In 1962, plaintiff invented a novel, sophisticated and valuable bed assembly which, by joining originally designed side rails at their centers with an interconnecting tension member, eliminated the need for wooden slats to support bed springs and solved significant problems associated with the use of conventional bed assemblies. His invention was clearly patentable. The questions presented are (1) whether claim 3 of his patent,1 which may be read to describe defendant's crude device, is valid; and (2) whether defendant's simple approach to the goal of slatless beds is sufficiently equivalent to the inventor's to constitute infringement of claim 4.2

A description of the principal problems solved by plaintiff's invention will identify the difference between the issues presented under claims 3 and 4.

One of the most significant problems confronting bed manufacturers and users has been the use of slats extending between, and supported on, the bed's side rails. Normally, three wooden slats were used. Quite often outward deflection of certain portions of the bed rails and downward deflection of the central portion of the slats would cause the slats to drop off the ledges and thus no longer support the bed springs. Sitting on the edge of a bed would tend to move all of the slats toward one bed rail, and would also deflect the bed rail outwardly, permitting the slats to drop to the floor.3

The problem associated with the use of wooden slats as thus described could have been eliminated by replacing the slats with a tension member interconnecting the central portion of conventional side rails, and thereby preventing outward deflection of the rails. The bed springs could then have rested securely on the horizontal flanges or ledges projecting inwardly from the side rails. Claim 3 of plaintiff's patent would have described an adequate solution to the industry's bed slat problems were it not for a complicating factor resulting from lack of standardization in the industry.

The bed rails are supported by pins which are set in notches in the end boards. Although the width of bed springs assemblies has been standardized for some time, the distance between the pins on headboards and footboards supporting the bed rails varies considerably from the standard spring widths. Thus, although a standard spring assembly for a double bed is 52½ inches wide, the slots in the headboard and footboard in which the supporting pins are placed may be as much as 54½ inches apart.

The discrepancy between the standard spring assembly widths and the variable distances between the notches in the end boards defeated pre-1962 attempts to produce a satisfactory slatless bed assembly. Without the wooden slats, the springs rested on the horizontal flanges or ledges projecting inwardly from the side rails. The metal ledges on conventional side rails were only about 1¼ inches wide. If the side rails were no further apart than the width of the springs, presumably such ledges would support the spring assembly, provided that a suitable "anti-spread device" prevented outward deflection of the rails when the bed was being moved or when someone was sitting on the middle of it. But the narrow ledges were not adequate to prevent the bedding from falling through to the floor when the notches in the end boards were farther apart than the width of the standard spring assemblies.

Plaintiff discovered an ingenious and by no means obvious solution to the problem. He designed new side rails which differed from the conventional in several respects. They contained a broader horizontal ledge which provides firm support for the spring assembly. More significantly, they were designed to provide a snug clamping relationship between the vertical portions of the two side rails and the spring assembly, with the side rails retaining their parallelism with each other and their firm support for the spring assembly, notwithstanding variations in the distance between headboard notches.

The patented rail has both a horizontal and a vertical flange over the major portion of its length, but near its ends the horizontal flange flows downwardly into the plane of the vertical flange. In consequence, when the rails are affixed to the headboards and footboards, and interconnected by a centrally located tension member, deflection may occur at the ends of the rails while the major portions of the rails maintain their parallel relationship with each other. As stated in the patent specifications, the major portions of the rails — i. e., the portions with both horizontal and vertical flanges — "will not flex"4 when interconnected by a tension member, whereas the purely vertical end portions "* * * may be deflected as much as one inch thus accommodating headboards having variations of two inches more than or less than the standard width of the spring assemblies."5

The specifications explain some of the advantages of the parallelism thus achieved. The spring assembly actually becomes an integral part of the bed thereby precluding a substantial canting of the side rails in relation to the headboard and footboard.6 Furthermore, with the vertical flanges of the side rails disposed against the bedding, there is no space between the bedding and the side rails in which articles may become lost or deposited; this snug relationship also eliminates a spot in which dust normally accumulates.7 And, of course, the lateral flexibility at the ends of the rails achieved the goal of slatless beds despite varying distances between headboard notches.

After learning of the commercial success of plaintiff's slatless bed assemblies, the defendant, Harris-Hub Company, Inc., began to manufacture and sell the accused assembly. The accused device does nothing more than interconnect the centers of two conventional metal side rails with a strap that applies sufficient tension to bow the rails inwardly against the bedding.

Although a purely theoretical description of the geometry of the bed rails might suggest that the vertical flanges would make contact with the bed springs only at the apex of each curve, in fact, since the maximum deflection of about an inch on each side rail is such a small fraction of the length of the rail, there is contact between the spring assembly and the bows throughout the major portion of the length of the rails. Plaintiff contends that this difference between parallel rails with end portions that flex as much as one inch, and bowed rails which flex to the same extent, is not sufficient to excuse defendant's copying of his idea.

Ethical considerations provide persuasive support for plaintiff's position. But under controlling legal principles, they are subordinate to the question whether defendant has violated or merely exercised a right protected by federal patent law. Cf., Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225, 84 S. Ct. 784, 11 L. Ed. 2d 661. As already noted, the questions involve (a) the validity of claim 3 and (b) the alleged equivalence of claim 4 and defendant's bow.

