ITT Corp. v. United States

Related Cases

    OPINION

    NETTESHEIM, Judge.

    Plaintiff ITT Corporation (“plaintiff”) sued for recovery pursuant to 28 U.S.C. § 1498 (1982), for reasonable and entire compensation for the unlicensed use and manufacture by or for the United States of the inventions described and claimed in United States Patent No. 8,936,145 entitled “Fiber Optic Alignment Sleeve” (the “ ’145 patent”); United States Patent No. 3,947,-182, entitled “Fiber Optic Connector with Axial Tolerance Relief” (the “’182 patent”), and United States Patent No. 4,047,-797, entitled “Fiber Optic Connector” (the “’797 patent”). The ’145 patent was issued on February 3, 1976, to plaintiff as assignee of inventor Ronald L. McCartney; the ’182 patent was issued on March 30, 1976, again to plaintiff as assignee of Mr. McCartney; and the '797 patent was issued on September 13, 1977, to plaintiff as as-signee of inventors Bruce K. Arnold and Elias A. Moreno.

    Passively involved is Hughes Aircraft Company (“Hughes”), which took no role in the conduct of this litigation other than to enter into a protective order with respect to discovery. Plaintiff charges that Hughes manufactures and sells to the United States Army, Navy, and Air Force fiber optic connectors covered by each of three patents in suit.

    The issues of validity and infringement of specified claims of each patent in suit have been tried. Any accounting was deferred for later proceedings after determination of liability. At the outset of trial, plaintiff restricted its proof to two claims each of the ’182 and ’145 patents and four claims of the ’797 patent. Its remaining claims were withdrawn, and the complaint is deemed dismissed with prejudice with respect to past procurements involving those claims.

    The facts found by the the court relating to background are presented first, followed by a discussion that separately finds facts relating to the defense of invalidity and the charges of infringement.

    BACKGROUND

    I. Fiber optics

    The technology of this case is fiber optics, which involves the transmission of light through glass or plastic. Electrical technology, in contrast, deals with the transmission of electricity in wires or coaxial cable. A spirited tutorial in fiber optics was given by Leslie M. Borsuk, Director of ITT Cannon, plaintiff’s subsidiary involved in the design, development, and marketing of fiber optic connectors. Mr. Borsuk also served as plaintiff’s primary witness to read the claims of the subject patents on the accused Hughes devices and to rebut the evidence of invalidity. Notwithstanding Mr. Borsuk’s employment and role in presenting plaintiff’s case spark connotations of partisanship, his credibility could not be impugned. This witness demonstrated exceptional familiarity with the subject matter of the patents and ability to articulate it. Most of his opinions and other testimony were “sufficiently convine-*323ing.” See Ashland Oil, Inc. v. Delta Resins & Refractories, Inc., 776 F.2d 281, 294 (Fed.Cir.1985), cert. denied, — U.S.-, 106 S.Ct. 1201, 89 L.Ed.2d 315 (1986) (stating preference for disinterested experts).

    Mr. Borsuk explained that fiber optics was the generic term given to what in his view (shared by other of plaintiffs witnesses) was a new technology in the early 1970’s. The similarities between transmission of electricity and light, a proposition zealously contested by plaintiff, do exist and are instructive in understanding fiber optics. Fiber optics presented a promising means for transmitting signals through cable by converting electrical signals into optical energy and reconverting the light signals into electrical signals. Mr. Borsuk noted the following advantages of fiber optics over electricity: Noise in the area does not affect the signal strength of fiber optic cable. It has a larger band width or information-carrying capacity than electrical cable. Fiber optic cables are smaller and lighter than electric cables. Potentially lower costs are associated with fiber optics because glass is made from sand.

    An optical fiber is approximately 5/1000 inch in diameter, roughly the size of a human hair. At the inception of the technology, the cables in fiber optics included many individual fibers, or “bundle fibers,” due to the anticipated breakage of individual fibers. The same message was transmitted in each fiber in the bundle, so that each bundle of fibers would transmit only one data stream. Now, with advances in technology, single fibers are more prevalent.

    Optic theory treats how light within a fiber is propagated through the fiber by internal reflection. An optic fiber consists of a “core” surrounded by “cladding.” The core has a higher index of refraction than the cladding, which provides insulation. The light propagates in a zig-zag fashion down the core as it reflects off the cladding. The optic fiber itself is surrounded by a buffer, some form of “strength member” (usually fibrous), and finally encased in plastic jacketing. This assembly is the “fiber cable.”

