Appellant failed in its suit to hold appel-lees as infringers of claim 8 of patent No. 538,809, May 7, 1895, and claims 1, 2 and 3 of patent No. 691,332, January 14, 1902, and claims 7 and 9 of patent No. 721,644, February 24, 1903, all granted to B. Wolhaupter, appellant’s assignor. A report of the decision in the trial court is found in 193 Fed. 418. On a record identical with that now before us. the Court of Appeals of the Sixth Circuit, in Railroad Supply Co. v. Elyria Iron & Steel Co., 213 Fed. 789, 130 C. C. A. 447, adj udged that the claims in suit were void for lack of invention. While these decisions are entitled to, and have been given, great weight in our consideration of the case, neither of them can be allowed to control our judgment or relieve us in any degree from giving to the record a complete and independent investigation. We proceed to set forth the. findings of fact and conclusions of law which lead us to hold that the claims are valid and infringed.
[ 1 ] I. The inventions covered by the several claims in suit are susceptible of joint use in one article, a railroad tie plate, and therefore the three patents are properly brought forward in one suit. Of the three, we deem the first the most important, the others being improvements of the tie plate covered by claim 8 of the first patent. That claim reads as follows:
“8. A railway tie plate formed on the under side with devices more or less sharpened adapted to penetrate and engage the tie, and on its upper side with a series of flanges on which the rail rests, substantially as described.”
1. At the root of the case lies the inquiry whether it is true that:
“Wolhaupter endeavored to cover a field of invention far wider than he was entitled, to occupy, and the claim must be held void because it is broader than the real invention.”
This conclusion is based on the finding that:
“The distinguishing feature of his device is not a series of relatively narrow, flat-topped, supporting surfaces suitably spaced apart, i. e., a corrugated top .surface for the plate as distinguished from a flat or plain top surface; Tor, in alluding to the permissible variations in construction, none of which was to depart from the prime object of his invention, which is the stepped flanges or lower stepped projections, he states that the flanges may be divided; that instead of flanges stepped projections may be used; that a square, instead of a beveled, plate may be employed, and the under flanges alone be stepped; that but one end of the plate need be stepped; or that flanges having one or more of their’ends stepped might be applied to a flat *264plate or to tlae numerous other constructions of plates then on the market. His top surface construction may he entirely omitted. Disregarding his thirteenth claim, which relates wholly to spike holes, each of his remaining 13 claims covers his lower flanges, and but 4 of them mention his upper flanges. The body of his plate may be changed, or may assume the form shown in some prior invention, but his stepped form of lower flanges or projections must be used, else his purpose to prolong the life of the tie cannot be attained. The language employed in his patent precludes the substitution for them of any alternative form. . To locate his flanges parallel with the rail necessitates an abandonment of the ‘prime object’ and ‘essential idea’ of his invention, which he declared to consist in providing a plate having on its under side a series of stepped flanges or projections arranged in a diagonal line with respect to the rail flange.”
