Bullock Electric Mfg. Co. v. General Electric Co.

SEVERENS, Circuit Judge. This is an ^appeal from an interldcutory decree of the Circuit Court sustaining certain claims of a patent, awarding a perpetual injunction, and ordering á reference to a master for the ascertainment of profits and damages. The patent which is the basis of the controversy is No. 508,637, issued to H. G. Reist, as assignor to the complainant, November 14, 1893, for an “improvement in the construction of armature cores” in dynamo electric machines. The bill charged the defendant, the Bullock Electric Manufacturing Company, with infringement. The answer denied that Reist was the first inventor of the devices for which the patent was granted, averred the anticipation thereof by numerous former domestic and foreign patents and earlier publications disclosing the supposed invention of Reist, and also denied infringement. In his specification the patentee states that his object was to improve the construction of armature cores, so as to obtain ample ventilation for dissipating the heat generated therein without detriment to the inductive qualities of the core. It was well known that in the operation of such machines, heat was generated in the core of the armature by eddies of the magnetic flux in parts remote from the conductor whereby a waste of power was incurred; and besides, the obstruction thus encountered incited heat which might become injurious to the armature. Other factors are suggested which combined to create heat in the core, but that mentioned has been recognized as the principal one. Naturally this difficulty was most serious in large armatures consisting of massive collections of iron. The core of the patent in suit is, says the patentee, to be built up in the usual manner of annular iron laminae in layers, and is' supported by a spider having arms radiating from the shaft, but instead of making the core solid from end to end, he builds it up in sections or bundles of laminae, and, between each two sections he introduces skeleton separators which consist of “ribbed castings, riveted or otherwise suitably fastened to the side of one of the lamina, and each adapted to bear against the outside lamina of the next section.” In another place, he states that, instead of the ribbed castings, he proposes in some cases to make the separators of sheet metal, having one portion turned up at right angles to the other; which latter is “riveted or otherwise secured to” the outside lamina of a section, while the edge of the turned up portion bears against the outside lamina of the next section as in the first instance. These separators, in one form, consist of thin, flat plates secured to the lamina by rivets, and are provided with thin ribs extending outwardly to the outside lamina of the adjacent section, the ribs being radial to the centre of the armature. In another form the ribs are riveted directly on the lamina, and of themselves constitute the separators. In another form several equidistant ribs are -assembled on a skeleton form of separator, which are specially adapted to the toothed, or Pacinotti, style of armature; the ribs in such — 411 — case bearing against the opposite teeth of the separated sections. The material of which the separators shall be composed is “brass or other metal, cast in the shape desired.” For a more complete understanding of the structure it seems desirable to exhibit the drawings (except Fig. 6, which is unnecesary) attached to the specifications, as follows: Fig. 1 shows the separators in cross-section of the armature. C is no part of the separator; B is one of the laminae; A is the flat portion •of the separator; and a is one of the ribs set up upon it. Fig. 2 is a face view showing the separated sections of the armature and the separators in place. These two figures illustrate the separators when used in the form of armature wherein the periphery is entire, and on the surface of which the conducting wire is wound. The other figures are of toothed, or Pacinotti, style of armature wherein the core is in regular spaces recessed at the periphery, and so, of course, the rings or laminae of which it is composed. The wire is wound in the spaces between the teeth. In all the figures, b shows the holes in which the rivets are to be driven to attach the separators to the laminae. .Tt will be observed that in Figs. 4, 5, and 7, the separators consist only of flat pieces riveted to the laminae. And it will be noticed also that in Figs. — 412 — 1 and 3 the ribs of the separators extend outwardly to the periphery of the core, in both forms of the rings, the smooth and the toothed peripheries. The claims which it is alleged are infringed are these: “(1) A laminated armature core built up in sections, and separators attached to the laminas between two consecutive sections, as and for the purpose described. “(2) In an armature core the combination with sections built up of lamina, of separators consisting of ribs of metal between said sections, and in contact with adjacent laminae whereby ventilating space is afforded between the inner and outer surfaces of said core, as described.” “(4) In a toothed armature core built up of laminated sections, separators consisting of ribs extending outwardly from the teeth on one of said sections to the corresponding teeth on the adjacent section, whereby said sections are mutually supported and air passages radial to the center of said core afforded, as and for the purpose specified. “(5) An armature core consisting of laminae arranged side by side and separators attached to certain