 semi-alterna- 
 tions per minute, there appeared to be a 
 lag of about 90 above the lag which 
 ought to be got, which would really cor- 
 respond to the maximum lag possible. 
 In fact, there seemed to be a considerable 
 amount of self-induction in a transformer, 
 even when the secondary circuit was 
 employed to send a current through 
 glow-lamps, or a non-inductive circuit. 
 Without lag, the wave of the secondary 
 current must necessarily be 90 behind 
 the wave of the primary ; but instead of 
 that, the former was 180 behind the lat- 
 ter, so that there were 90 extra lag in a 
 transformer. The same result had been 
 observed by Mr. Blakesley later on, and 
 referred to in a paper read before the 
 
154 
 
 Physical Society.* He did not, there- 
 fore, think that the arrangement could 
 be taken as one which would give a non- 
 self-induction in the external circuit. 
 The author had stated (p. 61), "the 
 coefficient of self-induction was found to 
 be L = 0.955 in C.G.S. measure, and 
 0.0955 in practical units." C.G.S., of 
 course, meant centimeter-gram-seconds ; 
 what " practical " units were he did not 
 know. If they were the ordinary units 
 of self-induction, corresponding with the 
 true ohm, the volt, the ampere, and so 
 on then 1 practical unit was equal to 10 9 
 C.G.S. units or centimeters. The number 
 0.955 given by the author should have 
 been divided by 10 9 or one thousand 
 million ; or more strictly, in consequence 
 of the ohm not being the true ohm, the 
 practical unit of self-induction was 99,777 
 X 10* centimeters. Hence the coefficient 
 of self-induction referred to by the 
 author was 0.9571 X 10" 9 centimeters in 
 practical units, which would be about 
 
 'Proceedings of the Physical Society of London, 
 1888, vol. ix. p. 287. 
 
155 
 
 6,200 miles. Two years ago Professor 
 Perry and he had suggested a name for 
 the practical unit of self-induction, it not 
 having previously received any name. If 
 the practical unit were exactly a thousand 
 million centimeters, the name quadrant 
 would be an appropriate one, but it was 
 about ^ per cent, less than the earth's 
 quadrant, and they had therefore sug- 
 gested the name " secohm " (a contrac- 
 tion of second and ohm). Professor 
 Ayrton next referred to a standard of 
 self-induction which he exhibited on 
 behalf of Professor Perry and himself, 
 and which he thought was the first com- 
 mercial standard of self induction ever 
 issued. It consisted of two coils of 
 platinoid wire, one fixed in position, and 
 the other pivoted so that it could be 
 turned about a diameter, and placed in 
 different positions relatively to the fixed 
 coil. By putting the coils in different 
 positions the arrangement had different 
 and definite coefficients of self-induction, 
 the values of which were indicated on the 
 scale in secohms from 0.013 to 0.036 
 
156 
 
 secohm. There was also an apparatus on 
 the table familiar to some members, called 
 a secohmmeter, for comparing two self- 
 inductions. It rapidly alternated the 
 battery connections and the galvanom- 
 eter connections of a Wheatstone's 
 bridge, and it rendered the measurement 
 of self-induction, which before two years 
 ago usually had been very un sensitive, 
 having had to be effected with single 
 impulses, as sensitive as the ordinary 
 measurement of resistance with a Wheat- 
 stone's bridge. With a combination of 
 those two instruments the coefficient of 
 self-induction could at once be measured 
 with accuracy, without any other appa- 
 ratus except the ordinary Wheatstone's 
 bridge. Fig. 32 showed symbolically 
 the arrangement for measuring a coeffi- 
 cient of self-induction L 2 . G C and B C 
 were the galvanometer and battery com- 
 mutators of the secohmmeter, which were 
 rotated by turning the handle H. Bal- 
 ance for steady currents having been first 
 obtained, the secohmmeter handle was 
 rotated at any convenient speed, and the 
 
157 
 
 self- induction L! of the standard adjusted 
 until balance was again obtained, when 
 
 ~ = . The continuous lines showed 
 Lj r x 
 
 Fig. 32 
 
 MEASUREMENT OF SELF-INDUCTION WITH SECOHM- 
 METER AND SECOHJl STANDARD. 
 
 permanent connections in the secohm- 
 meter, and the dotted lines wires em- 
 ployed to join its terminals with the 
 Wheatstone's bridge. 
 
158 
 
 Mr. T. H. Blakesley said that the 
 paper had considerable breadth ; but, like 
 a dangerous piece of water, it was 
 unequally deep. The subject was a very 
 large one, and perhaps the author had 
 not been able entirely to do justice to it, 
 even in his comparatively long paper. 
 Mr. Blakesley had himself done a good 
 deal of work in diagrams, and the first 
 remark he would make was that in Figs. 
 7, 8, and 9, the direction was continually 
 reversed. In Fig. 7 the motion of the 
 diagrams was supposed to be clockwise, 
 in Fig. 8 it was the reverse ; and in Fig. 
 9 it was supposed to be the same as in 
 Fig. 7. The advantage of a diagram 
 of that kind was that it not only repre- 
 sented the magnitude of certain quanti- 
 ties of electromotive force in general, but 
 the relative positions of their maximum 
 value. The author constantly called O 
 A the current. It was true he guarded 
 himself by saying that it was propor- 
 tional to the current ; but those who 
 read the paper might suppose that he 
 thought O A to be in phase with the 
 
159 
 
 current. It was not so. O K was the 
 real effective electromotive force, requir- 
 ing mere division by the resistance to 
 represent the current, as in ordinary 
 steady flow. In a good diagram, it only 
 
 Fig. 33 
 
 needed a pair of dividers to work out 
 results which, so far as the scale allowed y 
 were as complete as the results of analysis. 
 That was why they were so extremely 
 valuable. He would now mention the 
 details of the diagram (Fig. 33) with 
 
160 
 
 a view of fulfilling the characteristics 
 which he had mentioned, and which 
 represented all cases of the follow- 
 ing problem : Given two alternate-cur- 
 rent machines coupled in series, and 
 having given electromotive forces, and 
 the same known period of alternation, a 
 given resistance in the whole series and 
 a given coefficient of self-induction for 
 the whole series : how will any time- 
 interval between the corresponding 
 phases of the two machines (phase angle), 
 affect (1) The work to be put into the 
 system ; (2) The work to be -got out of 
 the system ; (3) The stability of the 
 motion ; (4) Efficiency (or the relation of 
 (2) to (1)); (5) the work wasted in heat- 
 ing the circuit 1 Proceed thus : With 
 a pair of dividers lay down to any con- 
 venient scale, A B, proportional to the 
 electromotive force of the first machine, 
 and B C (in the same straight line but in 
 the opposite direction), proportional to 
 that of the second machine, and describe 
 a circle upon A B as diameter. Adopt- 
 ing as the positive direction of ro- 
 
161 
 
 tation that which was counter-clock- 
 wise, set off the angle B A D in the 
 negative direction, equal to that of 
 
 electric lag (Tan B A D = =-^, where 
 
 -L L\i 
 
 L = the coefficient of self-induction, B 
 . the resistance of the circuit, and T = 
 the semi-period). From B and C drop 
 perpendiculars upon A D, cutting it in 
 D and 5 (not given in Fig. 33 to avoid 
 confusion of lines). D would obviously 
 be on the circle already drawn. With D 
 as center, and D 5 as radius, describe a 
 circle, called the E-circle, cutting the 
 former circle in Q and E. Then all 
 possible phase-differences between the 
 two machines might be represented by 
 some angle A D E, where E was always 
 upon the E-circle. For any point E the 
 electromotive forces of the two machines 
 were related in phase and magnitude, 
 exactly as were A D, D E. For example, 
 if E was at the point 5, the phase-angle 
 was 380; or the electromotive force of 
 one machine was throughout its varia- 
 tions always opposed by the electromo- 
 
162 
 
 tive force of the other. Further, if for 
 any point E, A E was joined and pro- 
 duced, if necessary, to cut in F the 
 circle on A B as diameter, the values of 
 A E, E F, A F, could be taken with 
 the dividers, and measured on the scale 
 first employed. These lines had the 
 following properties : 
 
 If R was the resistance of the circuit, 
 then 
 
 The power employed on ) _ A E, AF 
 
 ) 
 
 the first machine (input) ) 2 R 
 
 The p 
 
 circuit 
 
 The power heating the ) _ A E 2 
 - } 2 R ' 
 
 AE,EF 
 
 employed on J -- ^p 
 
 the second machine - 
 The latter depending on whether E lay 
 within or without the circle upon A B. 
 If E lay within the circle upon A B, 
 work would be given out by the second 
 machine ; that was, it might be employed 
 as a motor. If, on the other hand, E 
 lay upon that portion of the E-cir- 
 cle without the circle upon A B, 
 power must be employed on the second 
 
163 
 
 machine from outside, and it would do 
 part of the work of heating the circuit. 
 Thus the two problems, of transmission 
 of power and working in series, were 
 really parts of the same general problem. 
 The points Q and B were the positions of 
 E where these problems passed one into 
 the other, namely, where the output was 
 zero. Through D draw a diameter to the 
 E circle parallel to A B, or 3 D 4, and a 
 second diameter through K, the bisection 
 of A B, or 2 D K 1. It would be seen 
 that these two diameters were equally 
 inclined to A D. The points upon the 
 E-circle so obtained had the following 
 peculiarities : If E coincided with 1, the 
 maximum power would be given out by 
 the second machine; if E coincided with 
 2, the maximum power existed which 
 could be employed upon the second 
 machine ; if E coincided with 3, the mini- 
 mum power existed of the first machine ; 
 if E coincided with 4, the maximum 
 power existed of the first machine. Fur- 
 ther, if E lay upon the semi-circle 1 Q 2, 
 there was stability for the second 
 