A reading of the specification discloses very clearly, we believe, that instead of the patent merely describing one invention in which the prime object or essential idea consisted in providing on the under side of the tie plate a series of stepped flanges or projections arranged in a diagonal line with respect to the rail flange, another invention was described, as distinct and separate an invention as if it had been contained i.n a separate patent, and it is in relation to this additional disclosure that claim 8 and some other claims of the patent were framed. We subjoin Fig. 1 and Fig. 2 of the patent and under headings A and B those parts of the description and also the statements of invention which point to the separateness of the two subject-matters:
A. “In carrying out the invention A represents the track rail, B the tie, and and G the tie plate. The plate is preferably, although not necessarily, formed into the shape shown more particularly in Fig. 2, that is to say, what may be termed a trussed shape so that flanges are provided on both the under side as at c, en, 02, ca and top side as at e4. On the top side a way for the rail or what may be termed the rail seat is provided. The flanges on the under side are preferably shaped with comparatively sharp edges co so that they will readily enter and separate the grain, of the tie. As will be seen when in position these under flanges extend parallel with the grain of the tie, the ends being substantially blunt or square and forming abutments to prevent so far as possible the plate from creeping along the tie. It will be observed that by this form of plate there is much less opportunity or chance of buckling than with the ordinary flat plate, since each point where the rail rests upon the plate is supported and trussed by metal which is more or less on edge instead of on its side, thereby forming for each resting point of the rail two girders 07 joined at the apex by the- top flanges c4 where the rail rests. By the above construction also the wood between the flanges o, or, c2, cs will, because of the sloping of the approaching sides of the flanges, be compressed between the flanges, thereby greatly aiding in sustaining the weight put upon the plate, and also in excluding moisture from the wood under the plate. *265A form of plate such as I hare just described Is also of material advantage in that I am enabled to make it much lighter in weight. In this construction of plate with the same amount of material as is ordinarily used in tie plates a much stronger plate Is obtained. Heretofore it has been a commercial necessity in order to get the price of plates where they could be sold for use on the cheaper grade of ties, to reduce and diminish as far as possible the material and consequently the weight of the plate. The result has been that with the plate flat and thin it has buckled when the rail pressure has been brought on it thereby damaging the tie as well as rendering the plate useless as a bearing for the rail; but in my construction since-the points where the rail rests are supported by metal more or less in upright position. th,e plate is greatly strengthened without addition of weight. So also, by a construction as above, the top of the plate where the rail is seated comprises a series of flat portions with a series of depressions between, so that while there is abundant wearing surface between the rail and plate, there is space for sand, cinders, etc., to work out. Again, these depressions permit the draining off of water and moisture so that in cold weather the top of the plate, particularly the space between the rail and plate, is not covered with ice and the friction between the rail and plate thereby either lessened or obviated entirely.”
And it was with respect to the above-quoted particularized description that the patentee made the following general statement of invention :
A. “The invention further contemplates a form of plate provided with top and under flanges which shall be extremely strong in its construction and effective in its designed use, and yet easily rolled or formed into shape as well as very light weight.”
And upon this general statement of invention and particularized description the patentee based not only claim 8, which by its terms is not limited to the preferred trussed shape shown in Fig. 2, but also other claims in which the trussed form and other details of construction are covered. Of these latter claims we quote 10 and 11 as illustrative:
“JO. A railway tie plate the body of which is formed into a series of arches, file lower edges of said arches being provided with tie entering and engaging devices or flanges, substantially as described.
“11. A railway tie plate the body of which is formed into a series of arches the lower edges of which are provided with tie entering and engaging flanges, and the upper side of the plate.having a series of depressions vertically above said engaging flanges, substantially as described.”
B. “I will now describe the second feature of the invention, viz., the ability to adjust the plate to take up any widening of the gauge due to the lateral movement of the plate and rail. The ends of the engaging flanges on the under side of the plate are regularly stepped or cut on a diagonal line as compared with the rail flange. It is of course obvious that the ends of the flanges might be cut on this diagonal line or the entire end of the plate cut, in this way. In my present drawings I have shown the latter construction, the entire end of the plate being sheared diagonally as at D.” (Then follows a particular description of the stepped fiances on the under side of the plate, which for present purposes can be sufficiently understood from an examination of Fig. 1, after which the specification proceeds.)
“It is obvious that the construction just described is susceptible of many changes, and the prime object of the invention still be attained. For instance. the under flanges might be divided, or instead of flanges they might be merely projections and the plate instead of having its end beveled might be square, and the under flanges alone be stepped; or but a single end of the plate need be stepped; or the flanges having one or more of tlieir ends stepped might be applied to a flat plate or to the numerous other construe-*266tions of plates now on the market. In all of the above-named variations the principle employed is the same and each would embody the essential idea of my invention which consists in providing a. plate on its under side with a series of stepped flanges or projections that is to say, flanges or projections arranged on a diagonal line with respect to the rail flange.”
With respect to the subject-matter of stepped flanges on the under side of the plate the inventor made the following general statement of' invention:
B. “My invention has for its object the production of a plate which, while being provided with the top flange to receive the thrust of the rail, is so constructed on its under side that, after it has been moved by the lateral pressure of the rail, it, together with the rail, can be easily moved back to the proper position to maintain the gauge, and be held in that position even more firmly than when originally laid.”