of the lamina to form a ventilating space or spaces in the core. “(6) An armature core consisting of layers of lamina built up in sections or bundles, and pronged or skeleton separators attached to an outside lamina of each of said sections, whereby ventilating space is provided between adjacent sections, as described.” “(8) In an armature a sheet or lamina having teeth or projections for the reception of the armature coils or armature conductors, and metal separators riveted or otherwise secured thereto, said separators extending toward the ¡joints or free ends, of said teeth or projections.” A very comprehensive, and, as we think, extravagant, interpretation is put upon these claims by the complainant’s expert in laying the foundation for his comparison with earlier structures. It is stated that the» invention is principally intended for dynamo-electrical machines in which the armature' revolves inside the stationary magnetic field. But it is further stated that it is also applicable if the machine* is one in which the field is the revolving member and the armature is stationary. And this is apparent.' But it will conduce to clearness to discuss the subject by reference to a machine organized in the first of the forms mentioned, though some of the characteristics of separators required in this first form might not be so important in the other, a circumstance resulting from the fact that in one form they are in a rapidly rotating organization while in the other that body is stationary. If this had been an original invention, first devised to secure ventilation of the armature by fixing open spaces in its organization through which currents of air could be carried, it must have been admitted that it exhibited a new and very useful device to supply a need that had been recognized in the art as a serious one. But it was not. It followed in the wake of numerous patented inventions intended to meet this need, and the defendant contends that no new idea is shown by the Reist patent, but only the mechanical skill which one conversant with the subject, and with what had already been pointed out as means for meeting the difficulty, might be expected to supply. It is contended that Reist discovered no new principle or mode of operation which is represented by the means he proposed, and that the means which he did propose effect no new result, but are the equivalent of former constructions operating in the same way, more skillfully designed perhaps, but carrying forward an old idea in better forms. This defense makes — 413 — it necessary to .find out what had already been discovered and known in the art when this patentee brought out his method of constructing separators in armatures to effect their ventilation. We cannot undertake to critically analyze and exhibit in detail all the former constructions of separators shown in tne patents found in the record, but shall confine ourselves to a few which seem to us most relevant and important; though there are many others in the record which illustrate the general idea of the proper method of providing for ventilating the armature, that is, by separating the core into sections and putting separators between them, to maintain the spaces for-currents of air. We may preface what we have to say by stating that the necessity, or at least the great advantage, of ventilating the cores of armatures had been known and appreciated several years before Reist gave his attention to it, and that the method of doing this by dividing the core into parallel sections and interposing between them separators to hold them apart, and provide space for the circulation of air in the body of the core had been devised, and for several years practiced, in the manufacture of such machines. In 1881, one Brown obtained a patent for improvements in armatures which showed them built' up of sections of soft iron bars, a method of construction since superseded by the use of thin laminae instead of bars. The need of' ventilation was the same as in the later construction. For the purpose of securing it, Brown employed plates of sheet metal, such as Russian sheet iron, secured to the outer bar of each section. These plates were provided with flanges turned up at right angles from each edge of the plate. The edges of the flanges-did not bear directly upon the opposite bar, as in the Reist patent, but contacted with the corresponding edges of a similar plate riveted to the opposite bar. These plates separated the sections and formed an air chamber. The sides of the plates were radial to the central line of the armature. Several openings were made in the flanges on both sides of the plates through which the air from the central chamber was drawn through the armature, some of it passing directly through the periphery and part sidewise through an opening in the bars, and thence out through the radial chamber. And in his third claim, he includes these plates to form, as he says, “an air chamber which communicates with the external air, whereby, during the revolution of said armature, air will be drawn into said chamber and circulate into and through the air spaces of the bobbins or sections.” In November, 1883, patent No. 288,051 was granted to Giles for an invention, the sole purpose of which was to provide for the formation of air spaces in the core of armatures. Instead of dividing the lamina; into bundles or sections, he attached separators upon “one or both” sides of each lamina (he calls them “rings”) “oblique ribs uniformly arranged and spaced, and inclining away from the direction of the armature,” a