    Along a transmission route, fiber cables and therefore optic fibers must be mated at junction points. Both electrical technology and fiber optic technology share the term “connector” to designate the hardware that implements a non-permanent coupling. (Splicing offers a permanent method.) Fiber optic energy can be lost, and transmission energy reduced, or attenuated, in the process of effecting this coupling because the index of refraction of the fiber core is not the same as air. “Fresnel loss” is the term descriptive of this phenomenon. The efficiency loss of the coupling is expressed as decibel (“dB”) loss of power. Today, 1 dB loss of power, or 20 percent, is an acceptable loss. Mating opposing fibers in the best of conditions thus implicates significant attenuation.

    There are three degrees of freedom that two opposing optical fibers can achieve as they are positioned for mating. The first is lateral alignment, which refers to alignment of fiber cores on the same axis, and it should be pointed out that cores can be off center with the fiber. The second is gap or separation loss—the gap between two opposing fibers. The third is angular alignment, which represents the angle present when two fibers are not on the same axis. Mr. Borsuk viewed these degrees of freedom as cumulative problems against which an acceptable connector must be engineered. He sponsored the opinion that plaintiff has done so in the ’182 and '145 patents by claiming devices for butting two fibers while allowing for axial tolerances and maintaining their alignment. The ’797 patent terminates the fiber cable to a transmitter or receiver.

    2. Development of the patented inventions

    Plaintiff presented two witnesses to explain how the subject inventions were made. Mr. McCartney, a physicist, worked for ITT Cannon from 1974-1980 and is the named inventor or co-inventor in 12 patents pertaining to fiber optics. He began working in fiber optics with ITT Cannon within two to three months after his arrival. Mr. McCartney said he was under instructions to investigate every idea, rational or irrational, for fiber optic connectors, and his team did so for three months.

    Eleanor L. Landgreen, a long-time employee of ITT Cannon with limited post-high school education, but a great deal of practical knowledge acquired in her advancement from draftsman to designer of electrical connectors, joined Mr. McCartney’s group in what she referred to as the “new technology of fiber optics.” Mrs. Landgreen used the phrase “blue sky” operation to suggest the process by which ideas were generated for the group’s pre*324mier task of gutting an electrical connector and fitting fiber optics into it for military-hardware. When asked to develop a connector for fiber optics, the original team was able to use external housings of electrical connectors, but both witnesses regarded the components as different due to the fragile nature of fiber optics. Mrs. Landgreen’s first project was to build fiber optic connectors for an A-Bomber, then to put fiber optics in as many military connectors as possible.

    Mr. McCartney took pains to distinguish fiber optics as an entirely different phenomenon from electrical conduction. In fact, as will be discussed, marked similarities exist between electrical and fiber optics technology. To begin with, the same people worked in both technologies. One does not disserve plaintiff’s inventions by observing the gradual evolution of fiber optics from electrical technology. Specifically, electrical connectors were looked to for solutions by everyone (according to the testimony) experimenting in fiber optics in the early 1970’s.

    The '182 patent, covering Mr. McCartney’s first invention in this litigation, was filed on July 21, 1975, as a continuation in part of an application filed on October 29, 1974, which had been abandoned. Mr. McCartney viewed axial tolerance as a major problem. The lexicon of fiber optics already includes “optic fiber,” “core,” “cladding,” and “fiber cable;” the “contact” surrounds the cable and consists of an assembly of other parts or one part. A fiber is terminated in a contact. Mr. McCartney observed that multiple contacts in a single connector would come out in different lengths due to vagaries of manufacture. As a result, some optic fibers crushed when pushed together. The solution he devised was to introduce a spring to each of the macing contacts to allow them to abut in a controlled fashion or, as he put it, a floating intermediate position. Mr. McCartney described his solution in an engineering notebook entry dated July 9, 1974:

    The use of connectors where axial movement to accomplish interconnection of optical fibers requires that each contact be individually suspended axially. To accomplish intimate butt contact and also allow for axial tolerance variations a spring loading of each contact is required. In addition a high force rate is needed due to small displacements along the axis during a mating. A means is also required whereby a bottoming out of the displacement is required. Inherent in the system must be a means whereby vibration control is provided which will eliminate high amplitude vibrations and control optical gaps and eliminate optical surface shattering.

    A solution to the above problems is provided by using a restrained O-ring which is required to climb a circular slope as the optical pin is displaced during mating.

