The cross-examination of witnesses, mostly experts, occupied several days, and the argument was full and complete ; no limitation having been placed thereon, and the whole scope of the prior art and of the patents in suit was gone into. The defendant’s brief contains 237 printed pages, and is profusely illustrated, and the court has had the benefit of seeing the practical operation of the devices in question, or some of them, accompanied by.explanations from the experts on both sides. The experts differ radically in some respects, but I think the whole controversy may be reduced to the simple proposition, Must the complainant’s patents be so narrowly and strictly 'construed as to permit others without infringing to make and use structures in which, operating on the same general principle and producing the same result as the device of the patents, the pull travels slightly from a center point to the right and left, or, in order to infringe the complainant’s patents, or either of them, must the resultant pull of the structures be constantly at one point and constant and uniform?
Claims 1 to .4, inclusive, of the senior patent and claims 1 to 3, inclusive, of the junior patent were in issue in this suit, and their validity has been established, as well as their meaning and scope, to an extent, as between these parties'. These claims read:
Of the senior patent:
“1. An electromagnet having a plurality of coils symmetrically disposed around a central axis, the individual axis of each of said coils being parallel to said central axis, and means for producing currents of different phase in said coils.
“2. An electromagnet having a cylindrical core and a plurality of symmetrically disposed coils thereon, the said coils having their individual axes parallel to the axis of said core, and means for- producing currents of different phase in said coils.
“3. An electromagnet having a cylindrical core, with symmetrically disposed pole-pieces at one end thereof, coils on said pole-pieces and means for producing currents of different phase in said coils.
“4. An electromagnet having a cylindrical laminated core, with integral symmetrically disposed pole-pieces at one end thereof, coils on said pole-pieces and means for producing currents of different phase in said coils.”
Of the junior patent:
“1. An electromagnet having a polygonal core and a plurality of symmetrically disposed coils thereon, the said coils having their individual axes parallel to the axis of said core and means for producing currents of different phase in said coils.
“2. An electromagnet having a polygonal core with symmetrically disposed pole-pieces at one end thereof, coils on said pole-pieces and means for producing currents of different phase in said coils.
“3. An electromagnet having a polygonal laminated core with integral symmetrically disposed pole-pieces at one end thereof, coils on said pole-pieces and means for producing currents of different phase in said coils.”
*711Claim 1 o£ the senior patent says nothing as to the shape of the core, while claims 2 and 3 of that patent call for a cylindrical core, and claim 4 for a cylindrical laminated core. Claims 1 and 2 call for a plurality of coils, in the one symmetrically disposed around a central axis, with the individual axis of each of the coils parallel to the axis of the core, and in the other for coils symmetrically disposed on the core and having their individual axis parallel to the axis of such core. Claim 3 calls specifically for symmetrically disposed pole-pieces at the end of the cylindrical core with coils thereon, and claim 4 calls for integral symmetrically disposed pole-pieces at one end of the cylindrical laminated core with coils thereon, and all four claims call for means producing currents of different phase in said coils.
The claims of the junior patent call for polygonal core and a polygonal laminated core instead of a cylindrical core.
These parties have models showing cores with pole-pieces or legs thus:
[1,2] It is seen that in each case we have pole-pieces, and that they are symmetrically disposed with reference to the common center. It is also evident that, if we have coils on these pole-pieces energized by alternating currents so as to attract or take hold of the armature all of same phase, all acting together in pulling and letting go, we would have a constant pounding called “chattering” and caused by the incessant pounding of the core or legs against the armature. It is also evident that if we have out of phase currents so that poles A and B pull or attract when C and D let go, and C and D attract and hold when A and B let go, we will have largely, if not entirely, done away with the chattering. The sum of the pull is centralized. It in also self-evident that symmetrical disposition and arrangement of the pole-pieces is unavailing, unless we have such a symmetrical action and pull of the currents as will exercise a uniform and a substantially constant attraction and pull on the armature at the central point or axis of the whole. If in Figs. 1 and 2, A and C hold and pull while B and D let go, or in Fig. 3, A and C hold while B and D let go, we will have chattering. If, however, A and B hold while C and D let go, we will have less chattering or none. This is merely illustrative. The essential thing and dominating idea of these patents are a substantially uniform distribution of symmetrically balanced magnetic forces, and such geometrical symmetry only as is necessary to secure this. We must have the electrical pull symmetrical and substantially uniform, and if this be so, then the geometrical arrangement is of little or no consequence. But to secure the former the latter is more or less essential. At this late day it is settled that a result is not patentable, but means for producing a result may be. A patent must intelligently describe the means to be employed in producing either a new result or an old result in a better way or a new way. The prior art may anticipate, and it may not, but it is always to be considered and given due weight in construing a *712valid patent and in ascertaining its scope. If a pioneer, it covers and protects everything within its general scope which operates in the same way to produce the same, or even a better, result, provided the elements of the patented structure are all present. Every patent is to be construed in the light of tlegally known structures and devices existing at the time the patent was applied for. In determining the true scope of a patent we are to place ourselves back in the same light the inventor had, or which the law says he had, whether he saw it or not. Here the problem was to use alternating currents in electromagnets, and when energized by such currents to not only attract the armature, but hold in position by a substantially constant pull and without chattering.