164 
 
 machine ; and if E lay upon the semi- 
 circle 2 E 1, there was instability for the 
 second machine ; if E lay upon the semi- 
 circle 324, there was instability for the 
 first machine ; and if E lay upon the 
 semi-circle 4, 1, 3, there was stability for 
 the first machine. Thus the region 1, 5 ? 
 3, had a double guarantee of stability. 
 The angle 1 D 3 was twice that of 
 electric lag ; which showed one of the 
 advantages of self-induction. Again, if a 
 line from A to 4 cut the E-circle in 6, 
 this would be the point for E giving 
 the maximum efficiency in the trans- 
 mission of power. It must lie in the 
 region of stability. The point 5 had 
 the property that the efficiency there 
 had the ratio of the two electromotive 
 forces, as in the case of steady cur- 
 rents, but it was not as great at 5 as 
 at 6. Fig. 34 had been constructed in the 
 same way as Fig. 33, and with the same 
 electromotive forces, but with a larger 
 angle of electric lag BAD. It was 
 intended to show by an example that 
 the point 3 did not necessarily lie within 
 
165 
 
 the circle upon A B. If circumstances 
 were as in Fig. 34, then E might come 
 between 3 and Q, and be outside the 
 circle upon A B. Thus there would be, 
 for such a point, stability for both 
 machines at the same time that both 
 
 Fig. 34 
 
 machines would take part in heating the 
 circuit. The possibility of such an 
 arrangement had been denied by Dr. 
 Hopkinson in this Institution, and wide 
 currency had been given to his statement. 
 It should be distinctly understood that 
 the statement was not one of general 
 
166 
 
 truth, and with short periods and large 
 self-induction, two machines in series 
 might easily be run to work with 
 stability. He had only been able to indi- 
 cate the use of such diagrams as he 
 recommended. A very slight addition to 
 the diagrams enabled the effect of hys- 
 teresis to be taken into account. Hyster- 
 esis was the work absorbed in the magnet- 
 ized masses in the field, upon change 
 taking place in the degree or direction of 
 magnetization. This was now a recog- 
 nized fact, and well-known measurements 
 of the effect had. been made by Dr. 
 Ewing and others. On the subject of 
 transformers, he might mention a system 
 of transformation which might be 
 usefully employed in many practical 
 ways. Suppose that an alternate- current 
 machine (Fig. 35) played into a con- 
 ductor possessing, besides resistance, 
 various causes of self-induction. And 
 suppose that at two points A, B, of the 
 circuit were coupled the poles of a con- 
 denser. Then in general it would be 
 found that : (1) The relative values of 
 
167 
 
 the current, in the sections near the 
 machine and remote from it, became 
 changed from unity, which was, of course, 
 the relation before the introduction of 
 the condenser : (2) There arose a differ- 
 ence in phase in the two portions of the 
 circuit. Both these changes depended 
 only upon : (i) the coefficient of self-in- 
 duction in the remote section only ; (ii) 
 the resistance in the remote section only; 
 
 Fig. 35 
 
 (iii) the capacity of the condenser ; (iv) 
 the period ; and not upon the coefficient 
 of self induction, or the resistance of the 
 section near the generator, or of the 
 generator itself. Under a proper arrange- 
 ment of the four properties quoted, a 
 considerable excess of current could be 
 easily effected in the remote section over 
 that in the nearer section. Mr. Nalder 
 had been kind enough to give him the 
 
168 
 
 opportunity to test the results of the 
 theory in practice, and he supposed they 
 were the first people to see, so far as 
 it was possible by means of a dyna- 
 mometer to do so, a current in one 
 conductor at some distance from the 
 generating machine very much in excess 
 of its value at points near to and in 
 the machine itself. Not merely could 
 such a transformation in relative magni- 
 tude be effected, but the phase rela- 
 tion could be adjusted, thus giving the 
 method very wide applicability. For in- 
 stance, the two portions of the circuit 
 could have their phases thrown into quad- 
 rature with each other ; while the values 
 might be kept equal, which was the condi- 
 tion of things desirable in the Tesla 
 motor, now obtained, as Professor Ayrton 
 had said, by two circuits. 
 
 Captain P. Car dew, R. E., said that in 
 the paper the reasoning was based upon 
 the usual assumption that the varying 
 quantities, magnetic induction, electro- 
 motive force, and current, followed the 
 sinusoidal curve. That was admittedly 
 
169 
 
 on]y an approximation for the purpose 
 of facilitating the calculations. It was 
 probably nearly true in the case of ma- 
 chines with no iron in the generative 
 portion, or armature, in which a copper 
 conductor was dragged through alter- 
 nating magnetic fields ; but, in the case 
 of machines in which the armature wires 
 
 Fig. 36 
 
 were coiled on iron cores, the electro- 
 motive force being due to the variation 
 of induction in those cores, he did not 
 think that law of variation could hold. 
 The maximum induction in the moving 
 iron was not reached until after the point 
 of nearest approach to the fixed magnet 
 had been passed. When once the lines 
 
170 
 
 of force had begun to clear out of the 
 iron, the action would go on rapidly, in 
 spite of the opposing action of the in- 
 duced current, owing to the molecules of 
 iron assisting the demagnetization, the 
 curve of induction being something like 
 the firm line, while the curves of current 
 were like the dotted line in Fig. 36. In 
 the same way, if the moving iron was 
 whirled past alternating polarities, the 
 curves of induction and of current would 
 still be unsymmetrical and unsymmetric- 
 ally disposed with regard to the times of 
 coincidence of the axes of the fixed and 
 moving magnets (Fig. 37). 
 
 Mr. W. M. Mordey was glad the author 
 had adopted his term " Alternator." The 
 old term " alternating- current dynamo- 
 electric machine" was very cumbrous. 
 Alternators might be divided into two 
 classes, those which had iron in the arma- 
 tures, and those which had not. As his 
 own machine, made by the Brush Corpora- 
 tion, and that of Ferranti, were the only 
 examples of the latter class described 
 by the author, he might be permitted 
 
171 
 
 to mention the reasons which had led 
 him to avoid the use of iron. He thought 
 that makers of dynamos would agree that 
 almost all the ills that the dynamo was 
 heir to were due to the presence of the 
 iron in the armature. Its use in direct- 
 current machines might be said to be a 
 
 Fig. 37 
 
 necessary evil, as, although attempts had 
 been made to do without it, all successful 
 types of such machines required it for 
 structural purposes. Iron in armatures 
 when worked at a high magnetic density, 
 and with rapid reversals or variations of 
 magnetisms, became heated, and wasted 
 a good deal of power. In alternators 
 this objection applied with very much 
 
172 
 
 greater force than in direct- current ma- 
 chines, for in the latter the reversals 
 of magnetism were comparatively slow. 
 Thus the first result arrived at, in quite 
 recent practice, was that the magnetic 
 density that could be used in iron cored 
 alternators must only be about one-half 
 that employed in direct current arma- 
 tures. This alone meant a considerable 
 increase in the size of armatures, 
 without any gain in. output or efficiency. 
 Although the loss per cubic inch was 
 reduced by decreasing the magnetic den- 
 sity, the armature had to be made larger 
 to compensate for it, and the total loss 
 was actually increased, not reduced. Iron 
 was used to reduce the magnetic resist- 
 ance, to afford mechanical support, and 
 to introduce self-induction into the cir- 
 cuit. The latter, an evil in itself, was 
 said to be a modern necessity, caused by 
 the convenience of working alternators 
 in parallel. If iron was indispensable 
 for this purpose, which he was not 
 prepared to admit, it could readily be 
 inserted in some part of the circuit where 
 
173 
 
 ample space and cooling surface could be 
 provided, and from which it could be 
 easily removed when not required, that 
 was when it was only necessary to run 
 one machine. The armature was the 
 very worst place for iron. He ventured 
 to think it was much the best to make 
 the armature stationary, as he had done. 
 It then had only to resist the tangential 
 drag of the field. He thought a great 
 mistake was made in some alternators in 
 using Pacinotti projections. In all cases 
 there should be, as nearly as possible, a 
 steady magnetic flux in the field. This 
 could not be done if projections were 
 used. The Zipernowsky and Parsons 
 machines were faulty in this respect. 
 Much had been said about the form of 
 the wave yielded by alternators. He had 
 some time ago made an experiment with 
 the first of his machines, using the 
 method which was now described by 
 Professor Ayrton, and found that the 
 curve had the form shown by Fig. 38, 
 which was almost a sine-curve. On the 
 subject of the lag in transformers, to 
 
174 
 
 which Professor Ayrton had also alluded, 
 he might be allowed to mention that, in 
 the discussion of another paper by the 
 author, he had first stated the fact,* and 
 had described a simple experiment, show- 
 ing that the primary and secondary cur- 
 rents reached their maxima and minima 
 practically at the same time. 
 
 Sir James N. Douglass considered that 
 the discussion had thrown a considerable 
 amount of light on the relative advan- 
 tages of direct and alternate-current ma- 
 chines for lighting and power purposes. 
 In 1862 the French lighthouse authori- 
 ties adopted the alternate-current machine 
 of the Alliance Company at the two light- 
 houses of Cape la Heve; and in 1866, 
 Holmes produced his alternate-current 
 machine, and a pair of these were made 
 for the Trinity House and exhibited at 
 the Paris International Exhibition of 
 1867. These machines were installed at 
 the Souter Point Lighthouse in 1870, 
 where they had worked regularly and 
 
 * Journal of the Institution of Electrical Engineers, 
 vol. xvii., p. 215. 
 