And upon this general statement of invention and particularized description the patentee made several claims to features of the stepped flanges on the under side of the plate. We quote claim 1 as an example :
“1. The combination with the rail and tie of the tie plate provided with a series of stepped tie-engaging flanges or projections on the under side, each arranged to fit into the perforation made by the flange or projection next adjacent, whereby when the plate is shifted transversely of the tie it is also shifted longitudinally thereof, substantially as described.” •
It is thus made clear, we believe, that a serious error is committed in saying that the stepped flanges on the under side of the plate is the prime object and essential idea of the patent. Undoubtedly such a construction is the prime object and essential idea of one of the two separate inventive concepts disclosed and claimed in the patent. Both inventive concepts contemplate the production of a complete tie plate, and both concepts are embodied in combination claims; but it is only with respect to the concept of stepped flanges on the under side of the plate that the patentee makes the statement that many changes of construction might be made without departing from the prime object of the invention, as for instance applying the stepped flanges to a plate having a flat top.
Turning now to claim 8, we perceive that the subject-matter is a railway tie plate formed on the under side with devices more or less sharpened adapted to penetrate and engage the tie, and on its upper side with a series of flanges on which the rail rests. Advantages in having a series, of flanges on the upper side of the plate for the rail to rest upon are clearly pointed out in the part of the' specification which we have quoted pnder the heading A. These advantages are manifestly independent of the character of the tie-engaging flanges on the under side of the plate. Claim 8 very clearly, we think, undertakes to stake out a monopoly of having rail-supporting flanges on the upper side of any railway tie plate on the under surface of which are devices of any sort adapted to penetrate and'engage the tie. The objects .to be attained by the rail-supporting flanges were separate and distinct from, those to be accomplished by the stepped arrangement of-*267the tie-engaging flanges. The mental concepts were different, and their coming into being was in no wise dependent upon each other. Therefore it is inadmissible, in our judgment, to read into claim 8 that part of the specification which is devoted to the stepped arrangement of the tie-engaging flanges and thus limit the right of the inventor to hold a monopoly of the rail-supporting flanges in a railway tie plate unless that tie plate also embodies the'independent concept of the stepped arrangement of the lie-engaging flanges.
And inasmuch as it was confessedly old in the art to have on tie plates tie-engaging devices that ran parallel with the grain of the tie and also across the grain, there is nothing in the prior art or in the patent itself that requires the tie-engaging devices of claim 8 to be parallel with the grain of the tie or across the grain, and assuredly nothing that requires them to be the stepped tie-engaging flanges of claim 1.
[2] Nor is the tie plate of claim 8 to be limited to a construction in which the. depressions between the rail-supporting flanges are vertically above the tie-engaging flanges, nor to a construction wherein the body of the plate is formed into a series' of arches. This is so because the alignment of the depressions on the upper side with the flanges on the under side is covered by claim 11, and the formation of Ihe body of the plate into a series of arches is covered by claim 10; and because in contemplation of law each claim of the patent must be considered as setting forth a complete and independent invention. Leeds & Catlin v. Victor Talking Machine Co., 213 U. S. 301, 29 Sup. Ct. 495, 53 L. Ed. 805; Celluloid Mfg. Co. v. Zylonite Brush & Comb Co. (C. C.) 27 Fed. 291; Chicago Woodenware Co. v. Miller Ladder Co., 133 Fed. 541, 66 C. C. A. 517; Twentieth Century Heating & Ventilating Co. v. Taplin, Rice-Clerkin Co., 181 Fed. 96, 104 C. C. A. 156. To construe claim 8 as being the same as either claim 10 or claim 11 would be to declare claim 8 void as a duplication. Such a course should not be pursued unless it cannot fairly be avoided. In the present instance not only can it fairly be avoided, but further the specification dearly requires a differentiation of claim 8 Horn the other claims of the patent. In quotation A the patentee shows that the form portrayed in Fig. 2 and particularly described by reference letters was only the preferred form — preferred because, while embodying the general characteristics and the law of the structure covered by claim 8 with respect to rail-supporting flanges on the upper side of a plate having on its under side any sort of tie-engaging devices, it also enabled the maker to embody at the same time the stepped arrangement of the lower flanges described and claimed in other parts of the patent.
