This is a bill in equity for the infringement of letters patent No. 541,320, granted June 18, 1895, to the complainant, Samuel B. Donchian, for improvements in rug fasteners. The only claim is as follows:
“In combination in a carpet fastener, a stud section provided with means for attachment to a carpet, the concave under surface of said section having a projecting stud, the stud having a locking groove, a base section shaped to snugly fit the concaved under surface of the upper section and having a central opening, the walls of which closely fit about the stud at both its ends, and having a recess for a split ring within the base section, and the split ring loosely fitted within the recess and adapted to engage the stud at the grove, all substantially as described.”
The object of the invention is stated in the specification as follows :
“My invention relates to the class of devices designed for use in securing rug, carpet, or like article to a floor, and the object of my invention is to provide a device of this class .that shall be thin and compact, so as not to cause an unevenness on the rug by reason of a hump at the place where the fastening device is used, and that shall at the same time securely hold the rug in place; and a further object is to provide a device of this class that shall be strong enough to withstand the strains which shall be brought upon it under the peculiar conditions of use.”
In pursuance of this general object, the device consists of two sections, one a dome-shaped rug section having a concaved under surface, and adapted to be stitched to the under side of the rug; from this concaved under side of the rug section extends a grooved stud. The other section is a base section, nailed to the floor by slender nails. This floor section is fully described by the expert:
“This floor section has a central opening which will receive the stud. About this opening is a portion shaped to provide a recess, and in this recess is a split spring ring which is slightly smaller than the stud. When the stud is thrust in*891to the opening the ring grasps the reduced portion 'of the stud and prevents its withdrawal until sufficient force is applied to cause the ring to expand and let the stud out. The opening through the floor section is of such size and shape that as the spring yields under transverse strains the stud engages the solid walls of the floor section and is held by them rather than by the spring. And as the floor’section fits within the concavity of the rug section the wall of the rug section comes into contact with the edge of the floor section and assists the stud in sustaining transverse strains. The concavity of the rug section not only provides a receptacle for the floor section, but also produces a dome-shaped structure, which, while low, is very stiff and strong, so that it will not crush when stepped upon, nor produce a hump in the rug which is perceptible to the eye or to the foot. The Donchian specification points out that the holes made by the small wire nails used to fasten the floor section can be filled with wax if. the section is removed, and that the threads which are used to fasten the rug section project so as to prevent any marring of the floor by that section.”
The testimony tends to show that previous to the production of this fastener there had been several methods of temporarily fastening a rug to a floor. One of these was by the so-called “pin and socket” method. This method involved a hole bored in the floor, into which was forced a metal sleeve or socket. The rug was punctured above the hole. A pin was thrust through the rug into the hole, the pin having a flat head, which overlaid the rug and held it in position. Another mode of fastening was by the “peg and grommet.” By this method the rug was punctured, and the grommet or eyelet inserted in the puncture. This was then slipped over the projecting head of a peg which had been screwed or driven into the floor. Another method of fastening was by using a fastening screw passing through the rug and threaded into the floor, or a socket in the floor, the screw having a head projecting above the rug, by which it could be grasped for turning. The above methods were unpatented. There were also two methods provided in the patented art — one by an old patent to Culver in 1851, which provided for a metallic plate attached to the under side of a carpet having a hole which was slipped over the projecting head of a wood screw partially screwed into the floor. Another method was provided later by the Hellmuth patent, which was a modification of the “peg and grommet” fastener. All these methods of fastening involved the mutilation of either the floor or the rug, and in most cases of both.
With reference to the state of public knowledge at the time of the Donchian invention, the evidence shows that ball and socket fasteners were in common use in the fastening of gloves and in many other fastenings. But the ball and socket fasteners, as applied to gloves, were, when screwed in the floor, crude, unsatisfactory, and easily displaced by pressure. The advantages of the invention are fully stated in the specification quoted and in the evidence of the expert, to which reference has been made.