175 
 
 efficiently ever since, and without a single 
 failure or the necessity for any repair. 
 In 1877, the Trinity House made a series 
 of competitive trials at the South Fore- 
 land, with the Holmes and Alliance alter- 
 nating, and the more powerful direct-cur- 
 
 Fig. 38 
 
 MORDEY- VICTORIA ALTERNATOR. OPEN CIRCUIT CURVE 
 OF INDUCTION, HALF PERIOD. THE DOTTED LINE 
 IS A SINE-CURVE. 
 
 rent machines of Gramme and Siemens, 
 which resulted in the adoption of the 
 Siemens machines for the two Lizard 
 lighthouses in 1878. In 1879, the alter- 
 nate-current machine of Baron de Meri- 
 
176 
 
 tens was tested at the Royarinstitution 
 by Dr. Tyndall and himself, and one was 
 purchased by the Trinity House in 1880, 
 and sent to the Lizard lighthouses, where 
 it had been in successful operation every 
 night since that date, giving no more 
 trouble than the original alternate-cur- 
 rent machines of the Alliance Company 
 and Holmes. Three of these machines 
 of larger size, giving 'currents of about 
 160 amperes and 42 volts, were purchased 
 by the Trinity House, and used at the 
 trials, at the South Foreland, in 1884-85, 
 into the relative merits of electricity, gas, 
 and mineral oil as lighthouse illuminants. 
 These machines gave great satisfaction, 
 and two of them had since been perma- 
 nently installed at St. Catharine's light- 
 house. 
 
 Mr. G. Kapp, in reply, said he had 
 first thought that his paper was more 
 scientific than it should have been, but 
 the discussion had beaten him completely 
 in the amount of science which had been 
 developed. Considerable importance 
 seemed to have been attached to the 
 
177 
 
 nomenclature in the paper. Whether a 
 certain electromotive force should be 
 called " mean," " average," " equivalent," 
 the "square root of the mean squares," 
 or by any other name, might be a fit sub- 
 ject for discussion from a pure academic 
 point of view, but practical engineers 
 were satisfied with the terms ordinarily 
 in use, provided definitions were added 
 to prevent misunderstanding. The dia- 
 grams had not been designed with a view 
 of finding out everything which could 
 possibly be ascertained about alternators, 
 but simply as mental tools, perfected 
 only so far as necessary for the work of 
 designing alternators, but not further. 
 The statement in the paper, that there 
 was no self-induction in a bank of trans- 
 formers working glow-lamps, although 
 described as incorrect by Professor Ayr- 
 ton, had in reality been confirmed by the 
 remarks that gentleman had made, and 
 also by what fell from Mr. Mordey. Both 
 had said virtually the same thing, namely, 
 that the lag was nearly 180 between the 
 phases of primary and secondary current. 
 
178 
 
 But the lag between the induced counter 
 electromotive force in the primary and 
 the secondary current was also 180, 
 and therefore the phases of terminal 
 electromotive force, induced electromo- 
 tive force, and current in the primary, 
 must very nearly coincide, which was 
 only possible if there was practically no 
 self induction. He wished to offer an 
 explanation with regard to the units used 
 in the paper. His reason for adopting 
 these units was that they were convenient, 
 and since he first brought them forward 
 many others had adopted them. All 
 knew that the scientifically correct unit 
 was inconvenient for the shop, and as 
 the object of practical engineers was to 
 build machines in the easiest way, he had 
 adopted that system of units which many 
 years' practice had shown him to be the 
 most convenient. Professor Forbes had 
 objected to the parallel working of dyna- 
 mos, chiefly on the ground that it was 
 necessary to cut out certain districts in 
 case of fire. Mains might be joined in 
 parallel at the stations, but they need not 
 
179 
 
 be joined outside where the fire might 
 occur. There was no objection to work- 
 ing all the circuits from a pair of omnibus 
 bars in the station, but they should be 
 disconnected in the districts to be lighted. 
 Another objection urged was the drop of 
 external pressure with an increase of 
 current, provided the exciting current 
 was not varied. That was a serious 
 objection when there was one dynamo 
 machine only at work, and when a large 
 part of the total load would, on certain 
 occasions, be taken off suddenly. For 
 example, in ship-lighting it often hap- 
 pened that the lamps were divided be- 
 tween a few circuits, and a large number 
 of lamps must thus be switched off at 
 one instant. This would, with an alter- 
 nator having a large self-induction, pro- 
 duce in the remaining lamps a great 
 jump. But these machines were not 
 intended for ship-lighting or small iso- 
 lated installations ; they were intended 
 for large central stations where there 
 were many thousand lamps, and where 
 the effect of switching on or off even as 
 
180 
 
 many as one hundred lamps simultane- 
 ously would be infinitesimal in compari- 
 son with the large number of lamps 
 alight. He did not think that a switch 
 that contained mercury troughs which 
 would bob up and down was as good 
 as a switch with solid contacts. The 
 switches used at the Grosvenor Gallery 
 were of the latter type, and in his opinion 
 far superior to the mercury switches 
 which Professor Forbes had praised so 
 much. Professor Forbes was a little too 
 severe on English engineers when he 
 said that they merely copied foreign 
 dynamos. It was quite true that in a 
 certain sense all machines resembled 
 each other, in so far namely as all con- 
 tained field-magnets and an armature ; 
 but this was also the case with continu- 
 ous-current dynamos, of which many 
 different types were nevertheless recog- 
 nized. He did not think high frequency 
 was objectionable on the ground of 
 parallel working, but it was on the 
 grounds which he had mentioned in the 
 paper. He was sorry Professor Forbes 
 
181 
 
 had found fault with the transformers 
 made in England. He appeared to have 
 a low opinion of English engineers since 
 he had seen some of the central stations 
 on the Continent. His contention was 
 that in England electricians were employ- 
 ing too long a magnetic circuit. But 
 all these transformers which Professor 
 Forbes criticised adversely were not 
 made yesterday; they were begun to be 
 made three years ago. Many hundreds 
 were at work, and they had been altered 
 and altered till the present forms had 
 been developed. The magnetic circuit 
 was long, but the electric circuit was 
 short, and a short electric circuit was far 
 more important than a short magnetic 
 circuit. As a matter of fact, even the 
 . Zipernowsky transformer, which Pro- 
 fessor Forbes admired so much, had a 
 long magnetic circuit. Three or four 
 years ago, at the Inventions Exhibition, 
 that wonderful transformer with a spe- 
 cially short magnetic circuit was first 
 seen. It was a Gramme ring turned 
 inside out. It was stated to be a mar- 
 
182 
 
 velous improvement; but not one of 
 those transformers was now at work ; all 
 had been exchanged for the ordinary 
 Gramme ring, with iron inside and cop- 
 per outside, because in this form the 
 electric circuit was short ; consequently, 
 the regulating power of the transformer 
 was better than it had been previously. 
 The other question raised about trans- 
 formers was whether a high or a low 
 frequency was advantageous. A fairly 
 high frequency was advisable ; but above 
 a certain limit, say 70 complete periods 
 per second, an increase of frequency did 
 not reduce the size of the transformer. 
 On the table there was one of his transfor- 
 mers, weighing 200 Ibe., and feeding fifty 
 lamps, giving a weight of 4 Ibs. per lamp. 
 It was built for a frequency of 70. The 
 Westinghouse transformer (taking the 
 figures from a paper by Professor Forbes 
 at the last meeting of the British Associa- 
 tion) was for forty lamps, and it weighed 
 160 Ibs., or 4 Ibs. per lamp. In that 
 case the frequency was 133 as against 
 his 70. That was a proof that the 
 
183 
 
 increase of frequency did not diminish 
 the weight of the transformers. Pro- 
 fessor Forbes had brought a diagram 
 showing the output from the Grosvenor 
 Station (Fig. 28), and he argued that the 
 transformers by themselves absorbed an 
 energy equal to four thousand lamps. 
 That was surely not to be taken as a 
 serious argument. The diagram was not 
 one of HP., but one of current sent out. 
 Now the current sent out of course 
 depended on the lamps which were 
 alight. The argument was that in the 
 early hours of the morning there were 
 no lamps on circuit. But that was not 
 so certain. In the case of stations sup- 
 plying electric light by contract, for a. 
 fixed rental without a meter, people were 
 wasteful with the lamps and allowed them 
 to burn all night long. Professor Forbes 
 had overlooked one important circum- 
 stance, namely, that if the transformers 
 were on the circuit without giving off 
 current to the lamps, there was a lag of 
 70, 80 or more between the terminal 
 electromotive force and the primary cur- 
 
184 
 
 rent in each transformer. The energy 
 must therefore not be taken as repre- 
 sented by the product of electromotive 
 force and current, as given in the diagram, 
 but by this product multiplied by the 
 cosine of the angle of lag. On purely 
 theoretical grounds all the transformers 
 now in use were equally good and their 
 efficiency was high. Professor Ayrton 
 had tested some, and found that it was 
 as high as 95 per cent. The difference 
 in the transformers was in their details, 
 such as would secure low first cost, 
 facility of manufacture, ease of changing 
 a coil, freedom from humming, ability to 
 keep cool, and, above all things, absolute 
 safety of insulation. 
 
 COEKESPONDENCE. 
 
 Mr. J. D. F. Andrews, in reference to 
 the author's observations on eddy cur- 
 rents (p. 64), and his inability "to 
 suggest an entirely satisfactory explana- 
 tion for the effect of the iron core in 
 
185 . 
 
 reducing eddy currents in the copper," 
 stated that he had contributed the results 
 of a series of experiments in a letter to 
 Industries,* in which he showed that the 
 strength or amount of eddy currents 
 depended on the speed with which the 
 wire moved into and out of the magnetic 
 field, and that they could be greatly 
 reduced by shaping the magnet-poles so 
 that the wire gradually entered and left 
 the field. The shape of the conductor, 
 and its position relative to the armature 
 core and the field-magnets, had a greater 
 bearing on the reduction of eddy currents 
 than the shape of the field- magnets ; and 
 round wire of any diameter was nearly 
 proof against eddy currents if the 
 machine was otherwise correctly propor- 
 tioned. 
 