We therefore conclude that the Patent Office was right in determining that there was a sufficient description and disclosure of invention to support claim 8 as presented; and, further, that claim 8 should be given its prima facie value unless matters outside the patent make such a construction impossible.
[3] 2. A contention is made that the association of elements in claim 8 does not constitute a true mechanical combination, but is merely an aggregation. In Pickering v. McCullough, 104 U. S. 310, 318, 26 *268L. Ed. 749, the one statement of Mr. Justice Matthews that, “in a patentable combination of old elements, all the constituents must so enter into it as that each qualifies every other,” if separated from the context and taken to mean that the natural action of each element must by the action of the other elements upon it be' converted into some sort o'f action other than its natural action, seems to have gone too far. The Patent.Office has continued to grant and the patent courts have continued to sustain combination claims on the theory that it is sufficient if the elements are associated in a unitary structure and there co-operate to produce either a new mode of operation or a new result or the old result in a modified and improved way. In Oshkosh Grass Matting Co. v. Waite Grass Carpet Co., 207 Fed. 937, 125 C. C. A. 385, we upheld, against the defense of aggregation, in a machine for making grass twine, claims for combinations of “means for forcing the material forwardly, a funnel into which the material is received, compression-rolls adapted to receive the material therebetween after said material leaves the funnel, and means, after the material is compressed, for wrapping a twine therearound,” although it is apparent that the natural action of the feeding means could not be changed into some other sort of action by the subsequent action of the compression-rolls or by the still later action of the twine mechanism. In short, the action of no one of the elements could or did in any wise affect the inherent law of# action of any of the other elements. But because the grass twine machine was a unitary structure and because all of the elements co-operated to produce an improved grass twine, the defense of aggregation was overruled. And in Krell Auto Grand Piano Co. v. Story & Clark Co., 207 Fed. 946, 125 C. C. A. 394, citing Burdett-Rowntree Mfg. Co. v. Standard Plunger Elevator Co., 196 Fed. 43, 197 Fed. 743, 117 C. C. A. 206, and other cases, we upheld, against the defense of aggregation, in an automatic player piano, claims for the combination of the controlling levers with the fall-board, although no coaction between those elements was possible except through the mediation of the operator. In the instant case there is a unitary structure in which all of the elements of claim 8 co-operate to produce a new tie plate with new and improved qualities. We are therefore of the opinion that, if claim 8 is to be overthrown, it must be on account of some other defense.
[4] 3. The utility of the invention is abundantly demonstrated by the instant and growing demand, amounting at the beginning of this suit to over 40,000,000 plates a year, and is further shown by appel-lees’ move to share in this commercial success in the face of the certainty of vigorous and expensive litigation. This last-mentioned fact might also be taken as sufficient to evince the novelty of appellant’s structure, since appellees probably would not venture on the possibility of being convicted of trespass if they could have found upon the public commons what would- serve their commercial purposes. And an examination of the 70-odd patents in the record leaves no doubt that appellant produced something that in fact was new. The real question at this point .in the case is whether the inventive faculty or no more than the skill of the mechanic was employed in producing the new *269thing, for, as declared in Bates v. Coe, 98 U. S. 48, 25 L. Ed. 68, anticipation or denial of novelty of a combination claim cannot be based upon picking and choosing pieces here and there from prior patents and publications and thus showing that each of the elements was old. The real test is whether the mental concepts that are distinctly disclosed in and by the laws of the prior nonanticipatory patents and publications left no room for a new and independent mental conception in bringing into working form the new machine or product or process under investigation.
