This is an appeal from the decision of the District Court enjoining the manufacture, sale, and use of the Superior rail anchor, made pursuant to letters patent No. 1,015,129, granted January 16, 1912, to John A. Bodkin, as being in infringement of claims 1 and 2 of letters patent No. 656,470, granted August 21, 1900, to John E. Pope, claim 3 of letters patent No. 720,362, granted February 10, 1903, to Edward Eaas and Hiram H. Sponenburg, and claim. 2 of letters patent No. 1,021,387, granted March 26, 1912, to David F. and David E. Vaughan.
The following drawings from the patent in suit, the Bodkin patent, and certain alleged anticipating patents to be considered, together with the claims in suit, will help to clarify the situation.
Pope’s claims are:
1. A device for preventing creeping- of rails, comprising a crossbar extending transversely under the rail and having rigid abutments thereon bearing against the opposite edges of the rail flange and adapted to clamp the same, substantially as described.
2. A device for preventing creeping of rails, comprising a holder extending under the rail and bearing against the opposite edges of the rail flange, said holder being set between the ties and having a connection extending laterally therefrom to a tie, substantially as described.
Laas and Sponenburg’s claim 3 reads:
A railstay consisting of a bar extending across the under side of the rail and provided with jaws gripping said rail, and a flange depending from one *634(>nd portion of the bar and bearing on tlie side of the tie and disposed at an angle in relation to the length of the bar to hold the opposite end portion of the bar normally out of contact with the tie as and for the purpose set forth.
Vaughan’s claim 2 reads:
The combination, with a cross-tie and a railroad rail having a base flange, of an anticreeper comprising a flat bar extending beneath the base flange and having flat upwardly extending end portions provided with inner edges facing the side edges of the base flange, said bar being provided with means for supporting it on the base flange, and maintaining its side faces and the side faces of its end portions substantially in vertical position, and means acting on the cross-tie and on one end of tlie bar and maintaining the other end of the bar in spaced relation to the cross-tie, one of said end portions having a lateral bend therein.
*635Takrant, Fig. 1.
Noonan, Fig. 2.
Bkown, Fig. 7.
Sponenbuhg Patent, No. 668,423, Fig. 2.
Rail creeping, which offers a most perplexing problem to rail way-engineers and which rail anchors are designed to prevent, is due to several causes — primarily, to the wave motion of the rail induced by the passage of engines and cars thereon. Under the weight of the train or wheel load, the rail is deflected downwards causing a small though appreciable rise in the rail immediately in front of the on-coming wheels which, in their forward thrust, tend to force the rail in the direction of the traffic. Another cause of this longitudinal movement of the rails is the continued or excessive use of brakes when approaching stations or grades. The shock transferred to the rails by the frictional gliding of the heavy wheels over the rails tends in a marked degree to drive them forward. When heavy rails are laid upon a carefully tamped rock ballast and the traffic does not preponderate in either direction, there is comparatively little creeping. When, however, the rails are light and the roadbed not well ballasted or when the traffic is entirely, as in the case of double tracks, or preponderantly, in one di*636rection, the problem of rail creeping becomes serious. If the movement of the rails is not checked, the ties will skew and the roadbed will become weak and unsettled; the space always left between the adjacent rail ends to allow for the longitudinal expansion and contraction of the rails under extremes of temperature will close and in consequence the rails will be likely to bend or buckle laterally and vertically.
A feature common to practically all rail anchors is that one part of the anchor is clasped upon the rail and another part is attached to or abuts the tie, so that the longitudinal movement is communicated to the tie, which, being embedded in the ballast, is better able to resist that motion.
The development of the art, shown by the patents in suit and by the testimony, reveals the discovery of and adherence to certain principles and general features as highly desirable, if not indispensable, in an efficient rail anchor, and the rejection of others as inadequate and impracticable. Many of the early anchors were applied only at the rail joints ; but anchors spaced so far apart were not strong enough to hold the rail and increased the tendency of the tie to skew at the rail joint. The early devices were bolted to the rail and spiked to the tie. This bolting not only weakened the rails by the drilling therein of holes to receive the bolts, but necessitated continual inspection to keep the nuts tight on the bolts. No less weighty are the objections to spiking the anchor to the tie. A reverse movement, caused by a temporary reversal of traffic, by the expansion or contraction of the rails, or by the reaction from the wave movement, must either carry the tie with it, thus weakening the ballast, or cause the spike to work loose in its hole in the tie. The loosening of the spike from the tie destroys the advantage of a firm and secure connection sought to be obtained by its use. Moreover, it allows moisture to enter the spike hole and thus counteracts in part the effect of the chemical treatment to which softwood ties, now in increasing use because of the growing scarcity of hardwood ties, are subjected in order to render them capable of resisting the elements. To avoid these difficulties, the more recent devices are so constructed as to keep a firm grip on the rail, even as against a reverse movement, and by supplanting bolts with wedges or springs, and by dispensing with spikes by abutting instead of fastening the anchor to the tie, the maintenance inspection is reduced to a minimum.