 Mr. A. Du Bois-Keymond remarked 
 that some minor points in the diagrams 
 seemed open to criticism. With regard 
 to Figs. 9 and 10, he could not fully 
 comprehend the author's meaning when 
 he said, u the length O A, that is to say, 
 
 * Vol. iv. 1888, p. M 9. 
 
186 
 
 the current," and accordingly defined the 
 mechanical output of the machine as 
 " the product O A and O D." As far as 
 he could make out, O A was the resultant 
 electromotive force obtained by the 
 .addition of the terminal and induced 
 electromotive forces E t and E, and by the 
 author's own showing (Fig. 7) the current 
 ought to be the resultant of this and of 
 the electromotive force due to self-indue" 
 tion, that was, it ought to coincide with 
 O K ; or assuming the terminal resistance 
 to be one, it would be equal to O K. 
 Accordingly, the mechanical work gener- 
 ated would be the product of O E and O 
 D, namely, E. I. cos B O F, I standing 
 for the current. Moreover, he should 
 prefer assuming some definite position of 
 E t to start from as being more conforma- 
 ble to reality, aud letting E shift about, 
 thereby altering the positions and magni- 
 tudes of the other quantities. There 
 would then be perfect agreement between 
 the author's diagram and the geometrical 
 solution of the same problem explained 
 by him in a paper, " On the difficulties in 
 
187 
 
 the way of transmitting power by alter- 
 nating currents."* The use to which the 
 author put the diagrams, namely, to 
 determine the safety of running a given 
 machine in parallel with others, was 
 admirable, and quite put an end to the 
 doubts hitherto entertained by practi- 
 tioners on this head. 
 
 Professor J. A. Ewing said that at p. 
 65 of the paper there was a remark 
 about the dissipation of energy in the 
 armatures of dynamos through magnetic 
 hysteresis, from which it seemed to 
 him that the author had perhaps some- 
 what misunderstood in one respect the 
 position he had taken up in the paper 
 there referred to. What he had shown 
 was that the energy dissipated in iron, 
 through magnetic hysteresis in reversals 
 of magnetism, might be very much 
 reduced if the metal were subjected to 
 vibration during the process. When a 
 soft iron wire, for example, was sharply 
 
 * The Telegraphic Journal and Electrical Review 
 vol. xxiv. 1889, p. 112; and Elektrotechnische Zeitschrift 
 vol x. 1889, p. 1. 
 
188 
 
 tapped during repeated reversals of its 
 magnetism, there was practically no dissi- 
 pation of energy through hysteresis, and 
 even a very slight mechanical disturbance 
 had some influence in reducing the dissi- 
 pation. In applying this result to 
 dynamos, all t that he had said was this : 
 " Hence, in a dynamo, where vibration 
 occurs to a greater or less degree when- 
 ever the machine is running, the energy 
 dissipated through changes of magnetiza- 
 tion is even less than these experiments 
 on still metal would lead us to expect."* 
 He had meant, in these words, to draw 
 attention to the fact that whatever vibra- 
 tion occurred in the armature of a 
 dynamo would tend to lessen the dissipa- 
 tion of energy, produced by hysteresis? 
 below the value measured by his experi- 
 ments on still metal. He had not meant 
 to suggest that any large effect of this 
 kind was actually produced, and in fact 
 he agreed with the author in believing 
 that in a well-balanced dynamo the vibra- 
 tion was not likely to be seriously influen- 
 
 * Phil. Trans. Royal Society, 1885, p. 554. 
 
189 
 
 tial, so that the values of the dissipation 
 calculated from his experiments were 
 probably not too high for application to 
 such a case. With reference to " viscous 
 hysteresis " (p. 67), he had made no 
 direct observations showing that the 
 energy dissipated through hysteresis 
 increased - with the speed at which the 
 cyclic change of induction was performed. 
 So far as any such observations had 
 been made,, he understood that the 
 results had been negative, and that the 
 heating effect per cycle had been sensibly 
 the same whether the same cycle was 
 repeated with greater or with less 
 frequency. But there was a certain 
 amount of evidence that magnetization 
 took some time to be fully developed in 
 iron after the magnetizing force was 
 applied, especially when the magnetizing 
 force was low, and so far as this was true 
 it must have the effect of increasing the 
 dissipation of energy in cycles of high 
 frequency, since larger changes of mag- 
 netic force would then be required to 
 produce equal changes of magnetic 
 induction. 
 
190 
 
 Dr. John Hopkinson said that one of 
 the questions with which the author 
 dealt was how many alternations per 
 second was it most appropriate to use for 
 distribution by transformers'? It was 
 clear that the answer depended upon a 
 balance of advantages. No one would 
 use exceedingly few alternations ; no one 
 would advocate a greater number than 
 some superior limit much less than the 
 greatest it was possible to produce. In 
 favor of a high frequency was the fact 
 that, for a given efficiency, the trans- 
 former would be cheaper to manufacture. 
 On the other hand, in favor of a low 
 frequency, the whole of the conductor 
 was not equally used with alternating 
 currents. The author had alluded to the 
 question of a viscous hysteresis in mag- 
 netization of iron. Dr. Hopkinson 
 doubted the propriety of the term vis- 
 cous hysteresis, but he knew what was 
 meant. There was, he believed, no satis- 
 factory evidence of the existence of so- 
 callecl viscous hysteresis, and it was 
 certainly a phenomenon by no means so 
 
191 
 
 marked as true hysteresis when the 
 magnetism of iron was reversed. In his 
 lecture before the Institution in 1883, he 
 not only explained why alternate- current 
 machines could be run parallel but also 
 mentioned that an alternate-current 
 machine could be run as a motor. The 
 theory of both cases had been fully 
 developed by himself in a paper read 
 before the Society of Telegraph En- 
 gineers ;* and, at his suggestion, Pro- 
 fessor Adams successfully tried the 
 experiment of running a machine as a 
 motor at the South Foreland. In 1887 
 he presented to the Royal Society two 
 short papers. f The first related to trans- 
 formers, and showed how to treat them, 
 taking proper account of the true mag- 
 netic properties of the iron. The second 
 treated the theory of alternate-current 
 machines ; showed how the true differ- 
 ential equation of currents could be 
 
 * Journal of the Society of Telegraph Engineers 
 and Electricians, vol. xiii. 1884, p. 496. 
 
 t Proceedings of the Royal Society, vol. xlii. H-. 
 164 and 167. 
 
192 
 
 obtained, and that it differed materially, 
 in some cases, from the linear equation 
 generally used, but was of a character 
 not convenient for practical use. 
 
 Mr. Elihu Thomson, in regard to the 
 relation between the width of the field- 
 poles and the armature-winding, observed 
 that the author's statement that " width 
 of poles equal to half the pitch, smooth 
 armature, and winding covering only one- 
 half the surface," was most frequently 
 met with in practice, was without doubt 
 true. But was it the best practice ? 
 While designing such machinery, it oc- 
 curred to him that a higher yield in 
 electromotive force would be had from a 
 given armature, particularly under full 
 load, or "dynamic" working, as dis 
 tinguished from open circuit, or " static " 
 condition, if the open space in the 
 interior of each coil was made less than 
 the width of the pole-face. It was seen 
 that this must result as a consequence of 
 the crowding or distortion of the field- 
 lines forward by the currents in the 
 armature-coils, having the practical effect 
 
193 
 
 of somewhat narrowing the effective pole- 
 face presented to the armature. A trial 
 machine was built, on the type in use by 
 the Thomson-Houston Electric Company, 
 which consisted of a field structure with 
 inward radial cores and a cylindrical 
 armature with very flat coils laid on and 
 bound down, their curved ends out of 
 the field; not, however, being turned 
 down toward the shaft, as is the case 
 
 Fig. 39 
 
 with the Westinghouse machine. The 
 field-poles had a width approximately 
 equal to one half the pitch, the armature 
 being smooth. In the trial machine the 
 coils were laid on so as to practically cover 
 the whole exterior of the armature- core, or 
 they were wound with but a narrow line 
 of open space in the center of each coil, 
 
194 
 
 as in Fig. 39. Arrangements were made 
 that the inner turns might be progress- 
 ively cut out, so as to virtually remove 
 wire from each coil. Under static condi- 
 tion, or no load, the terminal electromo- 
 tive force was practically unchanged, 
 even though considerable fractions of 
 wire-turns were cut out from the center 
 of each coil. But under a load a limit 
 was soon found, or a point reached, in 
 which the electromotive force was the 
 maximum under a given excitation of the 
 field. This maximum was reached after 
 a few turns had been removed or cut 
 out ; and the open space in the center of 
 the coil so obtained was adopted for use 
 in the apparatus subsequently manufact- 
 ured on the large scale. The vacant 
 space was less than one-half the width of 
 the pole-face. The relations would be 
 as shown in Fig. 40. He regarded this 
 result as in a measure growing out of the 
 condition of displacement of lines, indr 
 cated in Fig. 41, where the current in 
 the armature -coil drew the lines together, 
 leaving the field, or bunched them 
 
195 
 
 through its own center in the position 
 indicated. In dealing with the heating 
 of armature conductors by eddy-currents, 
 the author had drawn a distinction be- 
 tween coils with and without iron cores, 
 showing the need of an explanation of 
 the differences noted. His own view of 
 the matter had for some time been this : 
 It is well known that, in a transformer 
 with a well-closed iron circuit, very large 
 sections of conductors might be employed 
 