feature shown in other patents, evidently designed to facilitate the draft of air. The ribs were attached to the “rings” and extended the entire width of them, but were insulated from them to “prevent electrical currents from traversing the core.” All his — 414 — claims'call for these ribs as "means ’for separating the rings and affording ventilating spaces. On March 3, 1884, R, E. B. Crompton obtained a patent in England for an invention of improvements in dynamo electric machines which was patented also in Belgium in 1885, and in the United States August 7, 1888, where it was numbered 387,343. He states that the objects of his invention are, referring to the armature: “ * * * Third, the arranging of the winding and its insulation so as to allow of the free access of a ventilating current of air to every part of its surface, and the providing' this ventilating current by simple means.” His preferred form of armature is built up of annular rings of soft sheet iron mounted on the dove-tailed radial spokes of the spider. Corresponding dove-tails are made in the inner edges of the laminas. Fig. 2 shows one of the rings (or “laminae,” or “washers,” as they are indiscriminately called in the art) of the Gramme form of core. The rings are composed of sections arranged in parallel form with ventilating spaces between the sections. In building up the armature, he lays the rings one upon another, properly insulated, and "“at regular intervals as shown at g, gl, g2, g3, etc., the washers may be spaced from one another by strips of material, i, i, which must also insulate them from one another and thus leave radial ventilating channels or space.” He says “radial” ventilating channels or spaces. Reference to his drawing shows that the .separators and spaces are in a general sense radial, but not quite so, for they are turned backward somewhat (having regard to the motion of the armature indicated by the arrow) so as to promote the exit of the air, as in other constructions heretofore -referred to. And, he adds cautions applicable to the plain ring core and to the toothed ring core to so construct them and form the winding as not to obstruct the ventilating spaces. At this point, we remark upon a matter which will’ be referred to later. In his drawings Crompton shows in Fig. 2 in cross-section a smooth armature and the spacing ribs, i, i, extending to, its periphery. In Fig. 4 he shows a toothed armature, but the spacing-ribs are not shown. And a question is whether in the latter form he extends the ribs along the teeth. Inasmuch as he had extended them in the smooth form to the surface, and the teeth are so thin and fragile as to' require a supporting separation to prevent their collapsing into the ventilating space, and no other way of preventing it is suggested, we think it is a reason — 415 — able supposition that he extended the separators back of the teeth to the periphery of the armature. He could not preserve the air space without resorting to some such expedient. In claim 3 he takes in the separators, called “spacing pieces,” as follows: “(3) In an armature for a dynamo-electric machine, the combination of a detachable spindle with radial bars and iron disks arranged with spaCing-pieces and radial ventilating passages, substantially as set forth.” It is to be observed that it is not here stated of what kind of material his separators are composed. In the specification he terms them “strips of material” and, as just noted, in his claim, “spacing pieces.” But it is apparent that any kind of insulating material sufficiently rigid to maintain the separation would meet the requirements. We say “insulating material” because the patentee takes pains throughout to lay stress upon the insulation of the rings. A later patent to R. R. Cohen, No. 397,340, was granted February 5, 1889, the object of which was to so build up the armature “that ample opportunity will be afforded for the perfect ventilation of the armature during the working of the machine.” In this the armature was divided into sections, at one end of which were plates extending to the periphery of the core. On these plates were projected “lugs” as they are called in one place, or “distance pieces” as they are elsewhere called, or “spacing pieces” as designated in the claims, bearing against the opposite section and providing the ventilating spaces. Fig. 3 shows one of these plates with the spacing pieces, a, a, projected thereon. Two features of these spacing pieces may he noticed here. They are radial to the center of the armature, and they are uniformly set with their wider dimension in the line of the current of air plainly indicating that the patentee had in mind the idea of so constructing his spacing pieces as that they should economize room for the current of air in the ventilating spaces. Claim 1 in this is as follows: “(1) The combination of the shaft of an armature with the shell comprising hollow bobbins and interposed spacing-pieces, each hollow bobbin consisting — 416 — oí opposite end disks and an interposed tubular shell, all substantially as specified.” In a later patent to Cohen, of March 11, 1890, he converts the “wings or lugs” in each radial line into one wing extending across the width of the rings. In other respects the organization of the armature exhibited in the Cohen patents is unlike the forms in the other patents to which we have referred; but the difference does not materially affect the particular purpose common to all these constructions in the provision of ventilating spaces, which we are now considering. The new thing which Reist