[3] To constitute infringement it is not necessary that the infringer use the form of the patented structure, unless form be of the essence of the patent, and hence mere changes in form do not avoid infringement, except in, the case mentioned. If all the elements of a complainant’s valid patent are used by another so combined and arranged as to produce the same general result as the patent, operating in the same way and in obedience to the same laws, we have infringement, even if there be some variation in arrangement which makes the result imperfect. Of course if this departure be a return to the prior art, and but a use of the prior art, then the alleged infringing structure is not within or covered by the patent. This, as I understand, covers the contention of the defendant here. Some new references are cited, and evidence relating thereto was permitted as bearing on the question of infringement, for if defendant’s magnets are in substance the same as the prior art, then they cannot infringe Lindquist.
[4] The new citations or prior art, not before this court either on the final hearing of the original suit or the motion for a preliminary injunction, are the Tesla patents, No. 511,559, No. 416,193, and No. 381,968; Thompson patent, No. 428,650; Fleming patent, No. 430,-898; French patents, Nos. 308,145, and 322,254; Electrical Review Publication of November 29, 1910; and Publication Le Genie Civil. These two French patents and the two publications show magnets attracting an armature and holding it to a core. The others do not. The Thompson and Tesla patents are for electric motors which rotate an armature in a direction transverse of the magnetic core. The Thompson patent shows a shading coil, and the patent is for the rotation of a disk in a meter or motor. Tesla is for a motor device. Mr. Waterman says, in substance, that Thompson would be defendant’s structures D or E, if the core were movable, but it is not. This he admits. Neither Thompson nor Tesla ever suggested in the art that they could be used as an alternating current traction magnet. It is obvious that the Fleming patent owned by defendant has little, if anything, to do with this controversy. Unitary core structure and magnetic symmetry are absent. It is said—and I think correctly said—that the pull would shift about in a circle. The Scott patent and Schuckert were both fully considered in this court and by the Circuit Court of Appeals originally. Scott is a three or a four legged structure, energized by three phase currents, and the resultant pull shifts from side to side. One or more of defendant’s witnesses admits that the pull *713shifts. Is it substantially a central pull? I think not. It would be a waste of time to go through all the defects of this and the French patents which are pointed out in the testimony and not disproved, and which differentiate them from Lindquist. But defendant says that in its structures, A to G, inclusive, the pull shifts, and therefore they are of the prior art, and not within the Lindquist invention when properly construed, which demands a pull that does not or will not shift at all. Defendant’s magnets, A, B, and C, alleged to infringe, have magnetic symmetry exact, except that in the large magnets the pull shifts one-fourth of an inch away from the center. Magnets D and B, because of a reluctance introduced by a curved upper surface causing the magnetic flux to crowd to the center, have substantially perfect magnetic symmetry and magnets P and G, which are furthest away from infringement, have, substantially, magnetic symmetry at the business end or main pole where a shading coil is located. The three-legged Scott and French patents, Nos. 308,145 and 322,254, are fairly illustrated, as is the pull, quite simply.
Energize poles A and B while the'current at leg C is at zero and the pull will center at B. Now energize B and C while the current at pole A' is at zero, and the pull will center at P. In short the pull is traveling or shifting from B to P and back again. The defendant’s structures do not have this shift in pull except to a slight degree.