 Fig. 40 
 
 Fig. 41 
 
 in the coils without introducing trouble 
 from eddy-currents. The extreme case 
 occurred in transformers for electric 
 welding, in which the section of the 
 single secondary turn might be many 
 square inches without showing the effects 
 of eddy-currents. The reason, as it 
 appeared to him, for such a condition 
 
196 
 
 was that the development of lines, taking 
 place as circular magnetism around the 
 primary conductors, was carried out with 
 great speed to the lodgment of the lines 
 in the iron shell or core. Thus all parts 
 of the secondary conductor were cut at 
 almost the same instant, by each line 
 evolved, in its passage to the iron core, 
 the same actions being repeated at each 
 alternation. The great speed at which 
 the lines passed the conductor in such 
 cases, in going to the iron core, left no 
 time during which differences of electro- 
 motive force in parts of the conductor 
 might act to induce eddy-currents. Now, 
 when a coil without an iron core passed 
 through a field, the speed of cutting did 
 not exceed the speed of motion, so that 
 if the lines being cut by a conductor 
 were denser at one point than at another, 
 eddy- currents resulted. Provide the 
 coil with an iron core and the lines re- 
 main in the core until dragged out, as it 
 were, by the movement given to it in 
 front of the field-cores, at which time the 
 lines move through the space between 
 
197 
 
 the successive cores with great speed, 
 and cut all parts of the conductor at the 
 
 same instant. The lines snap out of the 
 leaving core, as it were, and pass quickly 
 
198 
 
 to the incoming core, producing a \veak 
 field between the two composed of few 
 lines moving at great speeds. Figs. 42 
 and 43 represented the two conditions 
 comparatively. Fig. 42 showed a coil 
 without core cutting lines between N and 
 S ; and Fig. 43 the progressive move- 
 ment of coils with iron cores through 
 the field. The rapid movement of the 
 lines backward across the space between 
 the iron armature-cores, or projections, 
 II, Fig. 43, as the armature moved for- 
 ward, pointed also to the necessity for 
 laminatioi} of the field-poles when such 
 cores or projections were used. The 
 turns of armature-conductor nearest the 
 enclosed iron core would be subject to 
 eddy-currents, on account of the field 
 density being made more irregular by 
 the bending of the lines leaving the iron. 
 He was of opinion that the use of " horns '' 
 on pole-pieces was not akin to the effects 
 just mentioned, excepting in so far as the 
 entering and leaving field was more 
 spread out or graduated in density, and 
 so placed the parts of a large section 
 
199 
 
 armature- conductor in substantially equal 
 field densities at any instant during 
 movement into or out of the field. This 
 was not the case where no horns were 
 used, or where the field was abrupt and 
 permitted, say, one-half the section of 
 the armature-conductor to be at any in- 
 stant cutting dense lines, and the other 
 half to be moving in a field of little density. 
 Making the conductor of insulated strands, 
 and twisting the strands, was the remedy 
 indicated. 
 
* Any book in this Catalogue sent free by mail on receipt of price. 
 
 VALUABLE 
 
 SCIENTIFIC BOOKS 
 
 PUBLISHED BY 
 
 D. VAN NOSTRAND COMPANY, 
 
 23 MURRAY STREET AND 27 WARREN STREET, N. V. 
 
 ADAMS (J. W.) Sewers and Drains for Populous Districts. 
 Embracing Rules and Formulas lor the dimensions and 
 construction oi works of Sanitary Engineers. Second edi- 
 tion. 8vo, cloth $2 50 
 
 ALEXANDER (J. H.) Universal Dictionary of Weights and 
 Measures, Ancient and Modern, reduced to the standards 
 of the United States of America. New edition, enlarged. 
 8vo, cloth 3 50 
 
 ATWOOD (GEO.) Practical Blow-Pipe Assaying, izmo, cloth, 
 
 illustrated 2 oo 
 
 AUCHINCLOSS (W. S.) Link and Valve Motions Simplified. 
 Illustrated with 37 wood-cuts and 21 lithographic plates, 
 together with a Travel Scale and numerous useful tables. 
 8vo, cloth 3 OO 
 
 AXON (W. E. A.) The Mechanic's Friend : a Collection of Re- 
 ceipts and Practical Suggestions Relating to Aquaria 
 Bronzing Cements Drawing Dyes Electricity Gilding 
 Glass-working Glues Horology Lacquers Locomo- 
 tives Magnetism Metal-working - Modelling Photogra- 
 phy Pyrotechny Railways Solders Steam-Engine Tel- 
 egraphy Taxidermy Varnishes Waterproofing, and Mis- 
 cellaneous Tools, Instruments, Machines, and Processes 
 connected with the Chemical and Mechanic Arts. With nu- 
 merous diagrams and wood-cuts. Fancy cloth ..150 
 
 BACON (F. W.) A Treatise on the Richards Steam-Engine 
 Indicator, with directions for its use. By Charles T. Por- 
 ter. Revised, with notes and large additions as developed 
 by American practice ; with an appendix containing useful 
 formulae and rules for engineers. Illustrated. Third edi- 
 tion. i2mo, cloth i oo 
 
2 D. VAN NOSTRAND S PUBLICATIONS. 
 
 BARBA (J.) The Use of Steel for Constructive Purposes ; 
 Method of Working, Applying, and Testing Plates and 
 Brass. With a Preface by A. L. Holley, C.E. 12010, cloth.fti 50 
 
 BARNES (Lt. Com. J. S., U. S." N.) Submarine Warfare, offen- 
 sive and defensive, including a discussioa of the offensive 
 Torpedo System, its effects upon Iron-Clad Ship Systems 
 and influence upon future naval wars. With twenty litho- 
 graphic plates and many wood-cuts. 8vo, cloth ............ 5 oo 
 
 BEILSTEIN (F.) An Introduction to Qualitative Chemical 
 
 Analysis, translated by I. J. Osbun. i2mo, cloth .......... 75 
 
 BENET (Gen. S. V., U. S. A.) Electro-Ballistic Machines, and 
 the Schultz Chronoscope. Illustrated. Second edition, 410, 
 cloth ...................................................... 3 oo 
 
 BLAKE (W. P.) Report upon the Precious Metals ; Being Sta- 
 tistical Notices of the principal Gold and Silver producing 
 regions of the World, represented at the Paris Universal 
 Exposition. 8vo, cloth ..................................... 2 oo 
 
 - Ceramic Art. A Report on Pottery, Porcelain, Tiles, 
 
 Terra Cotta, and Brick. 8vo, cloth ........................ 200 
 
 BOW (R. H.) A Treatise on Bracing, with its application to 
 Bridges and other Structures of Wood or Iron. 156 illustra- 
 tions. 8vo, cloth ..... ............... .................... i 50 
 
 BOWSER (Prof. E. A.) An Elementary Treatise on Analytic 
 Geometry, embracing Plane Geometry, and an Introduc- 
 tion to Geometry of three Dimensions. i2mo, cloth ....... i 75 
 
 An Elementary Treatise on the Differential and Integral 
 
 Calculus. With numerous examples. i2mo, cloth ......... 225 
 
 BURGH (N, P.) Modern Marine Engineering, applied to Pad- 
 dle and Screw Propulsion. Consisting of 36 colored plates, 
 259 practical wood-cut illustrations, and 403 pages of de- 
 
 , 
 
 scriptive matter, the whole being an exposition of the pre- 
 sent practice of Tames Watt & Co., J. & G. Rennie, R. Na- 
 pier & Sons, and other celebrated firms. Thick 410 vol., 
 
 , . . 
 
 cloth .............. ..................................... 10 oo 
 
 Half morocco .............................................. 15 oo 
 
 BUR T (W. A.) Key to the Solar Compass, and Surveyor's Com- 
 panion ; comprising all the rules necessary for use in the 
 field: also description of the Linear Surveys and Public 
 Land Astern of the United States, Notes on the Barome- 
 ter, .t.gestions for an outfit for a survey of four months, 
 etc. Second edition. Pocket-book form, tuck ............. 250 
 
 BUTLER (Capt. J. S., U. S. A.) Projectiles and Rifled Cannon. 
 A Critical Discussion of the Principal Systems of Rifling 
 and Projectiles, with Practical Suggestions for their Im- 
 provement, as embraced in a Report to the Chief of Ord- 
 nance, U. S. A. 36 plates 410, cloth ....................... 6 00 
 
D. VAN NOSTRAXD S PUBLICATIONS. 3 
 
 CAIN (Prof. WM ) A Practical Treatise on Voussoir and Solid 
 
 and Braced Arches. i6mo, cloth extra. $i 75 
 
 CALDWELL (Prof. GEO. C.) and BRENEM AN (Prof. A. A.) 
 Manual of Introductory Chemical Practice, for the use of 
 Students in Colleges and Normal and High Schools. Third 
 edition, revised and corrected. 8vo, cloth, illustrated. New 
 and enlarged edition. I 50 
 
 CAMPIN (FRANCIS). Onthe Construction oflron Roofs. 8vo, 
 
 wit h plates, cloth 2 oo 
 
 CHAUVENET (Prof. W.) New method of correcting Lunar 
 Distances, and improved method of finding the error and 
 rate of a chronometer, by equal altitudes. 8vo, cloth 2 oo 
 
 CHURCH (JOHN A.) Notes of a Metallurgical Journey in 
 
 Europe. 8vo, cloth 200 
 
 CLARK (D. KINNEAR, C.E.) Fuel: Its Combustion and 
 Economy, consisting of Abridgments of Treatise on the 
 Combustion of Coal and the Prevention of Smoke, by C. 
 W. Williams ; and the Economy of Fuel, by T S Pri- 
 deaux. With extensive additions on recent practice in the 
 Combustion and Economy of Fuel : Coal, Coke, Wood, 
 Peat, Petroleum, etc. i2mo, cloth I 50 
 