claims to have invented is his separators, and the question is whether he made any advancement beyond the former art which partakes of the quality of invention. And if he did in what did it consist? Separators in a considerable variety of forms had, as we have seen, already been provided. There was nothing new in this general fact. Their purpose in such machines was understood. The principles and method of construction had been illustrated and were well known. If he could develop some new principle or some new means having the qualities of originality and utility, he might, if the sources of his new light were obscure and required more than the intuition and skill of those in the practice of the art to discover or recognize them, be entitled to a patent. But, ’ if he was in the beaten path and saw only how the means already known and practiced could be improved in detail, such as by changing the form, extending the size; carrying the same thought into another part of the same structure with no change in the character of its function there, or in any like case which does not display something extraordinary, something beyond the sphere of the common intelligence of those conversant with the art,, such exercise of skill would not be invention. Referring to his claims, it would seem as if Reist considered himself the originator of the idea of using separators as a device for the division of a solid armature and transmitting the air between the sections. But we have seen how far this was from the. fact. It had for a number of years been embodied in various forms and put into practical use. He simply changed the form of such separators and attached them to one of the outside lamina of the sections. By this attachment before, using those parts in building the armature, he made them integral as they were in Cohen’s patent, and in Giles’, and in the still earlier patent to Brown. As has been already shown, in some of the previous structures, the separators were attached to the sections; in others they were not, and the clamping of the sections together seems to have been relied upon for holding the separators in place. And, indeed, there could be no invention in making' such attachment to something, if it seemed necessary to maintain the separators in position, and -prevent their flying out in the rotation of the armature; 'and what mode of attachment could be more obvious than securing it to the laminas with which it was in contact in former constructions? It is contended that there is an advantage in Reist’s conception of making the strips and spaces radial to the center of the armature. Whether there is any advantage in an adherence to that form we can — 417 — not say. It would seem so. If there is, it was so formed in Crompton’s patent, and such was the form in Cohen’s and in others. - But the principal novel feature which Reist contemplated was that his separators should be thin; the object being to obstruct as little as possible the ventilating spaces. It seems hardly possible to say that there could have been any invention in this even if Reist was the first to advance in set form such an idea. Every electrician would appreciate that it was desirable to use as little inert material as might be, as well as that what was used should be put in the line of the current of air, ,and not cross-wise of it. This Crompton did to make his “radial ventilating spaces” and it was what Cohen did in both his patents when he set the sides of his spacing pieces in the direction of the air current. But Reist does not state how thin his separators shall be, nor how far they shall project toward the opposite section. If it he answered that he leaves these things to the judgment of the builder, the result would he that he appreciated that the builder would ba competent to determine what would be the right dimensions for properly ventilating the» armature, and this had been the only problem, if we call it such, that Reist himself had before him, for the state of the art was already advanced to that question. Supposing the artisan to take up the work of building an armature and using Reist’s separators having ribs set up on a base attached to a lamina, how far shall they project? And of what thickness shall they be? Any intelligent person would say that these dimensions should be in proportion to the size of the armature. the bulk of the material to be ventilated. And so what would be thin in a large armature would be thick in a small one. Thus it is a question of degree. Or suppose the artisan wishes to construct a ventilated armature, and does not care to use Reist’s separators, how shall he know in what manner he shall avoid the Reist patent? How thin must his separators be to do this ? Or suppose a purchaser wishes to buy one. He must look out for the patent. By what test or comparison shall he be guided? These inquiries enforce the rule that the patentee must describe with sufficient certainty the particulars of his invention so that the artisan and the public may know the character and limits of it, and how it is to be distinguished from others which the one may make, or the other purchase, in safety. Again some of the forms which Reist suggests as embodying his invention consist only of a flat member lying on the face of the lamina. It must be obvious that in such case it must be thicker than in case he forms a rib or lug upon it. For ventilation could not be adequately secured without a considerable thickness of his flat separator. The same conditions in respect to the thickness of his separator would exist as in the Crompton