The defense is noninfringement. In seeking to find what the inventive thought of the patentee was, and what construction should be given to the claim in the patent, aid is derived from following the course of the invention through the Patent Office. In his application the patentee did not claim that there was anything new in adapting the ball and socket fastener to a rug. He set forth his *892general purpose in the beginning of his specification which has been quoted. He went on further to say in his specification:
“My invention consists in the details of the several parts making up the device as a whole and in the combination of such parts as more particularly hereinafter described and pointed out in the claim.”
In describing his device more fully, he said:
“A central opening, 11, extends through the base section and is formed of a size to closely fit the stud, 5, on opposite sides of the recess, 9. The base section, 7, is made to conform to the shape of the concavity in the stud section so as to lie closely therein, and a flat spring is preferably used in the recess, 9, as greater strength can be secured in this form of spring with a given thickness; and this construction of the spring, of the concaved stud section, and the base section closely fitting therein provides an extremely thin device, while at the same time possessing the necessary holding qualities. The split ring, 10, is preferably formed of a size to loosely fit within the recess, 9. The base section is secured to the floor as by means of very small wire nails or the like, and this does not materially damage the floor, as when they are removed the comparatively small holes can be easily filled with wax and polished over, and the thread used to secure the plate or stud section to the rug, as shown in the drawings, projects on the under surface of the plate, and prevents any marring of the floor by the slight rubbing of the plate in use. An important advantage of my improved device resides in the construction of the stud to closely fit the central opening through the base section. This affords means whereby the stud is securely held against afiy lateral strain which would tend to loosen it from the grasp of the spring, these strains being continuously exerted in the peculiar use of the device as by walking on the rug.”
An examination of the file wrapper shows that in his original application to the Patent Office he concluded as follows:
“I am aware that various devices have been used in which a stud from one part has been frietionally held within an opposite part, and I do not broadly claim such a .construction. Such devices, however, have been used in an entirely different manner from the use to which my device is put, and in this latter use would become inoperative from the fact that they would be destroyed by the pressure which would be brought to bear upon them as by being stepped on, and my improved device possesses the advantages over such devices in that it is practically indestructible in the peculiar use to which it is put, and is so compact that an unevenness of the rug or carpet is not caused by its use.”
In that application he made three claims, as follows:
“(1) In combination in a rug fastener or the like, a stud section provided with means for attachment to a rug concaved on its under surface and having a stud projecting therefrom, a base section fitting the concavity of the stud section and having a central opening therethrough for the reception of the stud, a recess within the base section, and a split ring adapted to grasp the stud at the groove therein, all substantially as described.
“(2) In a rug fastener or the like, in combination, a stud section concaved on its under portion and provided with means for fastening to a rug or the like, a base section fitting the concavity of the stud section, and having a central opening closely fitting about the stud on the stud section, a recess, and a split ring loosely fitting in the recess to securely grasp the stud at the groove therein, all substantially as described.
“(3) In combination with a rug or the like, a rug fastener consisting of a stud section secured thereto and provided with a concavity on its under surface, a stud extending from the stud section, and having a groove extending around the stud, a base section fitting the concavity of the stud section and with a central opening closely fitting the stud, a recess in the stud section, and a split ring loosely fitting within the recess and firmly grasping the stud at the groove; all substantially as described.”
*893It thus appears that in the first claim he described in a “rug fastener or the like” a base section fitting the concavity of the stud section, but did not say anything about its fitting snugly or closely. In claim 3, however, he added the additional limitation of a base section fitting the concavity of the stud section, and with a central opening closely fitting the stud, but said nothing about the central opening fitting the stud closely at both its ends. These three claims, as filed, the Patent Office at once rejected, saying:
“The claims of this application seem to be all of substantially the same scope, and they are objected to for this reason.
“The claims seem to cover merely a double use of the fasteners for gloves covered by patents cited as follows:
“Richardson, Mar. 5, 1889, 399,161.) Clasps &
“Kreatzer, Mar. 22, 1887, 339,015. !■ Buckles
“Kershaw, Dec. 2, 1890, 442.056. ) Glove Fasteners.