 A Manual of Rules, Tables, and Data for Mechanical 
 
 Engineers. Based on the most recent investigations. Illus- 
 trated with numerous diagrams. 1,012 pages. 8vo, cloth... 7 50 
 Half morocco 10 oo 
 
 CLARK (Lt. LEWIS, U. S. N ) Theoretical Navigation and 
 Nautical Astronomy. Illustrated with 41 wood-cuts. 8vo, 
 cloth I 50 
 
 CLARKE (T. C) Description of the Iron Railway Bridge over 
 the Mississippi River at Quincy, Illinois. Illustrated with 
 21 lithographed plans. 410, cloth 7 50 
 
 CLEVENGER (S. R.) A Treatise on the Method of Govern- 
 ment Surve\ing, as prescribed by the U. S. Congress and 
 Commissioner of the General Land Office, with complete 
 Mathematical, Astronomical, and Practical Instructions for 
 the use of the United States Surveyors in the field. i6mo, 
 morocco 2 y> 
 
 COFFIN (Prof J. H. C.) Navigation and Nautical Astrono- 
 my. Prepared for the use of the U. S. Naval Academy. 
 Sixth edition. 52 wood-cut illustrations. I2mo, cloth 3 50 
 
 COLBURN (ZERAH). The Gas-Works of London. i2mo, 
 
 boards 60 
 
 COLLINS (JAS. E.) The Private Book of Useful Alloys and 
 
 Memoranda for Goldsmiths, Jewellers, etc. i8mo, cloth. .. 50 
 
4 D. VAN NOSTRAND S PUBLICATIONS. 
 
 CORNWALL (Prof. H. B.) Manual of Blow Pipe Analysis, 
 Qualitative and Quantitative, with a Complete System of 
 Descriptive Mineralogy. 8vo, cloth, with many illustra- 
 tions. N. Y., 1882 $250 
 
 CRAIG (B. F) Weights and Measures. An account of the 
 Decimal System, with Tables of Conversion for Commer- 
 cial and Scientific Uses. Square 32010, limp cloth 50 
 
 CRAIG (Prof. THOS.) Elements of the Mathematical Theory 
 
 of Fluid Motion. i6mo, cloth 125 
 
 DAVIS (C. B.) and RAE (F. B.) Hand-Book of Electrical Dia- 
 grams and Connections. Illustrated with 32 full-page illus- 
 trations. Second edition. Oblong 8vo, cloth extra 2 oo 
 
 D1EDRICH (JOHN). The Theory of Strains : a Compendium 
 tor the Calculation and Construction of Bridges, Roofs, and 
 Cranes. Illustrated by numerous plates and diagrams. 
 8vo, cloth 5 oo 
 
 DIXON (D. B.) The Machinist's and Steam-Engineer's Prac- 
 tical Calculator. A Compilation of useful Rules, and Prob- 
 lems Arithmetically Solved, together with General Informa- 
 tion applicable to Shop-Tools, Mill-Gearing, Pulleys and 
 Shafts, Steam-Boilers and Engines. Embracing Valuable 
 Tables, and Instruction in Screw-cutting, Valve and Link 
 Motion, etc. i6mo, full morocco, pocket form . ..(In press) 
 
 DODD (GEO.) Dictionary of Manufactures, Mining, Ma- 
 chinery, and the Industrial Arts. I2mo, cloth I 50 
 
 DOUGLASS (Prof S. H.) and PRE3COTT (Prof. A. B.) Qual- 
 itative Chemical Analysis. A Guide in the Practical Study 
 of Chemistry, and in the Work of Analysis. Third edition. 
 8vo, cloth 3 50 
 
 DUBOIS (A. J.) The New Method of Graphical Statics. With 
 
 60 illustrations. 8vo, cloth i 50 
 
 EASSIE (P. B.) Wood and its Uses. A Hand-Book for the use 
 of Contractors, Builders, Architects, Engineers, and Tim- 
 ber Merchants. Upwards of 250 illustrations. 8vo, cloth, i 50 
 
 EDDY (Prof. H. T.) Researches in Graphical Statics, errbiac- 
 ing New Constructions in Graphical Statics, a New General 
 Method in Graphical Statics, and the Theory of Internal 
 Stress in Graphical Statics. 8vo, cloth ^ i c>o 
 
 ELIOT (Prof. C. W.) and STORER (Prof. F. H.) A Compen- 
 dious Manual of Qualitative Chemical Analysis. Revised 
 with the co-operation of the authors. By Prof. William R. 
 Nichols. Illustrated. i2mo, cloth i 50 
 
 ELLIOT (Maj. GEO. H., U. S. R.) European Light-House 
 Systems. Being a Report of a Tour of Inspection made in 
 1873. 51 engravings and 21 wood-cuts. 8vo, cloth 5 oo 
 
D. VAN NOSTRAND S PUBLICATIONS. O 
 
 ENGINEERING FACTS AND FIGURES. An Annual 
 Register of Progress in Mechanical Engineering and Con- 
 struction for the years 1863-6.1-65-66-67-68. Fully illus- 
 trated. 6 vols. i8mo, cloth (each volume sold separatel>), 
 per vol $2 50 
 
 FANNING (J. T.) A Practical Treatise of Water-Supply En- 
 gineering. Relating to the Hydrology, Hydro dynamics, and 
 Practical Construction oi Water-Works in North America. 
 Third edition. With numerous tables and 180 illustra- 
 tions, 65opages. 8vo, cloth 500 
 
 FISkE (BRADLEY A., U.S. N.) Electricity in Theory and 
 
 Practice. 8vo, cloth 2 ;o 
 
 FOSTER (Gen. J. G., U. S. A.) Submarine Blasting in Boston 
 Harbor, Massachusetts. Removal ot Tower and Ccrwin 
 Rocks. Illustrated with seven plates. 410, cloth 3 50 
 
 FOYE (Prof. J. C.) Chemical Problems. With brief State- 
 ments of the Principles involved. Second edition, revised 
 and enlarged. i6mo, boards 50 
 
 FRANCIS (JAS. B., C E.) Lowell Hydraulic Experiments : 
 Being a selection from Experiments on Hydraulic Motors, 
 on the Flow of Water over Weirs, in Open Canals of Uni- 
 form Rectangular Section, and through submerged Orifices 
 and diverging Tubes. Made at Lowell, Massachusetts. 
 Fourth edition, revised and enlarged, with many new ex- 
 periments, and illustrated with twenty-three copperplate 
 engravings. 410, cloth , 15 oo 
 
 FREE-HAND DRAWING. A Guide to Ornamental Figure 
 and Landscape Drawing. By an Art Studen,. i8mo, 
 boards 50 
 
 GILLMORE (Gen. Q. A.) Treatise on Limes, Hydraulic Ce- 
 ments, and Mortars. Papers on Practical Engineering, U. 
 S. Engineer Department, No. 9, containing Reports or 
 numerous Experiments conducced in New York City during 
 the years 1858 to 1861, inclusive. With numerous illustra- 
 tions. 8vo, cloth 40O 
 
 Practical Treatise on the Construction of Roads, Streets, 
 
 and Pavements. With 70 illustrations. I2mo, cloth 2 oo 
 
 Report on Strength of the Building Stones in the United 
 
 States, etc. 8vo, illustrated, cloth . 150 
 
 Coignet Beton and other Artificial Stone. 9 plates, views, 
 
 etc. 8vo, cloth 2 50 
 
 GOODEVE (T. M.) A Text-Book on the Steam-Engine. 143 
 
 illustrations. I2mo, cloth ? oo 
 
 GORDON (J. E. H.) Four Lectures on Static Induction. lamo, 
 
 cloth 80 
 
6 D. VAN NOSTRAND'S PUBLICATIONS. 
 
 GRUNER (M. L.) The Manufacture of Steel. Translated 
 from the French, by Lenox Smith, with an appendix on the 
 Bessemer process in the United States, by the translator. 
 Illustrated. 8vo, cloth $3 50 
 
 HALF-HOURS WITH MODERN SCIENTISTS. Lectures 
 and Essays. By Professors Huxley, Barker, Stirling, Cope, 
 Tyndall, Wallace, Rpscoe, Hoggins, Lockyer, Young, 
 Mayer, and Reed. Being the University Series bound up. 
 With a general introduction by Noah Porter, President of 
 Yale College. 2 vols 12010, cloth, illustrated 2 50 
 
 HAMILTON (W. G.) Useful Information for Railway Men. 
 Sixth edition, revised and enlarged 562 pages, pocket form. 
 Morocco, gilt 200 
 
 HARRISON (W. B.) The Mechanic's Tool Book, with Prac- 
 tical Rules and Suggestions for Use of Machinists, Iron- 
 Workers, and others. Illustrated with 44 engravings. 
 I2mo, cloth i 50 
 
 HASKINS (C. H.) The Galvanometer and its Uses. A Man- 
 ual for Electricians and Students. Second edition. I2mo, 
 morocco . . i 50 
 
 HENRICI (OLAUS). Skeleton Structures, especially in their 
 application to the Buildingof Steel and Iron Bridges. With 
 folding plates and diagrams. 8vo, cloth I 50 
 
 HEWSON (WM.) Principles and Practice of Embanking 
 Lands from River Floods, as applied to the Levees of the 
 Mississippi. 8vo, cloth 2 oo 
 
 HOLLEY (ALEX. L.) A Treatiseon Ordnance and Armor, em- 
 bracing descriptions, discussions, and professional opinions 
 concerning the materials, fabrication, requirements, capa- 
 bilities, and endurance of European and American Guns, 
 for Naval, Sea-Coast, and Iron-Clad Warfare, and their 
 Rifling, Projectiles, and Breech-Loading; also, results of 
 experiments against armor, from official records, with an 
 appendix referring to Gun-Cotton, Hooped Guns, etc., etc. 
 948 pages, 493 engravings, and 147 Tables of Results, etc. 
 8vo, half roan 10 oo 
 