An analysis of all the prior patents we deem unnecessary. If the strongest references cannot prevail, it would be profitless to review the others. Two classes of references will be noticed, rail chairs and tie plates :
(a) Patent No. 142,020, August 19, 1873, to Hudson, and patent No. 203,570, May 14, 1878, to Wells are the strongest references of the rail chair class. Eig. 2 of the Hudson patent is as follows:
[8] The projection from the under side of the rail chair is not a device “more or less sharpened adapted to penetrate and engage the tie”; it is a heavy rectangular rib made to fit snugly into a corresponding depression cut into the tie in advance. And the upper side of this Hudson chair is not provided with “a series of flanges on which the rail rests,” for the depression in the upper surface of the chair so corresponds with the snugly fitting rib depending from the rail that the rail throughout its breadth is seated in uniform contact with the chair. The Hudson patent, in our judgment, furnished no light to those concerned with the problem of producing an efficient tie plate. As wooden ties became more and more expensive and as rails and trains and traffic became heavier and heavier, the cutting of the ties by' the action of the rail flange due to the wave motion of the track became a very serious matter. The shingles or flat plates of wood with which the practical tie plate art began were soon found to be insufficient. These were succeeded by flat plates of iron, and down to Wol-haupter’s time the best that the practical art had was an iron tie plate having a flat upper surface and having tie engaging projections on the under side. Physical exhibits irf this case show that these flat plates became warped or worn through in a comparatively short time. The wearing relation between the rail and tie plate is the same as between the rail and the tie; but with ra'il chairs the situation is different. “Rail chairs” are devices used where the ends of rails come together; they are intended to hold the separate rails firmly together and in alignment and so give them as far as possible the effect of being one continuous rail; they are bolted or otherwise permanently and securely affixed to the rail; and so, as far as cutting or grinding action between the rail and the tie caused by the wave motion of the track is concern*270ed, they belong with the rail. Tie'plates are intended to protect the wooden ties from being worn and cut by the action of the flange of the T-rail due to the wave motion of the track; they are laid upon the surface -of the ties and are attached thereto by means of, tie-engaging projections; spike holes are provided in the plates through which are driven the spikes whose heads clamp over the flange of the T-rail; and so, as far as cutting or grinding action between the rail and the tie caused by the wave motion of the track is concerned, they belong with the tie. And thus, due to the wave motion of the track, w.ater and sand or grit would get in between the rail and the surface of the flat tie plate and would grind and destroy the metal. This was a known condition of the practical tie plate art; but the record does not show that this same condition had to be met by workers in the rail chair art, and the law of the rail chair structure would seem to exclude it; and it was not until after Wolhaupter’s disclosure that any mechanic, expert or otherwise, claims to have discovered in the 22 year older Hudson .patent any instruction regarding the way in which to build an efficient tie plate.
Figs. 1 and 3 of the Wells patent are as follows:
This chair, as Wells directed it to be made, extends over four ties. The depending rib B is not a device “more or less sharpened adapted to penetrate and engage the ties”; it is a rectangular projection, about one-half as broad as the head of the rail; and made to fit snugly into a.prepared channel,.cut across .the body .of the tie; the-channel in the. *271tie is of such a depth as manifestly to weaken it; the purpose of the depending rib is to prevent lateral displacement of the chair and to impart increased strength and stiffness to the central portion of the chair; this strengthening function is called into play where the depending rib acts as a girder from tie to tie; and it is therefore a clear departure from the law of the structure to cut from Wells’s chair a piece the vridth of a tie and then call the rib or girder U a device in. a tie plate adapted to penetrate and engage the tie. The rail is supported upon the ledges P, F*, and C. These ledges result from the presence of the grooves B which, as the patentee says, enable the chair to be made much lighter than the ordinary chair and which also serve as receptacles for sand and grit which may enter between the chair and the rail. But inasmuch as Wells’s device was a rail chair extending over four ties, a distance of about 5 feet, it seems fairly clear to us that Wells never had in mind, and the record does not show that the railway chair art was ever concerned with, overcoming the emery-like grinding and cutting action upon the tie or upon armor on the tie as that action is brought into play upon the ties successively by the wave motion of the track. For, though Wells’s chair may not be as securely and inevitably made a part of the rail as is the case with Hudson’s chair, yet it is clear that Wells intended the law of his structure to be that the chair should be part of the rail, just as a tie plate is a part of the tie. The sides D extend up flush with the upper surface of the flange of the rail; the spikes shown in Fig. 3 and described in the specification are not the ordinary rail spikes, but are specially made of such a width as to lock the rail into the chair; and, though this locking might not be as efficient as the permanent bolts in the Hudson chair which was of no greater length than the width of the tie, yet by reason of the greater length of the chair as specified by Wells, probably the locking was. in fact just as effectual, and there is no doubt that Wells intended the rail to be locked securely in the chair. If this is so, then the upper surface of the tie plate of claim 8 was distinctly novel in itself. But this is not vital to the validity of the claim. The claim is for a combination, and it is very evident that the devices of both Hudson and Wells would have to be redesigned, reorganized, and rebuilt in order to bring into existence the combination of claim 8. That something more than the skill of a mechanic who was acting in the line of the teaching of these patents was required to produce the efficient tie plate of claim 8 might be inferred from the great lapse of time between Hudson and Wells and the disclosure of Wolhaupter, during which time the need of an efficient tie plate was recognized by many skillful persons engaged in improving the roadbeds of railroads; but the fact is really demonstrated by the practical development of the tie plate art.