“See, also, patent of Culver, Nov. 17, 1857, 18,631 (Builders Hardware— Carpet Fasteners).
“The claims are rejected in view of the prior art disclosed in the references cited above.”
Whereupon the patentée canceled the larger part of the closing paragraph of his specification, which took the ground that prior ball and socket fasteners had not been used in connection with rugs. He substituted the following:
“In none of these prior devices has provision been made for the close fitting of the engaging parts and for a secure holding of the stud against a strain tending to upset it. A material feature of my invention resides in providing the base section with a stud socket above and below the spring-member whereby any rocking actibn of one part on the other is effectually prevented, and any strain transversely of the stud prevented from prying or tilting the latter out of engagement.”
He pointed out specific features of difference between his patent and the patent cited against him in the prior art. In the above substitution in the specification he introduced for the first time the fit of the base section upon the stud section above and below the ring, and this he claimed to be a material feature of his invention. He struck out the three, claims, and substituted the following two claims:
“(1) In combination in a sectional fastening device a stud section provided with attaching means and having a stud projecting' therefrom, the stud having a locking recess, a base section having a central opening for the reception of the stud, a recess in the base section for holding a split ring, the walls of the socket embracing the stud both above and below the split ring, and the ■ split ring located in the socket and adapted to grasp the stud at the groove therein; all substantially as described.
“(2) In combination in a carpet fastener, a stud section provided with means for attachment to a carpet, the concave under surface of said section having a projecting stud, the stud having a locking groove, a base section shaped to snugly fit the concaved under surface of the upper section and having a central opening, the walls of which closely fit about the stud at both its ends, and having a recess for a split ring within the base section, and the split ring loosely fitted within the recess and adapted to engage the stud at the groove; All substantially as described.”
An examination of these claims shows that the patentee had abandoned the limitation to a rug fastener; that he had set forth the material feature of the invention in claim 1 as “the walls of the socket *894embracing the stud both above and below the split ring”; and in claim 2 as “the walls of which closely fit about the stud at both its ends.” It will be seen further that in claim 1 he made no restriction as to the fitting within the concavity of the lower section, but in claim 2 he made this limitation: “A base section shaped to snugly fit the concaved under surface of the upper section.”
In discussing the Culver patent, he referred to the closely fitting element in his patent as being a limiting element and an important feature of the claims as he then submitted them. He concluded thus: .
“It is extremely important in the construction of a fastener of the form devised by applicant that the socket piece should have the closed bearing for the stud at the opposite extremities of the latter, as has been described and set out in the amended claims. Without the peculiar form of close-fitting parts and the special arrangement of the holding ring in a recess midway of the stud, the transverse strains upon the fastener would disengage the parts, and make the device practically inoperative. It is held that applicant has now clearly set out and distinguished the peculiar features of invention, and a reconsideration and allowance of the case is requested.”
After the filing of this amendment the Patent Office again rejected the claims, the examiner giving the following grounds for rej ection:
“It is found that it is.old to provide a catch plate for a stud to engage the stud both above and below an annular groove in the stud. ■ See the following patents:
“Cushman, Aug. 12, 1890, 434,177,
“Drake, Sept. 23, 1890, 437,103 (both in Clasps, Buckles & Buttons — Separable). It seems that to utilize this expedient in a carpet fastener instead of in clasps for other purposes is a colorable change.”
The complainant then canceled claim 1, and amended claim 2, restricting it to a carpet fastener. The effect of this amendment was to eliminate the breadth of claim 1 and to confine the invention to use as a carpet fastener, and restrict it to a combination having in one structure two elements, namely, “a base section shaped to snugly fit the concaved under surface of the upper section,” and the feature “that the walls of the base section shall closely fit about the stud at both its ends.” These two elements were in combination with the split-ring fastener. The argument of the patentee in reference to the Drake patent is important as bearing upon the construction which he then gave to his claim. He said :
“The latter patent does not show a fastener section having a concaved un- ■ der surface which is a limitation of the claim as now presented. This con-caved lower surface is of importance in giving to the fastener as a whole a stability not possessed by the fastener of the Drake device. The socket in the Drake fastener does not embrace the stud, but contains an intermediate spring. The claim now presented in the case recites a central opening ‘the walls of which closely fit about the stud at both its ends.’ This feature is not found in the Drake patent.”