In 1900, the date of the Pope patent, the art was in a crude and undeveloped state. In his specifications, two devices are described. In Figure 1, a crossbar 6 extends transversely under the rail at a point between two ties. Rigid abutments thereon, 7 and 8, bear against the edge of the flange on either side. One end of the bar is provided with a lug which fits into a notch on the bar J¡.; this bar extends between two ties to which it is spiked. Figure 4, a modified form, is primarily involved in this suit. There the holding device 6, with its abutments 7 and 8, is made integral with the bar I¡.. When the device is hooked over the rail base, it is rocked horizontally in either direction until the rail is firmly clamped between the abutments, and then spiked to the tie.
Pope thus led the actual development of the rail anchor by not restricting the use of his anchor to the rail joint and by making the relation of the gripping jaws to each other rigid. Furthermore, while he *637may not have been aware of it, the cross bar of his device shown in Figure 4 would probably tend to rotate after it had been spiked to the tie'; there would seem to be a lateral skewing about the spike as a center when the longitudinal force acting upon the rail is communicated to the anchor. Pope’s device, while a distinct step forward, nevertheless had all the shortcomings of a spiked anchor; it could not adequately meet the situation arising from a reverse movement of the rail.
The defendant has urged strongly and at length the Edmonston patent, No. 141,478, issued in 1873 for a scaffold support, as fully anticipating the Pope and the other patents in suit. With this contention we cannot agree. Edmonston discloses the use of the shackle grip, an okl mechanical expedient. Its specific application is, however, entirely dif - ferent in a scaffold support and in a rail anchor. In the one, it is used to obtain a firm hold upon a stationary support, such as the pole or post of a scaffold, in order to sustain the ledgers and putlogs over which the platform is built; in the other, it aims to arrest and prevent the movement of a rail which cannot, by its own resistance, maintain its stationary position. Especially in view of the fact that for several decades’ after the patent to Edmonston, the most astute and thoroughly trained railway engineers in this country and abroad grappled without success with the problem of rail creeping, the contention of the defendant that Edmonston more than 40 years ago revealed to the public a device substantially similar, not only to the primitive Pope, but to the Eaas and Sponenburg, the Vaughan, and the Bodkin anchors, is wanting in merit.
Tarrant patent, No. 492,446, and Noonan patent, No. 349,447, arc also strongly relied upon as anticipations. The Tarrant patent is for a combination railway tie and chair. The chair has a holder extending transversely under the rail and rigid jaws or abutments thereon engaging the rail flange. The chair is rotated on its pivot in the tie to produce the desired engagement and is then secured in place by a bolt 29 inserted in each side of the tie at opposite ends. While one of the purposes of the structure is to prevent rail creeping, Pope differs from Tarrant in this: The Tarrant chair must set upon a tie, specially designed to receive it, while Pope’s device is constructed to set between the ties; the latter is more easily constructed, more readily applied to the rail, and more economically maintained in proper condition. For these reasons, the progress made by Pope represents a patentable advance over Tarrant’s combination tic and chair.
Nor does the Noonan patent, No. 349,447, for a railway track system, satisfactorily rebut the novelty of the Pope claims. This cumbersome device has a pair of blocks H, H', each of which is bolted to two adjacent ties; a grip, resting upon these blocks and extending transversely under the rail, comprised a metallic plate F bent downwardly in its center to form a recess in which the rail fits. The opposite end parts †, f, of the grip lie loosely in transverse recesses of the blocks H, H'. A block on one side of the rail is provided with a threaded bolt which bears against the edge of one flange of the gripping plate; the other block is provided with a pin and spring arrangement whereby the pin bears against the opposite flange of the plate. The recess of the grip F in which the rail rests is slightly wider than the rail base so that *638when the pin is forced inwardly the grip P will be canted or inclined horizontally to cause the diagonally opposite edges 1, 2 of the grip to bite on the opposite sides of the rail thus preventing a backward while allowing a forward movement of the rail. Although the elements of this device come close to. those of Pope, the structure is based upon a different conception and designed to achieve a different result. By the grip arrangement, Noonan sought, not to prevent the movement of rails in the direction of the traffic, which he sought to absorb in another way, but to lock the rail against any backward movement. In his system, the grip is normally loose, offering no obstruction to the free forward movement of the rail; in Pope, it is normally fast and unyielding. Noonan’s grip, moreover, must rest upon two bars, one on each side of the rail and running parallel with it, which bars span the space between two successive ties to which they are spiked. ■ In Pope and the later anchors, one connection with the tie and on one side of the rail suffices.