    Mr. McCartney developed a contact in the form of a termination pin assembly that comes in mirrored parts for mating. Each assembly contains a termination pin that houses the fiber optic cable. Surrounding each pin is a slidable sleeve (which is not movable); a retention clip, also known as a casear clip; and an O-ring. As the pin slides within the slidable sleeve, the O-ring, which lies in a groove, provides a resilient force between the slidable sleeve and the outside pin diameter. The retention clip is not a fixed part of the assembly in the sense that an extraction tool works against it to release the termination pin assembly. When the assembly is inserted into the connector the retention clip is activated to hold the slidable sleeve, and hence the pin assembly, in place.

    Mr. McCartney also was responsible for the ’145 patent, filed November 7, 1974, which he described as a tolerance compensating sleeve for aligning two opposing fiber optic contacts. He confronted two problems in developing solutions that became the ’145 patent. One was to accomplish a close fit between the pin and socket contacts. He took a tube, slit it from end to end, and made the inside diameter of the sleeve smaller than the outside diameter of the pin contact. The second problem arose from the need to accommodate bundle fibers. In Mr. McCartney’s view, aligning bundle fibers was a more difficult task than single fibers, apart from considerations of size. Mr. McCartney was interested in connecting hexagons, or hexpacs, consisting of groups of fibers lying in the groove of two others. He devised an alignment sleeve, with partial slits, none of which ran from one end to the other or the sleeve would have fallen apart. The obtruding contacts were engaged by the partial slits (or slots), which were rifled or spiraled over the sleeve to engage each contact wherever each of the contacts ran*325domly encountered a slot opening. Rotational alignment was thus achieved.

    Mrs. Landgreen was familiar with both the inventions leading to the ’182 and ’145 patents and had retained several termination pin assemblies developed by the group during the early 1970’s. She drew an early fiber optic connector for ITT Cannon in July 1974 illustrating and labeling a fiber optic cable, a “relief” (or slidable) sleeve, a contact retention clip, and an elas-tomeric axial spring (a rubber device), as well as a split alignment sleeve to align different size bundles.

    Mrs. Landgreen offered insight into the development of the ’797 patent. She was involved with Mr. Arnold in drafting the concepts and making most of the drawings. The invention leading to the ’797 patent was a single fiber optic contact support yoke and retention assembly device. The theory behind this patent was to take the Kevlar (brand name) strength member in each cable, which did not have much elasticity; to strip the plastic coating; and to bend back the Kevlar fiber in order to allow more slack behind the termination pin assembly, so that any force pulling on the cable would pull on the Kevlar, not the glass fiber. The device was designed for six single fibers in Kevlar strength members terminating in six contacts all set in a yoke. The ’797 patent thus sought to provide strain relief by protecting the fibers against physical abuse when subject to pulling. An application was filed on June 9, 1976.

    II. Claims in Issue

    The following claims are in issue:

    The ’182 patent

    1. A termination pin assembly for adapting a fiber optic cable for insertion into a connector assembly, comprising:

    a hollow elongated pin body having an axial bore of such a diameter over a predetermined axial dimension from a first end of said pin body so as to accommodate insertion of the optical fiber of said cable through said bore for at least said predetermined axial dimension, said first end of said pin body being the terminal end thereof when said cable is inserted within said pin;
    a circumferential groove about a central portion of said pin body, said groove having first and second side walls, said first side wall being nearest said first end of said pin body;
    an axially slidable sleeve mounted to slide over at least a portion of the outside surface of said pin body extending from said groove away from said first end of said pin body; and
    resilient annular ring means installed within said groove, said ring means being compressively engaged between said slidable sleeve and said first side wall of said groove and providing a resilient force tending to resist axial movement of said pin body first end toward said slidable sleeve.

    2. In a fiber optic connector for providing a removable light signal connection between the ends of at least one pair of fiber optic cables, said pair consisting of first and second cables, each containing a fiber optic bundle comprising at least one light transmitting fiber, said connector having first and second mating connector shells and corresponding guide means for holding said first and second cables in substantially colinear positions with said ends of said first and second fiber bundles in substantial abutment when said shells are mated; the combination comprising:

    first and second elongated hollow cable terminating pins corresponding respectively to each of said first and second cables, said pins each having an axial cavity therethrough, having an inside diameter for a predetermined axial length from a first pin end for accommodating said fiber bundle of the corresponding cable inserted therein, said first pin end being substantially coextensive with said end of the corresponding cable;
    means for bonding said fiber bundle within at least a portion of said axial cavity to provide substantial resistance to axial movement of said fiber bundles within said hollow pins;
    means comprising a coaxial sleeve slid-ably fit about the outside diameter of a central portion of at least one of said first and second terminating pins;
    first means associated with the internal cavity of at least the corresponding one of said connector shells for restraining axial translation of said coaxial sleeve in a direction away from the point of abutment of said cable ends;
    and second means including a resilient member associated with both the body *326of said termination pin and said coaxial sleeve to permit relatively small axial translation of said termination pin within said sleeve thereby to provide a compression force at the point of said abutment of said first and second fiber optic bundles thereby also providing relief for axial manufacturing tolerances.

    The ’145 patent

    1. In a fiber optic connector for providing a removable light signal connection between the terminal ends of at least one pair of fiber optic cables, said pair including first and second cables each having a plurality of optical fibers and each including a termination pin of predetermined cross-sectional ends thereof, said connector further including first and second mating connector members adapted to contain at least said pair of cables and corresponding termination pins, the combination comprising:

    retaining means within said first and second connector members, respectively, for retaining said termination pins and therefore said terminal ends of said corresponding first and second fiber optic cables in substantial axial light-transferring abutment when said first and second connector members are mated; and
    means including an alignment sleeve of resilient material within said first and second connector members for effecting lateral alignment of said terminal ends of said termination pins and corresponding fiber optic cables, said sleeve including at least one axially extending slot through the wall of said sleeve over an axial dimension overlapping the point of said abutment, said sleeve being of an initial diameter over said axial dimension such that insertion of said termination pins produces resilient lateral expansion of said sleeve to produce inwardly directed radial forces tending to hold said terminal ends in lateral alignment. ******

    13. A fiber optic cable coupling assembly comprising:

    an alignment sleeve of resilient spring material;
    a pair of fiber optic cables each having a fiber bundle terminated by a termination pin;
    at least one axially extending slot through the wall of an axial section of said sleeve, said axial section having a cross-section less than that of said pins; and
    said pins being mounted in the opposite ends of said sleeve and abutting each other in said axial section thereof thereby causing resilient lateral expansion of said sleeve to produce inwardly directed radial forces holding said pins in lateral alignment.

    The ’797 patent

    1. A connector for one or more single optical fibers of a fiber optic cable having a strength member therein comprising:

    a support member having a forward section and a rear section;
    means on said forward section for releas-ably mounting a contact terminated to a single optical fiber;
    means for making fixed attachment to the strength member of a fiber optic cable; and
    means for removably mounting said attachment means on said rear section in a predetermined position relative to said forward section.
    ******

    6. A connector as set forth in claim 1 wherein:

    said support member comprises a yoke including two longitudinally extending legs connecting said forward and rear sections, said legs being laterally spaced apart defining an open cavity therebetween in which optical fibers of the cable may be spread apart. ******

    9. An optical fiber connector terminated to a fiber optic cable having a strength member and at least one optical fiber comprising:

    an elongated yoke having a forward section and a rear section connected by at least one longitudinally extending leg;
    a contact terminated to said optical fiber;
    means on said forward section for releas-ably mounting said contact thereon;
    means fixedly attached to said strength member of said cable; and
    means removably mounting said attachment means on said rear section in a longitudinally predetermined fixed position relative to said forward section.

    *32712. An optical fiber connector and cable assembly as set forth in claim 9 wherein:

    said attachment means includes a sleeve surrounding said cable and having an outwardly extending flange thereon; and
    said mounting means for said attachment means comprises a longitudinally extending recess in said rear section extending from the front to the rear thereof, and opening to one side of said rear section, and a transversely extending slot in said rear section intersecting said recess and opening to said one side, said sleeve being mounted in said recess with said flange being mounted in said slot.

    DISCUSSION

    I. Validity

    35 U.S.C. § 282 (1982), permits defendant as the party charged with infringement to contend that a patent is invalid because, inter alia, every element of each claim has been anticipated under 35 U.S.C. § 102 or because the subject matter as a whole would have been obvious under 35 U.S.C. § 103. For each of the patents in suit, defendant urges only the defense of obviousness under section 103, which provides in full:

    A patent may not be obtained though the invention is not identically disclosed or described as set forth in § 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negated by the manner in which the invention was made.