In the conclusion of his argument he distinctly stated what he regarded as new in his invention:
“The fastener is new in the provision for the close fit of the lower section in the concaved under surface of the upper. It is new in the direct contact of the stud and the walls of the socket when a split ring forms the holding *895■device. The claim recites the limiting and distinguishing features, and it is submitted that on the question of sufficiency of invention the applicant should be given the benefit of any doubt and be allowed to reap the benefit of his invention to that extent which the production of the novel article which fills, and has immediately supplied, a large demand entitles him to.”
The court cannot escape the conclusion that the claim of the patentee as amended embraces two distinct and material features, namely, a base section shaped to snugly fit the concaved under surface of the upper section, and a central opening in the base-section, the walls of which fit closely about the stud at both its ends. The inventive thought upon which the patentee chose to rest his patent was a carpet fastener having these two distinct, material, and closely defined elements. He invented this construction because it was a compact, solid, firm construction which would not yield to pressure. He preferred and patented this close, compact method of resisting pressure. He confined himself to this limited construction. A careful examination of the claim itself as it finally appears in the patent is sufficient to bring the court to this conelusion. But a court may often find, and does find in this case, assistance from the arguments of the patentee himself when obtaining his patent, and from the amendments which he purposely and understandingly accepted. In Ball & Socket Fastener Co. v. Ball Glove Fastener Co., 58 Fed. 818, 824, 7 C. C. A. 498, 504, Judge Putnam says:-
“The rule touching the effect of such amendments has been several times laid down by the Supreme Court in patent causes, although it is only a peculiar application of the general principles of law relative to the interpretation of instruments. In the ease at bar the amendments relate to the very pith and marrow of the alleged improvement, touch directly the question of novelty, and were understandingly and deliberately assented to.”
In the case now before us the amendments made by the patentee relate to the “pith and marrow” of the invention. It would be unjust to the public, and to all the parties involved in the construction of the patent, if a patentee were allowed to “understandingly and deliberately” limit the scope of his patent while he is obtaining it, and were afterwards allowed to escape from his limitation when the patent is construed. He ought not to be heard to demand one rule of interpretation in the Patent Office and another in the courts. The ordinary principles relating to the interpretation of a contract are the principles which prevail in construing a patent. The understanding of parties to an agreement at the time it is made is always held to be of importance in the construction of such agreement. Courts often find aid in construing a contract by considering what the parties have said and what they have done when the contract was made. This court has had occasion to refer to the language of Ford Sugden in Attorney General v. Drummond, 1 Drury & Warren, “Tell me what you have done under a deed, and I will tell you what that deed means.” In Greene v. Buckley, 135 Fed. 531, Judge Townsend, speaking for the Circuit Court of Appeals in the Second Circuit, says:
“Where the patentee specifies a particular form as a means by which the effect of the invention is produced, or otherwise confines himself to a particu*896lar form of what he describes, he is limited thereby in his claim for infringement. Walker on Patents (4th Ed.) § 363, and cases cited. Where a patentee acquiesces in a rejection of claims, and amends the same so as to be more specific, such claims must be read and interpreted with reference to the rejected claims and to the prior state of the art, and cannot be so construed as to cover either what was rejected by the Patent Office or disclosed by prior devices. Knapp v. Morss, 150 U. S. 221, 225, 14 Sup. Ct. 81, 37 L. Ed. 1059; Koemer v. Peddie, 132 U. S. 313, 10 Sup. Ct. 98, 33 L. Ed. 382. ‘Where a patentee has modified his claim in obedience to the requirements of the Patent Office, he cannot have for it an extended construction which has been rejected by the Patent Office ; and in a suit on his patent his claim must be limited, where it is a combination of parts, to a combination of all the elements which he has included in his claim as necessarily constituting that combination.’ Phoenix Caster Co. v. Spiegel, 133 U. S. 360, 368, 10 Sup. Ct. 409, 33 L. Ed. 663. The words of limitation inserted in a claim must be construed in the light of the circumstances surrounding the issuance of the patent in order to prevent an undue broadening of the scope of the invention. Singer Mfg. Co. v. Cramer, 192 U. S. 265, 285, 24 Sup. Ct. 291, 48 L. Ed. 437. In interpreting such a patent, the admissions and declarations of the patentee in the Patent Office, and amendments made by him which relate to the essence of alleged improvements, and are directed to the question of invalidity, when understandingly and deliberately assented to, are binding upon him. Reece Button Hole Machine Co. v. Globe Button Hole Machine Co., 61 Fed. 958, 10 C. C. A. 194” — citing also other cases.