The district court 311 F. Supp. 734, found claim 3 invalid, and we have no doubt that he was correct. In construing claim 3, we, of course, look at the language of the claim itself, not at the device manufactured by plaintiff which embodied features described in other claims.8 It is equally clear that plaintiff's commercial success with the assembly described in claim 4 has no bearing on the validity of claim 3.9

Plaintiff argues that the district court erroneously construed claim 3 as nothing more than an anti-spread device. As so construed, the claim is plainly anticipated by a 1933 patent10 covering an anti-spread device for beds consisting of an adjustable strap interconnecting the central portions of the side rails. Moreover, the language of the claim itself describes the purpose of the interconnecting tension member as "preventing outward deflection of the central portion of the rails."11 The specifications explain the importance of preventing outward deflection.12 Thus, there is ample support for the district court's interpretation of claim 3 in the language of the patent.

Plaintiff argues, however, that claim 3 should be read as also describing an inward deflection of the rails.13 But there is no reference to an inward deflection in the claim. Plaintiff seeks to equate the claim's reference to "maintaining" the relationship between the rails and the bedding with his argument "that the cross-member will pull the side rails into and underneath the box springs." But his argument relates to the favorable pulling action of the actual device as described in claim 4, rather than to a fair interpretation of the word "maintaining" in claim 3, which, as there used, relates to the avoidance of a spreading action. The claim does not use the verb "pull"; the verb "maintain," which it does use, does not describe active movement.

Neither claim 3 nor anything else in the patent suggests that the inventor believed conventional rails could be used to solve the slatless bed problem resulting from disparities between the standard width of springs and the varying distances between end board notches. The use of conventional rails with a simple anti-spread device would have been acceptable if no such disparity existed. It is thus reasonable to interpret claim 3 as describing an anti-spread device which would be an adequate substitute for wooden slats in beds using headboards with notch separations of the same distance as the width of standard springs. The problems solved by other claims in the patent would not be present with respect to such beds. And nothing in the patent suggests that claim 3 would be an adequate solution to the problems described in the specifications except in such cases. We, therefore, construe claim 3 as an anti-spread device contemplating a slatless assembly for beds with end board notch separations equal to the width of the springs. As such, it was plainly invalid.

Apart from the question of anticipation, we would have grave doubt about the validity of plaintiff's patent if his idea embodied nothing more than a tension member connecting two ordinary metal side rails. Cf., Great A. & P. Tea Co. v. Supermarket Corp., 340 U.S. 147, 152-153, 71 S. Ct. 127, 95 L. Ed. 162; Anderson's Black Rock v. Pavement Salvage Co., 396 U.S. 57, 61, 90 S. Ct. 305, 24 L. Ed. 2d 258.

Almost the entire text of claim 4 is an explanation of the special design of plaintiff's side rails.14 Claim 4 clearly does not describe defendant's device, which uses conventional rails. Plaintiff argues, however, that infringement is established by the doctrine of equivalence. Relying on Graver Tank & Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 607-609, 70 S. Ct. 854, 94 L. Ed. 1097, he contends that the accused device performs "substantially the same work in substantially the same way to obtain substantially the same result" as the conception described in claim 4.15

Plaintiff submitted proposed findings of fact on the issue of equivalence which the district court rejected. As the Supreme Court noted in Graver, a finding of equivalence is a determination of fact with respect to which the trial court's decision should not be disturbed unless clearly erroneous. 339 U.S. at 609-610, 70 S. Ct. 854. We are not persuaded that the district court's rejection of plaintiff's proposed findings on the issue of equivalence was clearly erroneous.

The text of the patent indicates that one of the inventor's assumptions, which required him to design a novel rail with its horizontal flange flaring into the vertical at its ends, was that the "right angular cross-sectional configuration will not flex."16 If flexing a conventional rail into a bow had been within his conception, and if that configuration produced a result which is truly an equivalent of his invention, he took unnecessary pains to develop his sophisticated rail design. We think a fair reading of the entire patent indicates that his contribution to the art was predicated on an assumption that the portion of the rail which provides both clamping pressure from the vertical flange and adequate support from the horizontal flange "will not flex."

We are also persuaded that plaintiff cannot deny that the true parallelism of his device produces a result sufficiently superior to the bowing of conventional rails to preclude application of the doctrine of equivalence. The benefits of parallelism are described in the specifications and, in large part, constitute the objective achieved by the special design of the rails. As we consider the rail design a crucial feature of plaintiff's patented invention, we are unwilling to extend the scope of plaintiff's monopoly to include all slatless bed assemblies using a tension member to interconnect conventional side rails. Plaintiff has no patent on the function performed by his invention.17 The substance of plaintiff's invention includes an original design for side rails.

We read the language which plaintiff used in his patent as foreclosing his claim that tension sufficiently strong to bow conventional rails is the equivalent of his unique side rail design. He may not thus demean his own ingenious concept.

The judgment is affirmed.