. Brown patent, No. 630,444, for a rail joint,- likewise fails to anticipate Pope. This device fills in the entire space between two ties. It has two jaws gripping the opposite edges of the base flange of the rail and connected thereunder by means of a bolt which runs transversely beneath the rail base. iVssuming that this bulky contrivance can be applied at any point along the rail and not merely at the joint, it fails to anticipate Pope, because the jaws are not rigid or integral with respect to the bar. extending beneath the rail base, but are held together by nut and bolt.
Schauman patent, No. 328,616, Trude patent, No. 329,867, and Griffin patent, No. 556,667, are for tie plates, the base of which is designed to form a protecting plate upon which the rail rests. The overlapping lips or jaws secure the rails to the plates without the use of spikes for this purpose and prevent the lateral displacement or buckling of the rails. When installed the devices are rotated until the lips overlap the rail, and then are firmly spiked to the tie on both sides of the rail. While the spiking aims to prevent a creeping of the plate on the tie, there is no suggestion that these devices are capable of preventing or retarding the longitudinal movement of the rail itself.
We conclude that the Pope patent is not anticipated and shows patentable novelty.
Infringement of claims 1 and 2 is clear. While the defendant’s spring lugs are so located with reference to the rail flange that they must engage its upper surface rather than its edges, and while this is admittedly an improvement upon the device shown in Pope’s drawings, this modification must be deemed to be within the fair range of equivalents to which Pope’s claims are entitled in view of the prior art. Furthermore, Pope’s “rigid” abutment of claim 1 does not exclude defendant’s spring metal used to accomplish the same function; and defendant’s flange is none the less a connection because integral with the rest of the device.
In the Laas and Sponenburg anchor, under patent No. 720,362, a metal bar a, placed flatwise, passes transversely under the rail and is formed at one end with a jaw which grips one edge of the rail flange. The rail flange on the opposite side is engaged by a detachable jaw c, *639which is bolted to the bar a. To hold the bar laterally on the rail R, a shoulder d abuts directly against the opposite edge of the rail base. This shoulder is formed, integral either with the bar a or with the detachable jaw. The bar a has a flange g depending therefrom at one side and disposed to bear upon the tie C. The specifications further provide:
“In order to allow the described rail-gripping jaws to increase their hold on the rail in proportion to the resistance required to prevent the rail from creeping, we form the flange ff at an angle in relation to the length of the bar a, so as to cause the opposite end portion of the bar to be normally out of contact with the tie, and thus yield to some degree to the longitudinal strain of the rail, and by said yield of the bar the jaws b e thereof are caused to more effectually pinch the rail R.”
In view of Sponenburg’s patent, No. 668,423, and Brown’s patent, No. 630,444, Laas and Sponenburg cannot claim the broad conception of an anchor self-sustaining on the rail base and without tie engagement. Their invention is found in the construction of such an anchor having the tie abutting face on one side, while the opposite end remains normally out of contact with the tie. This construction enables the device to rotate on the rail and thus to strengthen automatically the grip of the anchor upon the rail in relation to its longitudinal movement. The prior art shows no anticipation.
Fdmonston has been fully considered in discussing Pope.
Noonan merely provides a lock against a reverse movement of the rail, and is not calculated or designed directly to prevent the longitudinal movement of the rail in the direction of the dominant traffic. Even if Noonan’s device were reversed, the rotary action or shackle grip would be obtained by means of a metal plate that engages the rail and rests upon blocks which must be spiked or bolted to the ties; Baas and Sponenburg’s forward step was in obtaining an effective shackle grip action on a rail anchor without spiking or bolting it to the tie. It is clearly a substantial and patentable advance over the primitive Noonan device.