Does the defendants’ fastener infringe the claim of the patent in suit? It is undoubtedly in some respects similar to the fastener of the patent. Each has a stud section with means for attachment to the floor. Each has a projecting stud provided with a.groove. Each has a base section provided with a central opening. Each has a recess in the central opening provided with a split ring loosely fitted and adapted to engage the stud. The claim of the patent provides that the stud section shall be fastened to the rug and the socket section to the floor. The defendants’ device is arranged in the opposite way.- But one way may be considered the immaterial reversal of the other. I have already found, however, that the claim of the patent presents two material features: First, the close fit of the base section within the concavity of the upper section; and, second, the close fitting of the aperture of the base section about both ends of the stud. The inventive thought of the patentee made these two elements important and decisive. His inventive idea was to resist pressure and strain by the closeness of fit in respect to these two elements. The defendants’ device does not, in my opinion, infringe these two elements. The constructive idea of the defendants is distinctly and essentially different from the inventive idea of the patentee. The defendants’ device does not have the close fitting of one section within the concavity of the other. It is designed to allow the rocking or tipping of one part on the other, whereby undue strain of the parts is sought to be prevented. When in place, all pressure is transmitted through the rigid stud to the floor. The stud section thus resists the strain, and the other section is not required to be of close, solid construction. ' The close fitting of the aperture of the base section about both ends of the stud was made in the specification “a material feature” of the patentee’s invention. His claim provides that the walls must closely fit about the stud at both its ends; and he insisted that'an old pat*897ent, which had an intermediate spring functionally like the split ring of his patent, did not meet the terms of his claim. But in the device of the defendants the stud section can touch the socket section only through the medium of the split ring. It cannot be held to present a close fit of one section upon the other. In reference to the shape of the stud the defendant Kingston says:
“In the construction of the shape of the stud, I concaved it so that in use on the different kinds of velvet carpeting it would, by its shape and friction, on the gradual curve of the stud towards the top. act as a lever when the various strains would be brought to bear upon this particular part.”
The defendants do not attempt to form their construction upon the basis of a “locking groove,” but rather upon a gradual curve of the stud, and in general upon a loose mechanical construction. Some features of this construction are found in certain old patents. In speaking of the action of the socket section upon the stud, their expert says:
“This action resuits directly from the special construction of stud with a reduced shank portion instead of a locking groove, the loose fit of the ring about the stud, and the extremely loose fit of the stud in the aperture, which permits the socket section to rock to its furthest extent without bringing the walls of the aperture into the engagement with the stud.”
The construction of the stud in defendants’ fastener is designed to permit the free rocking or tilting of the socket section with relation to the stud section, instead of presenting the close fit called for by the claim of the patent.
The court is of the opinion that the loosely fitting arrangement of elements employed by the defendants cannot be held to be an equivalent of the close, solid construction of the complainant. In my opinion, the defendants’ device is based upon a constructive thought different from the inventive idea of the patent, and cannot be held to be an infringement of it.
Bill to be dismissed, with costs.