Defendant’s magnet P may be roughly illustrated as follows:
This magnet has a pivoted armature, indicated by A and R, and a shading coil, indicated by small c and the slot to receive same cut in C, which indicates the laminated core, the laminated armature being indicated by A. The laminae are clamped to a central rib, indicated by R, which rib forms their support and a means for pivoting the armature. The laminations of this magnet are made of silicon steel, not in use when the patents in suit were applied for, and it is conceded it has a higher electrical resistance than ordinary steel, and is less subject to eddy currents, which are troublesome when shading coils are introduced. The windings of the energizing coil are on pole 1 (P-1), and the introduction of the shading coil, small c, in effect to an extent divides the other pole into two poles, indicated as P-2 and P-3, and it is conceded the effect is the same as if the slot were extended to near the base of the core. We have a single phase alternating current which energizes the coil on pole 1, and sets up a single phase magnetic flux which passes to the armature and attracts same towards the pole faces. It will be. noted in the illustration above that pole 1 does not contact with the armature, while the other poles do. When the coil around P-1 is energized, the magnetic circuit is up into the armature, and then fob lows the laminations of the armature, A, to the poles 2 and 3, and *714down into them and the laminations of the core C, and back to the lower extremity of, pole 1. The copper wire, small c, is carried through the slot to the support on which the magnet is mounted, and is closed by a resistance which is adjustable. It is testified by complainant’s witnesses, and is conceded by the defendant, that the electrical effect of this shading coil is the same as , though made of a shorter turn of wire and closely surrounding the pole 3, assuming the slot to be cut deeper. That portion of the magnetic flux which is opposed by the shading coil in its path will lag behind, and is, of course, dephased. Without the shading coil there would be two pulls on the armature, one at pole 1, the left-hand leg, and one at the right-hand leg (now with the shading coil, poles 1 and #). These pulls emanate from the coil around the left-hand leg, and hence without the shading coil would be substantially in phase with each other, and both pulls would come at the same time, and the “let go” would come at the same time, and we would have chattering. With the shading coil dividing the right-hand leg into poles, ÍÍ and 3, and one of these dephased, we have three pulls, one at pole 1, one at pole and the other at pole 3, and as one pole is out of phase with the other, there is a constant pull on the armáture, more or less, and hence chattering is avoided up to a certain point certainly. Of course if the shading coil were not introduced into the structure, the magnetic flux in the face of the right-hand leg (as a whole) would be substantially symmetrical. It is- unnecessary to say that shading coils were old in the art. The contention of the defendant, as I understand, is that in structure F, we have a magnet made up of three poles, all in a straight line operated on a single phase circuit, and that the resultant of the magnetic pull of defendant’s magnet F does not remain at a fixed point, but is constantly shifting from a point near the middle of the right-hand leg to a point near the edge of the windings on the left-hand leg. It is obvious that in magnet F the poles 1, and 3' are not symmetrically spaced-as the distance from 1 to # is much greater than from 8 to 3. This, of course, is not the structure of the French patent, nor is it the structure of Lindquist as described by him. In the French patent the poles are equally spaced, and Lindquist symmetrically disposes his poles. However in magnet F, poles 0 and 3 are in effect symmetrically arranged or spaced as to each -other, although out of phase with pole 1, and both contact with the armature, while pole 1 does not contact therewith. Each pole has a magnetic circuit' of some sort, but these are not equal. This is obvious. There can be no chattering at pole 1, as the pole face does not contact with the armature. It seems to be conceded that in the operation of magnet F there is no chattering, of any account at least, up to a certain point, and that after that point is reached there is chattering more or less, as at a load of 50 pounds and more, although this magnet will support a load of 65 pounds. I take it that a magnet not within the claims of Lindquist here in issue does not infringe, even if not within the claims of the prior art patents. But the operation of the magnet F with the shading coil introduced is that when the coil at pole 1 is energized the magnetic current passes up into the *715laminations of the armature and along to poles 2 and 3 and then down, a part of the current meeting the copper wire, or shading coil, and there is then set up another current around this shading coil point of contact which lags behind the main current and affects the whole of poles 2 and 3, when but for this they would be at zero, and causes both to be out of phase with pole 1 when at zero and pulling on the armature when pole 1 is not, and as a consequence at one point of time poles 2 and 3 are in effect two poles, but at another point of time one pole, or an entire pole affected by the out of phase current. This causes poles 2 and 3 to- hold on the armature when pole 1 is at zero. The air space between pole 1 and the armature is purposely left, for if not left when the current is shut off, the switch would not open. In this structure, magnet A, when the main coil current passes through zero, the shading coil is to do, and does, its work.