patent, and in both cases the same question- would arise, how thick shall the separator be? And, if in Crompton’s case that question is left to the judgment of the builder, so it is in Reist’s, and it follows that the latter has taken no forward step whatever. This feature of the indefiniteness of description was adverted to by the Chief Justice in delivering the opinion of the courtin Guidet v. Brooklyn, 105 U. S. 550, 26 L. Ed. 1106, and influenced the,decision. The patent was for an improved stone pavement. In earlier practice the blocks had been comparatively smooth but slightly rough on, the sides, — 418 — and when put in place their upper edges were too near together to afford sufficient holding for the caulks of the shoes of draft horses used on the road, and strips of wood or stone had been put in the spaces or the edges of the blocks were chamfered. The patentee specified that the blocks he proposed should be sufficiently rough to hold them in composition sufficiently far apart to give the proper space between their edges, and so to relieve the objection. But the court, referring to (among other things) the fact that the patentee did not specify any degree of roughness, .held that he had only carried out an old idea making a change, not definitely indicated, in degree only, and that the patent could not be sustained. The case of Preston v. Manard, 116 U. S. 661, 6 Sup. Ct. 695, 29 L. Ed. 763, is of still more direct application to this feature of the patent. The patent was for an improved fountain hose-carriage. The new element brought into combination with old elements was “a reel of large dimensions to allow the water to pass through the hose when partially wound thereon.” Mr. Justice Gray, in the opinion, said: “The requisite diameter of tiie reel, and its proportion to the size of the hose, are not defined in the specification, but are left to be ascertained by experiment, or from general knowledge. If the patentee had discovered anything new in the size or proportions of the reel, requisite to allow the water to flow through the hose, he should have described it with such precision as to enable others to construct the apparatus.” It was reasonable to suppose that a large reel would carry a hose giving a more direct passage for the water, than one with narrower convolutions, but the improvement was only in the form and size of the reel, and there was no specification of the size except that it should be large. And the very gist of Reist’s invention was in the dimensions of his separators. In Howard v. Detroit Stove Works, 150 U. S. 168, 14 Sup. Ct. 68, 37 L. Ed. 1039, the patent was for an improvement in stoves. The novel feature of it was a flange cast or riveted upon the inside of the lower end of the fire pot. This flange formed the support of the grate. This part of the fire pot was exposed to undue expansion by reason of excessive heat at that place. Stoves had formerly been constructed with a flange on the inside of the upper edge of the ash-basin, just under the fire pot, and the grate rested upon this. The patentee conceived that it would be an advantage to put the flange at the bottom of the fire pot and make it so wide as to gather the ashes there, and thereby minimize the heat in that part of the fire pot. And he obtained a patent embodying that idea. But, though he described the flange and the purpose of it, he failed to state what the inward projection of it should be. Eor this reason the patent was held void. It was also held that the transferring the flange from the ash-basin to the fire pot did not involve invention. And see, also, the Incandescent Lamp Patent, 159 U. S. 465, 16 Sup. Ct. 75, 40 L. Ed. 221. These cases would seem to lead to the conclusion that the description of the separators in this patent is insufficient, and not in compliance with the requirements of the statute; and, if so, the patent is void. But there are other considerations leading to the same result, which we think should be stated. In none of the claims of the patent in suit — 419 — is it stated whether the separators are to be of the form of a flat spacing piece lying against the lamina or of sü'ch a piece with a flange or rib upon it, and either would answer the calls of the claims. And some of the forms in earlier patents would answer just as well. In some of the forms of Reist’s separators there is a web or skeleton form the function of which is to hold the ribs in proper relation to each other. But the flat sheet metal plate in the Brown patent attached to the side of the section performed that function. In Cohen’s, a disk in the same form as the core of the armature held the distance pieces in place. And as we understand the patent of Giles, his “ribs” were attached to the lamina. For he says his “rings” were provided with “ribs.” And in two of these patents the lamina performed the function of the web in holding the separators in place and therefore were the equivalents of a web. ' And in Cohen’s the web to which the separators were attached consisted of a disk. But Reist makes no claim which includes the feature of a web unless it be in the projection of the separators upon the teeth of a Pacinotti armature, a matter we shall consider presently. The eighth claim of the patent is for separators' in a toothed armature extending upon the teeth and attached thereto, and this, it'is contended, is a particularly advantageous feature. Referring again to the specification, we find it stated that the object of this construction is to hold the teeth of the opposite lamina; apart and prevent their