 Railway Practice American and European Railway 
 
 Practice in the economical Generation of Steam, including 
 the Materials and Construction of Coal-burning Boilers, 
 Combustion, the "Variable Blast, Vaporization, Circulation, 
 Superheating, Supplying and Heating Feed-water, etc., 
 and the Adaptation of Wood and Coke-burning Engines to 
 Coal-burning; and in Permanent Way, including Road-bed, 
 Sleepers, Rails, Joint-fastenings, Street Railways, etc., etc. 
 With 77 lithographed plates. Folio, cloth 12 o 
 
 HOWARD (C. R.) Earthwork Mensuration' on the Basis of 
 the Prismoidal Formulae. Containing simple and labor- 
 saving method of obtaining Prismoidal Contents directly 
 
D. VAN NOSTRAND S PUBLICATIONS. 7 
 
 from End Areas. Illustrated by Examples, and accom- 
 panied by Plain Rules for Practical Uses. Illustrated. 8vo, 
 cloth $l 50 
 
 INDUCTION-COILS. How Made and How Used. 63 illus- 
 trations. i6mo, boards 50 
 
 ISHERWOOD (B. F.) Engineering Precedents for Steam Ma- 
 chinery. Arranged in the most practical and useful manner 
 for Engineers. With illustrations. Two volumes in one. 
 8vo, cloLh 250 
 
 JANNETTAZ (EDWARD). A Guide to the Determination of 
 Rocks: being an Introduction to Lithology. Translated 
 from the French by G. W. Plympton, Professor of Physical 
 Science at Brooklyn Polytechnic Institute. I2mo, cloth i 50 
 
 JEFFERS (Capt. W. N., U. S. N.) Nautical Surveying. Illus- 
 trated with 9 copperplates and 31 wood-cut illustrations. 
 8vo, cloth 5 oo 
 
 JONES (H. CHAPMAN). Text-Book of Experimental Or- 
 ganic Chemistry for Students. i8mo, cloth I oo 
 
 JOYNSON (F. H.) The Metals used in Construction: Iron, 
 
 Steel, Bessemer Metal, etc., etc. Illustrated. i2mo, cloth. 75 
 
 Designing and Construction of Machine Gearing. Illus- 
 trated 8vo, cloth , , 2 oo 
 
 KANSAS CITY BRIDGE (THE). With an account of the 
 Regimen of the Missouri River, and a description of the 
 methods used for Founding in that River. By O. Chanute, 
 Chief-Engineer, and George Morrison, Assistant-Engineer, 
 fllustrated with five lithographic views and twelve plates of 
 plans. 4to,cloth... 600 
 
 KING (W. H.) Lessons and Practical Notes on Steam, the 
 Steam-Engine, Propellers, etc., etc, for young Marine En- 
 gineers, Students, and others. Revised "by Chief-Engineer 
 J. W. King, U. S. Navy. Nineteenth edition, enlarged. 
 8vo, cloth 2 oo 
 
 KIRKWOOD (JAS. P.) Report on the Filtration of River 
 Waters for the supply of Cities, as practised in Europe, 
 made to the Board of Water Commissioners of the City of 
 St. Louis. Illustrated by 30 double-plate engravings. 410, 
 cloth 1500 
 
 LARRABEE (C. S.) Cipher and Secret Letter and Telegra- 
 phic Code, with Hogg's Improvements. The most perfect 
 secret code ever invented or discovered. Impossible to read 
 without the key. i8mo, cloth I oo 
 
 LOCK (C. G.), WIGNER (G. W.), and HARLAND (R. H.) 
 Sugar Growing and Refining. Treatise on the Culture of 
 Sugar-Yielding Plants, and the Manufacture and Refining of 
 Cane, Beet, and other sugars. 8vo, cloth, illustrated 12 oo 
 
D. VAN NOSTRAND'S PUBLICATIONS. 
 
 LOCKWOOD (THOS. D.) Electricity, Magnetism, and Elec- 
 tro-Telegraphy. A Practical Guide for Students, Operators, 
 and Inspectors. 8vo, cloth .................... ....... g 2 e;o 
 
 LORING (A. E.) A Hand-Book on the Electro-Magnetic Tele- 
 
 graph. Paper boards ..................................... 50 
 
 Cloth .................. ; ............................. ..... 75 
 
 Morocco ................................................... I oo 
 
 MACCORD (Prof. C. W ) A Practical Treatise on the Slide. 
 Valve by Eccentrics, examining by methods the action of 
 the Eccentric upon the Slide-Valve, and explaining the prac- 
 tical processes of laying out the movements, adapting: th 
 valve for its various duties in the steam-engine. Second edi- 
 tion Illustrated. 4to, cloth .............................. 250 
 
 McCULLOCH (Prof. R S.) Elementary Treatise on the Me- 
 chanical Theory of Heat, and its application to Air and 
 Steam Engines. 8vo, cloth ................................. 3 50 
 
 MERRILL (Col. WM. E , U. S. A.) Iron Truss Bridges for 
 Railroads. The method of calculating strains in Trusses, 
 with a careful comparison of the most prominent Trusses, in 
 reference to economy in combination, etc., etc. Illustrated. 
 4to, cloth ........... ................................... 500 
 
 MICHAELIS (Capt. O. E., U. S. A.) The Le Boulenge 
 Chronograph, with three lithograph folding plates of illus- 
 trations. 410, cloth ....................................... 300 
 
 MICHIE (Prof. P. S.) Elements ot Wave Motion relating to 
 Sound and Light. Text-Book forthe U.S. Military Acade- 
 my. 8vo, cloth, illustrated ................................. 5 oo 
 
 MINIFIE (WM.) Mechanical Drawing. A Text-Book of Geo- 
 metrical Drawing for the use of Mechanics and Schools, in 
 which the Definitions and Rules of Geometry are familiarly 
 explained ; the Practical Problems are arranged, from the 
 most simple to the more complex, and in their description 
 technicalities are avoided as much as possible. With illus- 
 trations for Drawing; Plans, Sections, and Elevations of 
 Railways and Machinery; an Introduction to Isometrical 
 Drawing, and an Essayon Linear Perspective and Shadows. 
 Illustrated with over 200 diagrams engraved on steel. Ninth 
 edition. With an Appendix on the Theory and Application 
 
 of Colors. 8vo, cloth 4 oo 
 
 "It is the best work on Drawing that we have ever seen, and is 
 especially a text -book of Geometrical Drawing for the use of Me- 
 chanics and Schools. No young; Mechanic, such as a Machinist, 
 Engineer, Cabinet-maker, Millwright, or Carpenter, should be with- 
 out it." Scientific American. 
 
 Geometrical Drawing. Abridged from the octavo edi- 
 tion, for the use of schools. Illustraced with forty-eight 
 steel plates. Fifth edition. i2mo, cloth 200 
 
D. VAX NOSTBAND'S PUBLICATIONS. 9 
 
 MODERN METEOROLOGY. A Series of Six Lectures, de- 
 livered under the auspices of the Meteorological Society 
 101878. Illustrated. I2mo, cloth $150 
 
 MORRIS (E.) Easy Rules for the Measurement of Earth- 
 works, by Means of the Prismoidal Formula. 78 illustra- 
 tions. 8vo, cloth I 50 
 
 MOTT (H. A , Jr.) A Practical Treatise on Chemistry (Quali- 
 tative and Quantitative Analysis), Stoichiometry, Blow-pipe 
 Analysis, Mineralogy, Assaying, Pharmaceutical^ Prepara- 
 tions, Human Secretions, Specific Gravities, Weights and 
 Measures, etc., etc., etc. New edition, 1883. 650 pages. 
 8vo, cloth 4 00 
 
 NAQUET (A ) Legal Chemistry. A Guide to the Detection of 
 Poisons, Falsification of Writings, Adulteration of Alimen- 
 tary and Pharmaceutical Substances, Analysis ^f Ashes, 
 and examination of Hair, Coins, Arms, and Stains, as ap 
 plied to Chemical Jurisprudence, for the use of Chemists, 
 Physicians, Lawyers, Pharmacists, and Experts. Translat- 
 ed, with additions, including a list of books and Memoirs on 
 Toxicology, etc., from the French. By J. P. Battershall, 
 Ph.D., with a preface by C. F. Chandler, Ph.D., M.D., 
 LL.D. i2mo, cloth 200 
 
 NOBLE (W.H.) Useful Tables. Pocket form, cloth 90 
 
 NUGENT (E.) Treatise on Optics ; or, Light and Sight, theo- 
 retically and practically treated, with the application to 
 Fine Art nd Industrial Pursuits. With 103 illustrations. 
 I2mo, cloth I 50 
 
 PEIRCE (B.) System of Analytic Mechanics. 4to, cloth lo OO 
 
 PLANE \BLE 'THE). Its Uses in Topographical Survey 
 ing. From the Papers of the U. S. Coast Survey. Illustrat- 
 ed, cloth 209 
 
 " This work gives a description of the Plane Table employed at 
 
 the U. S. Coast Survey office, and the manner of using it." 
 
 PLATTNER. Manual of Qualitative and Quantitative An- 
 alysis with the Blow-Pipe. From the last German edition, 
 revised and enlarged. By Prof. Th. Richter, of the Royal 
 Saxon Mining Academy. Translated by Prof. H. B. Corn- 
 wall, assisted by John H. Caswell. Illustrated with 87 wood- 
 cuts arid one lithographic plate. Fourth edition, revised, 
 560 pages. 8vo, cloth 5 oo 
 
 MLYMPTON (Prof. GEO. W.) The Blow-Pine. A Guide to its 
 use in the Determination of Salts and Minerals. Compiled 
 
 from various sources. I2mo, cloth. I 50 
 
 The Aneroid Barometer: Its Constructioa and Use. 
 Compiled from several sources. i6mo, boards, illustrated, 50 
 Morocco I oo 
 
10 ffi D. VAN NOSTRAND'S PUBLICATIONS. 
 