The Lindquist patent, as we have seen, calls for a “substantially constant pull,” with a “plurality of coils symmetrically disposed around a central axis, the individual axis of each of said coils being parallel to said central axis.” This is an electrical device operated by an alternating electric current, and magnetic symmetry and magnetic parallelism was in the mind of Lindquist, and that is the spirit and gist of his invention. Geometrical symmetry and geometrical parallelism in the absence of magnetic symmetry and parallelism would accomplish nothing, and this court and the Circuit Court of Appeals, with reference to these claims in issue, have decided that a device of this kind having the magnetic symmetry and parallelism of Lindquist are infringements.. Change the form of the structure as much as we please, but do it in such a manner as to leave a plurality bf coils symmetrically disposed around a central axis with the individual axis of each of the coils parallel to the central axis, and we are within Lindquist. If 1 understand the contention of the defendant, it is, with other things, that Lindquist not only describes, but expressly calls for, a laminated cylindrical core in claim 4 of the senior patent and claim 3 of the junior patent, made up of laminations concentric with the support described in the specifications, and for the reason so much stress is laid on the language used in the description of the magnet described in the specifications, 1 assume the defendant’s counsel claims Lindquist had in mind a core with concentric laminations in all the claims. But this is neither said in the patent nor in the claims of the patent. It was not incumbent on Lindquist to describe every form of structure which could be used in practicing his invention, and he is not confined to- one structurally like or in form similar to the one described by 'him. If the alleged infringing structure has a plurality bf coils symmetrically disposed (magnetically) around a central axis and the individual axis of each of such coils is (magnetically) parallel to the central axis, and we have, a “substantially constant pull” and an absence of chattering, we have Lindquist. It is hardly necessary here to point out the difference between “humming” and “chattering” in these structures. As already stated, when the face or faces of the core and of the- armature come together with force, a noise is made, and if this is rapidly repeated, we have *716what is known as “chattering.” Take a number of thin pieces of metal (laminations) and place them side by side and either clamp them firmly or rivet them together, and, as it is impossible to have perfect constant contact at,all points of the adjacent flat laminae, when we apply an electric current, we get a jarring of the one sheet against another to a slight extent, and this causes a humming. In all laminated cores there will be some humming. The defendant contends that Lindquist’s invention and contribution to the art consists: (1) In providing a concentrically wound core forming a “cylindrically laminated core”; (2) providing mathematically equal lengths of paths of the magnetic circuits in which; (3) the expansion due to heating is radial, so as not to distort the pole faces out of their proper plane, and which when it occurs produces chattering; and, also, (4), in •which form wound coils may be used instead of expensive hand wound coils. Defendant’s counsel says:
“The invention—the contribution to the art—is found in the form of the core which has the structural and functional advantages just referred to.”
It cannot be doubted that Lindquist has these advantages to an extent in the structure which he describes in his patent, but this court cannot agree with the contention of defendant that therein lies the entire invention of Lindquist. It seems' to me that Fig. 5 of Lindquist’s senior patent sets at rest the proposition that his patent calls for and demands a constant uniform pull of the same magnitude at a particular and fixed point at all times. In Fig. 5 of the patent, we have a magnet with three poles symmetrically disposed and having the axis of each parallel with the central axis of the whole. We will indicate this thus:
Assume we have suitable windings and connections and one Pole must be out of phase with the two others to get a constant pull. If poles A and C pull when B is at zero, it would seem clear enough that the resultant pull would be somewhere on the line AC, and when C and B pull and A is at zero, the pull would seem to center at a point on the line CB, and when B and A pull and C is at zero, it would seem the resultant pull would be somewhere on the line BA. That is, while D is the central point of the whole, the resultant pull is not there, but shifts as from 1 to 2, and from 2 to 3, or approximately so. Of course the currents rise and fall gradually and'not abruptly, and it is useless to inquire just the line which the pull would follow in moving from point to point, but it seems clear the point of resultant pull would not center at D at any time.
Nonchattering alone is not the test of infringement, nor is non-chattering at the rated load of the magnet, and I do not understand complainant to so claim.
The defendant's magnets A to B, inclusive, are clear infringements of the claims in issue of the patents in suit. In substance this has been determined by the Circuit Court of Appeals in this suit. The addition of shading coils with shifting of location from right to left or left to right, so done as to maintain magnet symmetry, parallelism, and the other essentials of Lindquist, does not avoid infringement. *717This is mere change of form. There is no functional change, and the mode of operation is the same substantially. The displacement of the shading coils in D and B does not substantially change the center of pull because of the curved face of the plunger. This introduces a magnetic resistance which increases in width as you go from the center, and hence the magnetism is forced to the center. It is self-evident from an examination of the defendant’s devices and of the drawings, which show the location of the shading coils, that an effort was made to get as near as possible to Lindquist, but avoid infringement. It has not been done in magnet's A to U inclusive; but, 1 think, it has been done in magnets B and G. The complainant contends that the axis of each of these structures P and G is at the axis of the main pole at what is termed the “business end” of the magnet, and with which the armature actually contacts. But is this a central axis within the meaning of the Lindquist patent claims in issue? Complainant’s counsel says of these magnets P and G, as follows:
“The axis of these structures is, in our opinion, the axis of the main pole or ‘business end,’ with which the armature contacts and at which the shading coil is located. Before the injunction, the shading coil was located centrally of this polo face, but now it is slightly displaced. The action, however, is substantially the same. It is probably true that the resultant out-of-step pulls on this pole face shift very slightly because of this displacement, but not more than a small fraction of an inch, if at all. The sketch of magnet F on page 72 of defendant’s record by Dr. Sheldon shows clearly that, although the coil has been displaced somewhat, the pull does not shift materially, if at all. As he there shows, the flux produced by the main wound coil is indicated by the line A. This flux, of course, produces a pull substantially centrally at the contacting pole face where the shading coil is located. The local flux produced by the shading coil is indicated by 0 and merely passes around the shading coil into the armature and back again. Wherever it passes from the core into the armature, or from the armature into the core, it produces an attractive pull, as will be obvious. In other words, it produces a. pull on both sides of the shading coil, and, therefore, the resultant is at the shading coil, and, as Dr. Sheldon has shown it, it is an extremely small fraction of an inch from the center of the main coil flux and pull. In other words, if there is any shift at all at this business end, it is such a, minute fraction of an inch that it is not worth considering. So, for all intents and purposes, there is practically no shift of pull on this main or business end, which is the only one that contacts with the armature. The same is true of defendant’s magnet G.