collapsing into the ventilating channel when the sections of the armature arc clamped and pressed tightly together. But, .having understood the need of air channels between the sections, and having adopted a scheme of separators to maintain the spaces, which separators extended to the periphery of the entire rings of the Gramme form and used the same device in the body or entire portion of the rings, there could be no invention in extending the same scheme to accomplish an identical purpose in respect to the teeth in the other form. It is as plain a case as can be for the application of the rule laid down in Smith v. Nichols, 21 Wall. 113, 22 L. Ed. 566, and followed in many subsequent cases by the Supreme Court, and in several cases by this court. That rule, as applied to the facts of this case, is that the mere carrying forward a new or more extended application of the original thought, is not such invention as will sustain a patent. An apposite case among our own decisions is Soehner v. Favorite Stove & Range Co., 84 Fed. 183, 28 C. C. A. 317, where it was held that, it having been ascertained that an advantageous form in which to build the sides of a stove was to make them curved in along the upper and lower borders so as to enlarge the oven space, there was no invention in extending a similar form to the vertical flues in the rear end of the stove to accomplish a like purpose of enlarging the flues. And in American Carriage Co. v. Wyeth, 139 Fed. 389, 71 C. C. A. 485, 488, we held that the prolongation of a flanged strip of metal, which had been used to connect and hold in place the knees of a sleigh with the rave, so as to connect and hold in place the knee and the runner was not invention, citing L. Schreiber & Son v. Grimm, 72 Fed. 671, 19 C. C. A. 67, and Dunbar v. Myers, 94 U. S. 187, 34 L. Ed. 34. — 420 — While the specification o'f 'the patent in suit states that the separator must be of metal and expresses a preference for brass castings, there is no limitation to any kind of metal in any of the claims, and in the first, fourth, fifth, and sixth -claims, the separators are not required to be of metal, while in the second and eighth they are. And if some kind of metal is required, it is not stated in these claims what kind it shall be; whether it should be a kind which shall insulate the opposite laminae, or be of a kind which shall have conductivity for the magnetic current. ’ In some of the previous patents the material of the separators is not stated. In others it is stated to be of metal, and in one, of hard rubber. If the purpose of Reist was to introduce some kind of metal which had not been used before, as a good medium for the movement of the magnetic flux, or for insulation, or for any other purpose, he should have stated what it was. In the absence of any attempt to do this, we incline to think that he thought of metal for the material because of its greater strength and rigidity in proportion to the space it would occupy. But it is not material since he proposes no special kind of metal which would distinguish his own from former constructions. Another rule which has been only suggested thus far by the recital of the facts may now be formally stated. Among the facts it appears that all the features of the means proposed by Reist were to be found in earlier patents and that what he did was no more than selecting from them parts which in his combinations performed essentially the same functions as they did in the organizations from which they had been taken. We have often held that there is no invention in this, and such is the rule recognized by the Supreme Court. We refer to a few of our own decisions: Goodyear Tire & Rubber Co. v. Rubber Tire Wheel Co., 116 Fed. 363, 369, 53 C. C. A. 583; Burnham v. Union Mfg. Co., 110 Fed. 765, 770, 49 C. C. A. 163; Overweight Counterbalance Elevator Co. v. Henry Voght Machine Co., 102 Fed. 957, 761, 43 C. C. A. 80; Campbell Printing Press & Mfg. Co. v. Duplex Printing Press Co., 101 Fed. 282, 294, 41 C. C. A. 351. There is a recognized exception to this rule, or rather counterpart of it, when in the new combination a new mode of operation is effected which produces an original result, as we said in Dowagiac Mfg. Co. v. Superior Drill Co., 115 Fed. 886, 901, 53 C. C. A. 36. But this must be something novel in principle, and not a combination operating upon a principle which was common to the old machines from which the elements were taken, as in the case at bar. Ror the reasons stated, we are of the opinion that there was no patentable invention in the Reist patent. We admit that it exhibits a good, a skillful, and perhaps a more useful piece of workmanship than had before been attained. But it is in substance only an illustration of the principles discovered and applied by his predecessors by, means the samé as, or equivalent to, his own. It is stated and pressed upon our attention that separators like his have gone into extensive public use. But it is easy to see how this might happen without attributing to the patent anything- new in principle. A thing or a combination of cooperating parts may be invented and the original embodiment of it be shown in a crude and imperfect form. The skill of the trained work — 421 — man will develop the idea of the inventor in more refined, more delicate and more exactly suitable forms than the original. He may cut away. needless bulk, he may increase the size of parts, he may make them stronger, if need be, by the substitution of one familiar material for another, make them lighter or heavier, he may divide one part