(b) Prior to 1870 the tie protector was a wooden shingle or plate. From about 1870 to 1880 a simple metal plate with flat top and flat bottom was used. Between 1880 and 1895, when Wolhaupter made his disclosure, a channel plate of equal thickness throughout which fitted over the tie like a cap, a flat topped plate with tie-engaging devices on the bottom, a plate flat on the top and bottom with a shoulder on the top to resist lateral pressure of the rail, and the same sort of a *272plate with tie-engaging devices on the bottom, were in actual use; but never prior to Wolhaupter had the tie plate of claim 8 been produced. And since then the entire development has been along the line of improving the tie plate of claim 8. If the redesigning and reorganization of the rail chairs of Hudson and Wells was so obvious, why did not some one of the many men who were interested in the improvement of the roadbeds of the 1,300 railroads of our country bring forward this combination which has proven to be so desirable?
In addition to the prior practical art, some prior patents need to be noted. Fig. 2 of the Goldie patent, No. 485,030, October 25, 1892, is
The draftsman has indicated a space between the bottom of the rail flange and the top of the central portion of the tie plate, but this is contrary to the intention that Goldie had in mind. He points out that the under surface of the rail flange is slightly raised in the middle and that the depression in the tie plate is made to accommodate the bottom of the rail so formed, and in this way a firm and solid bearing for the rail is afforded throughout the width of the rail flange. Physical exhibits showing the result of the actual use of this tie plate demonstrate that Goldie’s idea accorded with the fact, and that the rail actually bears fully upon the tie plate throughout the width of the rail flange. The manufacture of this plate, as the patentee thereof himself testified, was abandoned, and its use was entirely superseded by the Wolhaupter plate.
Patent No. 494,692, April 4, 1893, to Reece, is for a method of making railroad tie plates. At one step in the process of manufacture there are flanges at each edge which might be rail supports if they were left in the completed plate. But the next step in the manufacture cuts away these flanges at the part where the rail is seated, and the rail rests upon the flat surface of a plate of uniform thickness, except that there are two holes left in the plate by the process of cutting away the flanges. It is difficult to see how Reece, by cutting away what otherwise might be rail-supporting flanges, was instructing the ordinary mechanic how to make a tie plate having on its upper side a series of flanges on which the rail rests, much less instructing him in regard to the desirability of such flanges.
Wilsbn’s tie plate, described in patent No. 522,867, July 10, 1894, is made in several pieces. ' One part of Wilson’s tie plate, by reorganizing it in the light of Wolhaupter’s disclosure, might be made into something similar; .but, when the whole plate is put together as Wilson instructed it to be, the rail rests upon a flat plate of uniform thickness throughout. ■ # °
So far there has been found in neither the practical nor the paper .prior art any tie-plate which approached the idea of having rail-sup*273porting projections upon the upper surface of a plate. Figs. 1 and 4 of the Dunham patent No. 469,386, are as follows:
This corrugated plate is of uniform thickness throughout.. Two purposes were to be served' by the corrugations: First, the patentee says that a corrugated plate three-sixteenths of an inch thick will be stiffer and better adapted to resist the bending or buckling strain to which it is subjected by the weight of the rolling stock than a flat plate which is five-eighths of an inch thick. This tie plate never made any appearance in the practical art. The testimony of practical men convinces us that Dun-ham’s three-sixteenths of an inch material would either break or flatten out under the weight to which it would have been subjected in actual use at the date of the application, and since that time the strains have greatly increased. Second, the corrugations were to afford air spaces between the bottom of the plate and the tie to obviate the decay of the wood. Regarding the bottom formation of this,tie plate of uniform thickness, it is to be noted that the corrugations could not be tic-penetrating and tie-engaging devices, else the air spaces would be lacking and the plate would fit tightly upon the wood as in Wol-haupter's plate. Regarding the upper surface, no suggestion with respect to the corrugations is made except that thus the plate may be made lighter and stronger. And that the character of the upper surface was not material is shown by the patentee’s statement in connection with Fig. 4 that a plate with a flat upper surface corresponds fully to his invention. In our judgment nothing in this paper patent corresponds in letter or in spirit to the tie plate of claim 8.