 PLYMPTON (Prof. GEO. W.) The Star-Finder, or Plani- 
 sphere, with Movable Horizon Printed in colors on fine 
 card-board, and in accordance with Proctor's Star Atlas. . .j$i oo 
 
 POCKET LOGARITHMS, to Four Places of Decimals, includ- 
 ing Logarithms of Numbers, and Logarithmic Sines and 
 Tangents to Single Minutes. To which is added a Table of 
 Natural Sines, Tangents, and Co-Tangents. i6mo, boards, C.Q 
 Morocco . i oo 
 
 POOK (S. M.) Method of Comparing the Lines and Draught- 
 ing Vessels propelled by Sail or Steam. Including a chap- 
 ter on Laying-off on the Mould- Loft Floor, i vol. 8vo, with 
 illustrations, cloth c oo 
 
 POPE (F. L.) Modern Practice of the Electric Telegraph. A 
 Hand-Book for Electricians and Operators. Eleventh edi- 
 tion, revised and enlarged, and fully illustrated. 8vo, cloth. 2 oo 
 
 PRESCOTT (Prof. A. B ) Outlines of Proximate Organic An- 
 alysis, for the Identification, Separation, and Quantitative 
 Determination of the more commonly occurring Organic 
 Compounds. i2mo, cloth *. i 
 
 Chemical Examination of Alcoholic Liquors. A Manual 
 of the Constituents of the Distilled Spirits and Fermented 
 Liquors of Commerce, and their Qualitative and Quantita- 
 tive Determinations. I2mo, cloth 150 
 
 First Book in Qualitative Chemistry. Second edition. 
 
 I2mo, cloth i tp 
 
 PYNCHON (Prof. T. R.) Introduction to Chemical Physics, 
 designed for the use of Academies, Colleges, and High- 
 Schools. Illustrated with numerous engravings, and con- 
 taining copious experiments with directions for preparing 
 them. New edition, revised and enlarged, and illustrated 
 by 269 illustrations on wood. Crown 8vo, cloth 3 oo 
 
 RAMMELSBERG (C. F.) Guide to a Course of Quantitative 
 Chemical Analysis, especially of Minerals and Furnace Pro- 
 ducts. Illustrated by Examples. Translated by J. Towler, 
 M.D. 8vo, cloth ' 2 25 
 
 RANDALL (P. M.) Quartz Operat9r's Hand-Book. New edi- 
 tion, revised and enlarged, fully illustrated. I2mo, cloth. . . 2 oo 
 
 RANKINE (W. J. M.) _ Applied Mechanics, comprising Prin- 
 ciples of Statics, Cinematics, and Dynamics, and Theory 
 ot Structures, Mechanism, and Machines. Crown 8vo, 
 
 cloth. Tenth edition. London 500 
 
 A Manual of the Steam-Engine and other Prime Movers, 
 with numerous tables and illustrations. Crown 8vo, cloth. 
 Tenth edition. London, 1882 5 oo 
 
 A Selection from the Miscellaneous Scientific Papers of, 
 
 with Memoir by P. G. Tait, and edited by W. J. Millar, C.E. 
 8vo, cloth. London, 1880 .10 oo 
 
D. VAN NOSTRAND'S PUBLICATIONS. 11 
 
 RANKINE (W. J. M.) A Manual of Machinery and Mill-work. 
 
 Fourth edition. Crown 8vo. London, 1881 $500 
 
 Civil Engineering, comprising Engineering Surveys, 
 
 Earthwork, Foundations, Masonry, Carpentry, Metal- 
 works, Roads, Railways, Canals, Rivers, Water-works, 
 Harbors, etc., with numerous tables and illustrations. 
 Fourteenth edition, revised by E. F. Bamber, C.E. 8vo. 
 London, 1883 6 50 
 
 Useful Rules and Tables for Architects, Builders, Car- 
 penters, Coachbuilders, Engineers, Founders, Mechan- 
 ics, Shipbuilders, Surveyors, Typefounders, Wheelwrights, 
 
 etc. Sixth edition. Crown 8vo, cloth. London, 1883 4 oo 
 
 and BAMBER (E. F.) A Mechanical Text-Book ; or, 
 
 Introduction to the Study of Mechanics and Engineering. 
 8vo, cl -th. London, 1875 3 SO 
 
 RICE (Prof. J. M.) and JOHNSON (Prof. W. W.) On a New 
 Method of Obtaining the Differentials of Functions, with 
 especial reference to the Newtonian Conception of Rates or 
 Velocities. i2mo, paper 50 
 
 isays 
 
 sils, and a description of the Coal Fields of North America 
 and Great Britain. Illustrated with Plates and Engravings 
 in the text. 3 vols. 410, cloth, with Portfolio of Maps 30 oo 
 
 ROEBLING (J. A) Long and Short Span Railway Bridges. 
 Illustrated with large copperplate engravings cf plans and 
 views. Imperial folio, cloth 25 oo 
 
 ROSE (IOSHUA, M.E.) The Pattern-Maker's Assistant, em- 
 bracing Lathe Work, Branch Work, Core Work,Sweep Work, 
 and Practical Gear Constructions, the Preparation and Use 
 of Tools, together with a large collection of useful and val- 
 uable Tables. Third edition. Illustrated with 250 engrav- 
 ings. 8vo, cloth 2 50 
 
 SABINE (ROBERT). History and Progress of the Electric Tel- 
 egraph, with descriptions of some of the apparatus. Second 
 edition, with additions, i2mo, cloth I 25 
 
 SALTZER(ALEX } Treatise on Acoustics in connection with 
 
 Ventilation. I2mo, cloth I oo 
 
 SCHUMANN (F) A Manual of Heating and Ventilation in 
 its Practical Application for the use of Engineers and Archi- 
 tects, embracing a series of Tables and Formulae for dimen- 
 sions of heating, flow and return pipes for steam and hot- 
 water boilers, flues, etc., etc. I2mo. Illustrated. Full 
 roan ... i 50 
 
 Formulas and Tables for Architects and Engineers in 
 
 calculating the strains and capacity of structures in Iron 
 and Wood? i.?mo, morocco, tucks 250 
 
12 D. VAN NOSTRAND'S PUBLICATIONS. 
 
 SAWYER (W. E.) Electric-Lighting by Incandescence, and 
 its Application to Interior Illumination. A Practical 
 Treatise. With 96 illustrations. Third edition. 8vo, cloth. $2 50 
 
 SCRIBNER (J. M.) Engineers' and Mechanics' Companion, 
 
 cles, the Mechanical Powers, Centres of Gravity, Gravita- 
 tion of Bodies, Pendulums, Specific Gravity or Bodies, 
 Strength, Weight, and Crush of Materials, Water-Wheels, 
 Hydrostatics, Hydraulics, Statics, Centres of Percussion 
 and Gyration, Friction Heat, Tables of the Weight of 
 Metals, Scantling, etc., Steam and the Steam-Engine. 
 Nineteenth edition, revised, i6mo, full morocco I 50 
 
 --- Engineers', Contractors', and Surveyors' Pocket Table- 
 Book. Comprising Logarithms of Numbers, Logarithmic 
 Sines and Tangents, Natural Sines and Natural Tangents, 
 the Traverse 'I able, and a full and complete set of Excava- 
 tion and Embankment Tables, together with numerous 
 other valuable tables for Engineers, etc. Eleventh edition, 
 revised, i6mo, full morocco I 50 
 
 SHELLEN (Dr. H.) Dynamo-Electric Machines. Translated, 
 with much new matter on American practice, and many il- 
 lustrations which now appear for the first time in print. 
 8vo, cloth, New York .". (In press) 
 
 SHOCK (Chief-Eng. W. H.) Steam-Boilers : their Design, 
 Construction, and Management. 450 pages text. Illustrated 
 with 150 wood-cuts and 36 full-page plates (several double). 
 Quarto. Illustrated. Half morocco 1500 
 
 SHUNK (W. F.) The Field Engineer. A handy book of prac- 
 tice in the Survey, Location, and Track-work of Railroads, 
 containing a large collection of Rules and Tables, original 
 and selected, applicable to both the Standard and Narrow 
 Gauge, and prepared with special reference to the wants of 
 the young Engineer. Third edition. I2mo, morocco, 
 tucks 250 
 
 SHIELDS (J. E.) Notes on Engineering Construction. Em- 
 bracing Discussions of the Principles involved, and Descrip- 
 tions of the Material employed in Tunnelling, Bridging, 
 Canal and Road Building, etc., etc. I2mo, cloth . . I 50 
 
 SHREVE (S. H.) A Treatise on the Strength of Bridges and 
 Roofs. Comprising the determination of Algebraic formu- 
 las for strains in Horizontal, Inclined or Rafter, Triangular, 
 Bowstring, Lenticular, and other Trusses, from fixed and 
 moving loads, with practical applications and examples, for 
 the use of Students and Engineers. 87 wood-cut illustra- 
 tions. Third edition. 8 vo, cloth 350 
 
THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 AN INITIAL FINE OF 25 CENTS 
 
 WILL BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE TO 5O CENTS ON THE FOURTH 
 DAY AND TO $1.OO ON THE SEVENTH DAY 
 OVERDUE. 
 
 MAR 8 t94tM 
 
TK37/I 
 
 02244 
 
 UNIVERSITY OF CALIFORNIA LIBRARY