into two, or combine two in one, or make any other transformation of details, so long as he is pursuing and working out the original discovery or invention by the exercise of the insight, good judgment, and expertness which he is expected to possess and apply. And this improvement may go on so long as any improvement in bringing the means already supplied to greater perfection can be made, and yet it continues to be only an embodiment of the primal idea. In all the useful arts such improvement as this is being continually carried forward. It is easy to understand that purchasers should prefer and insist on having the article, especially if it be a machine, in its best form, and the latest improvement would be likely to measurably supersede the older one. Another thing which affects the value of the presumption arising from the general acceptance and use of a thing is that in modern usages of trade such articles are often pressed upon the attention of purchasers instead of being chosen. And in many instances an article gets into general use from being pushed upon the market by energetic tradesmen with strong financial backing. No doubt the presumption still obtains, but it does not and for that matter never did, have a controlling influence, unless the question of invention hangs in doubt. And our impressions of the present case are such that we do not invoke the presumption as a helping factor. Exit, if we were to hold the patent valid, it must be because of his adaptation of the separators for attachment to the adjacent laminae for the purpose of securing them in place. In his description he enforces this point, and in all the drawings which indicate the manner of associating the separators with the laminae, holes are shown for the riveting. And in the form of separator wherein he proposes to turn up a part of the separator to constitute a rib, it would seem to be necessary that the part lying against the lamina would be secured to it, in order to hold the ribs in proper position. In his directions for building the armature, he says: “A suitable number of laminae are bundled together in sections, and a lamina provided with separators, as shown added thereto;” which indicates that the attachment has already been formed. It is true that in his second claim there is no requirement that the separator shall be attached to the lamina. But in others there is, which indicates an intended distinction. And a question might arise whether this claim is not broader than his invention, if the method of attachment is to be regarded as the substance of the invention as we are now supposing. But the defendant does not adopt this special feature which we are now crediting to the Reist patent. Its separators are not attached to the laminae, but in that one of its forms which most resembles Reist’s, depends upon a dove-tailed connection with the arms of the spider to prevent their being thrown out by centrifugal force during the rotation of the armature, a method of connecting the body of the core with the spider previously in use. The — 422 — defendant connects sections of the ribs together by a web to keep them in-proper distances apart. .But this was- accomplished in a similar way in the structures of older patents, not probably by a web so thin. But, as the complainant does not limit the thickness of the web, we may dismiss that subject from further consideration, with the added remark that the web used in their separators by the respective parties is, as we gather from the record, of substantially different thickness. The other form of the defendant’s separator consists of a triangular bar attached to the lamina by raising fingers cut into the body of the lamina on opposite sides of the bar and after the bar is inserted between these fingers, pressed down upon the outer surfaces of the bar; its other side resting upon the lamina. This triangular bar has no adaptation for such attachment to the lamina as would prevent it from being thrown out by centrifugal force, for the thin strips of fingers' of the lamina would have little griping force of their own and the security of the bars from longitudinal displacement would depend rather on the compression of the sections in clamping them together to form the core, than on the capacity of the fingers. We do not think that this can be the attachment intended by the Reist patent, which seems to require a more integral union, one which would of itself prevent the slipping out of the separator. The separator has little or no resemblance to that of the patent unless, it be to that form consisting of a flat bar thick enough to give the requisite spacing and that was the form of rib shown in the Giles patent. The complainant’s expert argues, curiously enough, that this form of defendant’s separator infringes the Reist invention notwithstanding it is too bulky. But he differentiates Reist’s separators from'.the earlier ones on that very ground. And, in fine, when this separator of the defendant is brought into combination with the lamina, it constitutes a substantially different combination from anything shown by the patent in suit, if it be conceded that the separators of the latter differ-substantially from those in earlier patents by reason of this attachment to the lamina. We are therefore of opinion that, if the controversy were brought to a question of infringement, the defendant’s form herein first considered does, not infringe, and we are strongly inclined to think that the second does not. The decree of the court will be reversed, with direction to dismiss the bill, with costs.