[5] 4. The law which we believe is applicable to these facts has been frequently declared and may be briefly summarized. Invention of a combination does not lie in gathering up the elements that are employed, but consists in first conceiving that a new and desirable result may be attained by bringing about a relationship of elements which no one has before perceived and then going forth to find the things that may be utilized in the new required relationship. In an old and well-developed field the apparent simplicity of a new device is often the highest evidence of inventive genius. So far as human minds are able, judges should exclude from view the disclosure of the pat-entee, should regard the patentee’s problem as of a time antedating the application, and should therefore not too readily accept the ex post facto wisdom of the bystander. Prior art structures are to be examined in view of the purposes and laws of such structures. It is not enough that a prior art device approach very near the idea of the patent in suit; it must so clearly disclose the idea that it would be ap-*274parent to a mechanic of ordinary intelligence who was not examining the device for the purpose of discovering in it the idea of the patent. For, if he already had that idea, he would not be getting it from the prior art device, but from his own imagination or some other source. Clough v. Barker, 106 U. S. 166, 1 Sup. Ct. 188, 27 L. Ed. 134; Topliff v. Topliff, 145 U. S. 156, 12 Sup. Ct. 825, 36 L. Ed. 658; Potts v. Creager, 155 U. S. 597, 15 Sup. Ct. 194, 39 L. Ed. 275; Diamond Rubber Co. v. Consolidated Tire Co., 220 U. S. 428, 31 Sup. Ct. 444, 55 L. Ed. 527; Ryan v. Goodwin, 3 Sumn. 514, Fed. Cas. No. 12,186; Regent Mfg. Co. v. Penn. Elec. Co., 121 Fed. 80, 57 C. C. A. 334; Faries Mfg. Co. v. Brown & Co., 121 Fed. 547, 57 C. C. A. 609; Ideal Stopper Co. v. Crown Cork & Seal Co., 131 Fed. 244, 65 C. C. A. 436; Wm. B. Scaife & Sons Co. v. Falls City Woolen Mills, 209 Fed. 210, 126 C. C. A. 304.
[6, 9] Another consideration is the presumptive validity of a patent. From long and continued repetition of the phrase the members of the patent bar and of the patent bench sometimes may seem to get into the condition of the man who repeats a word over and over until it fails to convey any meaning to his mifid. But this presumption should be given more than formal recognition. A “patent” is a contract between the government on behalf of the people and the pat-entee. The grant of a patent might have been made conclusive evidence of its validity except against suits by the government for fraud or mutual mistake in the issuance. But the fact that certain defenses are left open to 'the individual should not make us lose sight of the nature of the presumption that attaches to the grant. Not merely has the application been examined on behalf of all the people by experts who have access to all the prior patents and publications of the world; not only has the applicant spent his time and invested his money in procuring the patent; but in most of the important cases the patentee and those working under him have invested very large sums in buildings and machinery and have expended other large sums and put in great energy and effort to build up, by advertising and salesmanship, a profitable business. And this is done before any one challenges the presumptive validity of the patent. Courts therefore should not view the application as of the date of its filing and constitute themselves into a board of reviewing examiners and on nicely balanced considerations find that the Patent Office examiners were in error; but they should consider the patentee’s equities in his business which has developed under the presumptive validity of the patent, should give heed to the place achieved by the patented article in the field of the practical art since the date of the patent, and should therefore decline to sustain the defense of noninvention and to strike dovihi the patent and the business built upon it unless that defense has been established beyond a x'easonable doubt. In this case we find that the appellees have not so maintained their defense.