The defendant relies upon Figure 7 of Brown patent, No. 630,444, as showing a shackle grip anchor that is not spiked or bolted to the tie. .Figure 7 is a modification of a structure described at length in the specifications. The principal embodiment of Brown’s patent discloses a device filling up the entire space between two ties, and broad flat abutments extending the entire length of the anchor are in contact with both ties at both sides of the rail. In the modification shown in Figure 7, a section of the casting is left off, so that one of the depending flanges bears upon a tie on one side of the longitudinal axis of the rail and the other flange bears upon another tie on the other side of the longitudinal axis of the rail. While the specification suggests no reason for the modification, the only apparent object was economy of metal. The modified form, like the principal embodiment, fits snugly between the two ties, and allows no play for a rocking or rotating movement that would affect in any substantial degree the grip upon the rail. Moreover, Brown’s device is a rail joint; it was not designed to be used at any other point, and because of its size and form it could not practically and commercially be applied at several points along a single rail.
*640If Pope’s anchor as represented in Figure 4 is rotated to obtain a firm grip before it is spiked to the tie, it may have a slight leverage after it is fastened to the tie, resulting from its tendencjr to rotate upon the spike. But Pope’s leverage is not only less effective than Laas and Sponenburg’s; it is obtained by virtue of the attachment of the device to the tie — that is, at a cost and a sacrifice which the teachings of the art oppose, and the avoidance of which constitutes Laas and Sponen-burg’s merit. Their device, under a reverse movement, cannot injure the tie or become loose; the flange will merely tend to withdraw from ■contact with the tie.
Sponenburg, in his patent No. 668,423, shows no grasp of the shackle grip conception; in Figure 2 the tie abutting face extends beyond the longitudinal axis of the rail, so that there is only the slightest leverage, a practically negligible tendency, in the device to skew laterally.
The Bodkin structure responds fully to claim 3 in suit; that Laas and Sponenburg’s drawings and*specifications show a detachable jaw, does not limit this broader claim to the preferred construction; defendant’s integral jaw does not enable it to avoid the charge of infringement.
The device under the Vaughan patent, No. 1,021,387, consists of a spring steel bar or yoke 5 extending across and under the rail base and a malleable iron shoe 11. The yoke is formed from a flat blank of sheet steel arranged with its side or flat faces in a substantially vertical position. On each end it has an upwardly extending portion. One of these 7 is hooked over one edge of the rail flange, the other 6 is straight and fits into a slot in the shoe, which engages the other edge of the base flange. Two vital features of the yoke are thus specified:
“It will be readily understood that in employing the flat bar 5 with its end portions 6 and 7 arranged with their side faces in vertical position, as shown in the drawings, that the strain upon the end portions 6 and 7, as well as the tension on the body of the bar 6, is in the direction of the flat faces of the parts of the bar in preventing the rail 3 from creeping, thus giving us the maximum strength for a given quantity of metal employed in the bar. This forms an important feature of our invention.”
“The lateral bend in the end portion 7 of the bar 5 * * * is provided for the purpose of stiffening that end of the bar.”
The yoke and the shoe are held together by a spring action. The bar 5 is made of spring steel, and when the parts are assembled in their proper positions, the spring action of the bar in tending to assume a position from which it was sprung by the torsional twist given to it in fitting it into the shoe—
“exerts a forward and a downward pressure upon the end portion 7 and the part 10, pressing the part 10 against the upper face of the base flange, and a reverse or rearward pressure upon the end portion 6 pressing the rearward face of the end portion 6 against the projection 16 and the forward face of the bar 5 against the rearward face 13 of the shoe.”
The result of this torsional twist or spring action is to cause the parts of the anticreeper to grip firmly the rail flange without any bolts. The shoe 11 has a tie-abutting face 11, about which the anchor will tend to rock, so that the tendency of the rail to creep will increase the wedging action. This shackle grip action, quite similar to that found in the Laas and Sponenburg device, is thus described:
*641“When the parts of the anticroepor are assembled upon the base flange 4 of the rail 3 with the abutments 12 of the shoe 11 engaging the cross-tie 2, as shown In Figs. 1, 2, and 3, the end of the bar 5 opposite to the shoe 11, is maintained in spaced relation to the cross-tie % by the shoe 11. Should the rail 3 attempt to creep forwardly or in the direction of the arrows shown in Figs. 1 and 2, forward movement oí the shoe 11 will be prevented by the cross-tie 2, and the rail in creeping forwardly will carry with it the end of the bar 5 opposite to the shoe 11; and immediately upon the initial creeping movement of the rail 3, the bar 5 will tend to assume a diagonal position beneath the base flange 4, thereby forcing Ihe inner edge 9 of the end portion 7 firmly into engagement with one side edge of the base flange 4; and at the same time forcing the inner edge 9 of the end portion 6 against the outer face of tile body of the shoe 11, and consequently forcing the shoe It against the other side edge of the base flange 4, thus firmly gripping the rail and preventing it from, creeping.”