“There was a half-hearted attempt by one of the defendant’s witnesses in his' affidavit to lead the court to believe that there was a material shift, not of the pulls at the business end, but because there was a slight pull between the armature and the return flux leg. Look at Dr. Kennelly’s sketch of this magnet F and of magnet G, between pages 10 and 11 of complainant’s supplemental record, vol. 1. The return flux pole does not contact with the armature. The flux produced by the main-wound coil of course passes around the entire U-shaped core and into and out of the armature. A good deal of it passes through the pivot also, but some of it, of course, passes through the air gap between the armature and the return flux leg, and consequently produces a slight pull on the armature at that point, although probably most of it is blocked off by this air gap.
“Defendant’s contention in this respect is apparently that there is a slight pull at this point, and that therefore the resultant of the two simultaneous pulls produced by the main coil flux is somewhere between the center of the business end and a point at the left of that center. Prof. Ganz (page 201) states that the pull of the main pole would be greater. Ob*718viously it will be greater because there is no air gap between the main pole and the armature, but there is a considerable air gap between the return flux leg and the armature.
“Now it is true, in ail probability, that if you consider the instant of time when there is a pull due to the main wound coil on the business end and on the return flux pole, the resultant of those two pulls will be somewhat to the side of the center of the business end, but what effect does this have on the magnet or the operation thereof? Look at the sketch as shown below. It may be taken to represent defendant’s magnet. When the main coil flux is present, it will produce a pull indicated by K, in the center of the main pole, and a very small flux and pull L at the return flux pole. The resultant of these two pulls, we will say, is at G, shifted a little to the right of the center (in the Cutler-Hammer suit on a similar structure Prof. Sheldon admitted that the resultant of these two pulls shifted 33/100 of an inch on a 60 cycle 220 volt circuit, and 39/100 of an inch on a 30 cycle 110 volt circuit). Now look at Fig. 4 of the sketch. That is supposed to represent the pull exerted by the shading coil when the main coil flux is zero. Obviously the pull exerted by the shading coil can never be as large as that produced by the main coil. Consequently, the amount of pull exerted by the shading coil determines the chattering point of the magnet, and therefore determines the maximum useful work done by that magnet. Consequently, even if we threw away the pull L on the return flux pole, there would still remain the larger pull K in the center oil the main pole, and the magnet would operate precisely as it does at present, so far as the useful work or the chattering point is concerned. In other return-flux pole (although it may raise the words, the slight pull L at the pull-off point a pound or two) has absolutely nothing to do with the chattering point or the useful work of the magnet. Therefore, so far as the work of the magnet is concerned, it is all done by the out-of-step and substantially centrally located pulls E and K on the main pole or business end, as indicated in Fig. 5 of the sketch.”
Figs. 3, 4, and 5, referred to, are as follows:
This is a strong argument and almost convincing. We, in substance, get a result which is substantially the equivalent of Lindquist. It is neither a prior art structure nor a prior art operation. It seems h> me it is one not described by Lindquist or an equivalent covered by the claims of the patents. In this case, on the motion to bring structures A to G, inclusive, under this decree, I held ([D._C] 217 Led. _ 583) that all were infringements. This was before the cross-examination of the witnesses. Op going over the briefs of counsel and all the evidence I am constrained to modify that holding and hold structures A to B, inclusive, infringements, but that structures B and G, do not infringe.
The order or decree will be in accordance herewith.