[7] II. Claims 1, 2, and 3 of the second patent read as follows:
“1. A railway tie plate provided on its upper side witli one or more flanges on which the rail may rest or by which it is directly sustained and on the under side with one ór more tie-engaging flanges extending parallel with the upper flanges and directly beneath the latter, substantially as described-
*275“2. A railway tie plate provided on its upper side with one or more flanges on which' the rail may rest and by, which it is directly sustained and on the under side with one or more tie-engaging flanges extending parallel with the upper flanges and directly beneath the latter and sharpened to permit them to readily enter the tie, substantially as described.
“3. A railway tie plate provided on its upper side with one or more flanges on which the rail may rest and by which, it is directly sustained, on the underside with one or more sharpened tie-engaging flanges extending parallel with the upper flanges and directly beneath the latter, and on the upper side with an additional flange or flanges extending above the plane of the rail-sustaining flanges and adapted to receive the lateral thrust of the rail, substantially as described.”
In the prior art tlie nearest reference is the first Wolhaupter patent, which shows a tie plate having rail-supporting and tie-engaging flanges. The claims now in question are for a specific and limited improvement of the first Wolhaupter tie plate. By limiting the construction to one wherein the tie-engaging flanges are parallel with and directly beneath the rail-supporting flanges greater strength was obtained. Nothing in the prior art anticipates this particular construction. Of course the idea may he tlie same as that which induces the placing of the columns on one floor of a building immediately beneath the columns on the upper floor. But the question here, as was the case with the first patent, is one of invention, whether it involved more than mechanical skill to perceive that such a general strengthening principle might be brought into the relationship between the elements of a tie plate, and then practically to organize the new structure. The prompt acceptance of this idea, the commercial success to which this idea has contributed, and the fact that seven years elapsed after Wol-haupter had disclosed the general combination before any of the numerous engineers interested in railroad construction perceived the possibilities of such a relationship, lead us to resolve any doubt that might be created by the apparent simplicity of the step, in favor of the validity of the patent.
III. Claims 7 and 9 of the third patent are in these words:
“7. A tio plate provided in its rail-supporting surface with transverse grooves or channels, and at one margin of said supporting-surface with a transverse rail-abutting shoulder.”
“9. A tie plate provided in its rail-supporting surface with transverse grooves or channels, reaching to the edge of the plate, and at one margin of said surface a transverse rail-abutting shoulder.”
Here again is a specific and limited improvement upon the original Wolhaupter tie plate. The invention docs not consist in putting a rail-abutting shoulder upon a tie plate. And the concession that rail-abutting shoulders were old in the tie plate art presents no obstruction to the validity of a claim in which the rail-abutting shoulder is one of several elements. In this instance, also, the question is whether invention was involved in producing this specific and limited combination. No prior structure of the practical or of the paper art either meets these limited claims or .suggests, out of all possible relationships of the elements, the particular relationship that is here claimed. For the reasons stated in connection with the other patents we are unable to find that the defense of noninvention is properly sustained.
*276IV. In developing its commercial tie plate, appellant dropped from view many of the separate inventions embodied in other claims of these patents; but it brought into one structure the several inventions of the claims in suit. We have found that none of these claims is nullified by the prior art or limited by anything more than its own terms. So respecting infringement the question is whether these claims read upon the structure put forth by the appellees. A photographic reproduction of appellees’ tie plate is as follows:
As this tie plate is substantially a copy of appellant’s commercial structure, the real inquiry is whether appellant has abandoned its own inventions. To us it seems clear that this tie plate responds in letter and in spirit to the terms and meaning of the claims in suit. We have no difficulty in perceiving a railway tie plate formed on the under side with devices more or less sharpened adapted to penetrate and engage the tie and on its upper side with a series of flanges on which the rail rests. We also observe that the tie-engaging flanges extend parallel with and are directly beneath the rail-supporting flanges. It is likewise obvious that there is the specific and limited relationship of having the rail-abutting shoulder at one margin of and parallel with the rail-supporting flanges and of having both the rail-abutting flange and the rail-supporting flanges transverse to the tie.
The decree is reversed, with the direction to order an accounting on the claim of the first patent and to enter an injunction and or.der an accounting on the claims of the second and third patents.