At the time the Vaughan patent in suit was applied for, an anchor self-sustaining upon the rail without use of spike or bolt was not new in the art. Leighty (patent No. 809,193, 1906) and Lien (patent No. 816,926, 1906) sustained their devices upon the rail by a wedge arrangement; D. F. Vaughan (patent No. 897,037, 1908) and D. F. and D. L. Vaughan (patent No. 897,038, 1908), prior to their patent here in suit, had constructed an anchor sustainable on the rail by virtue of a spring action resulting from a torsional twist of the base of the yoke. All of these patents, moreover, have a shackle grip similar to that of the Laas and Sponenburg device. Even a lateral bend in the end portion hooking over the rail flange on the side not engaged by the shoe is disclosed in Figure 3 of D. F. and D. L. Vaughan patent, No. 8^7,038.
Vaughan’s advance over the prior art is in the use of flat sheet steel arranged edgewise — that is, substantially in a vertical position — -and the employment therewith of a lateral bend in one of its end portions which is likewise arranged edgewise. As the strain upon the end portions, as well as the tension on the body of the yoke, is in the direction of the flat faces of the bar, a maximum resisting strength for a given quantity of metal is thus attained. While in the Vaughan patent, No. 897,038, the lateral bend appears, it is not there used in combination with the flat vertical sides; so used it tends to stiffen the end of the yoke. These improvements justified the grant of letters patent, and defendant has paid them the tribute of copying them.
Tn defendant’s device, the anchor is made from a single piece of sheet steel. A flat strip or yoke 8 arranged edgewise, carrying at its ends hooks 1¡, and 5 which are adapted to overlie the outer edges of the rail base, passes under the rail. Extending from the yoke at an acute angle is a post or strut 6 which abuts against the tie & with its turned end 7. The parts are so shaped and proportioned that when the yoke is placed beneath the rail at right angles to the longitudinal axis of the rail, the hooks may be slipped over the base flange of the rail.
“The anchor is then,” quoting Iron? the specifications, “slipped along the rail until the post engages with the tie in advance of the anchor and then the end upon which the hook is carried is driven forward so as to bring the yoke at an angle to the longitudinal axis of the rail. This causes the hooks to take a firm grip upon the edges of the base flange of the rail, and any tendency of the rail to move forward causes the hook 4 to take a tighter grip, so that it will follow arty movement which the rail may have. Since the opposite end *642of the yoke is prevented from moving forward ¡by reason of tlie engagement of the post with the tie, a forward pull upon the hook 4 causes the yoke to assume a still greater inclination to the longitudinal axis of the rail and consequently produces a still tighter grip upon the rail.”
Bodkin further states in his specifications:
“In both of the forms which I have just described I prefer to make the angles of the under edges of the hooks somewhat smaller than the angles of the top of the base flanges of the rail; for example, where the angle of the base flange is 13 degrees, the angle of the under edges of the hooks may be made about 12 degrees. This enables me to obtain a very powerful wedging action as the anchors are swung around at an angle to the longitudinal axis of the rails and carry the hooks farther up' on the base flanges.”
From this description it may be seen that this anchor like Vaughan’s is free.of bolts and spikes and maintains a firm grip upon the rail even as against .a reverse movement; it, too, is so designed that, by reason of the shackle grip action, the greater the tendency to' creep, the firmer and more secure the engagement of the hooks with the rail flange becomes. It may be an improvement on Vaughan in being a one piece device;, but, as neither the language of Vaughan’s claim 2 nor tire prior art limit them to two pieces, and as Bodkin’s anchor possesses all the features and embodies the principles and conception for which the art is indebted to the Vaughans, it is none the less an infringement.
Assuming that the use of flat steel arranged edgewise is-of particular value only where the yoke is given a torsional twist and that the claim in suit should therefore be restricted to devices maintained on the rail by a spring action, the Bodkin anchor infringes; for as Vaughan testified without contradiction, when the end of the anchor spaced apart from the tie is driven toward the tie to take up the slack, the device is put under tension and the inwardly projecting lips are caused to ride upwardly on the rail flange, thereby rocking the lips in opposite directions and twisting the whole device.
The decree of the District Court will be affirmed.