    (Emphasis added.) The Federal Circuit, consistent with the United States Supreme Court’s opinion in Graham v. John Deere Co., 383 U.S. 1,17-18, 86 S.Ct. 684, 693-94, 15 L.Ed.2d 545 (1966), requires that the trier of fact undertake a four-part analysis:

    [1] [T]he scope and content of the prior art are to be determined; [2] differences between the prior art and the claims at issue are to be ascertained; [3] and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonob-viousness of the subject matter is determined. [4] Such [objective evidence of] secondary considerations as commercial success, long felt but unsolved needs, failure of others, etc., might be utilized to give light to the circumstances surrounding the origin of the subject matter sought to be patented. As indicia of obviousness or nonobviousness, these inquiries may have relevancy.

    (Citation omitted); accord Medtronic, Inc. v. Daig Corp., 789 F.2d 903, 905 (Fed.Cir. 1986); Loctite Corp. v. Ultraseal, Ltd., 781 F.2d 861, 872 (Fed.Cir.1985) (citing Simmons Fastener Corp. v. Illinois Tool Works Inc., 739 F.2d 1573, 1575 (Fed.Cir. 1984), cert. denied, 471 U.S. 1065, 105 S.Ct. 2138, 85 L.Ed.2d 496 (1985)). The Supreme Court stipulated that these inquires are factual, although the ultimate question of patent validity is one of law. Graham v. John Deere Co., 383 U.S. at 17, 86 S.Ct. at 693.

    A. Scope and content of the prior art—electrical art as analogous art

    The Patent & Trademark Office (the “PTO”) did not cite electrical art against any of the three patents in suit. In fact, the PTO classifies fiber optics separately. Nevertheless, the basic goal sought to be achieved by both fiber optics and electricity is the same—to propagate energy from a source to a- specific destination. However, light and electricity have different properties which significantly impact upon how the goal is achieved and how electrical and fiber optic connectors are fabricated.

    Energy travels down an optical fiber by total internal reflection. Simply put, light is continually reflected off the internal surfaces of the fiber. The light must proceed in a zigzag pattern. To obtain a maximum flow of energy, it is important to maintain the consistency of this zigzag pattern. Achieving this consistent flow of energy is difficult and becomes most critical at the junction of two optical fibers in a connector. At this point light must transfer between a glass surface, an air interface, and then another glass surface. Any significant bending of the fiber can distort the internal reflection of the light within the fiber.

    Energy in the electrical art is propagated by the conduction of electrons. This conduction, according to Mr. McCartney, operates by totally different physical principles from those which apply in fiber optics. A force or voltage is emitted from a source, causing the electrons to flow in a current *328by means of a continual bumping action. If the flow is interrupted due to the application of a connector, the electrons will continue to flow through that point without an energy loss. The transition at a connection need not be effected with the same critical precision that is required in fiber optics. Electrons need not flow in a lateral or straight pattern, but can travel in a tortuous path with little power loss. In comparing the alignment function of fiber optics with that of electricity, Mr. Borsuk said that connectors need only touch at any point to transmit electrons; to accomplish the propagation of light between two optical fibers, a precision alignment of optical fibers is critical. .

    The testimony of Mr. Borsuk and other witnesses described a variety of reasons why the technology applicable to electrical connectors differs from that applicable to fiber optics.

    First, lateral misalignment is a problem distinct to fiber optics. For the efficient transfer of photons and maintenance of total internal reflection, optical fibers must be joined in precise axial alignment. This means that the two fiber ends must both be aligned on the same axis. An electrical connector can be at “90 degrees if the flow of electrons [goes] through the contact,” according to defendant’s witness Donald B. Forman.

    Second, it is critical that no significant gap be present between two optical fibers sought to be joined. Mr. Borsuk explained: “When you have air, you always have three-tenths of a [dB] loss and you want to eliminate that. That’s a fixed loss in any connector, even a perfectly aligned connec-tor____” The presence of an extensive gap would cause detrimental light loss. This is not to say that optic fibers should touch; rather, proper abutment requires optic fibers to be in close proximity. The slight light loss that arises due to this gap is eliminated by placing a fluid between the fiber that has an index of refraction similar to that of glass, so that light can propagate through it. Electrical contacts, on the other hand, must touch. Electrical connectors utilize a pin and socket concept and thereby avoid transmission losses.

    Third, “axial tipping” or “angular misalignment” is a problem associated with fiber optics. This problem manifests itself by the existence of an angle between the axis of two fibers. If fibers are on an angle to one another, even if close together, some of the light from the emitting fiber will not enter the receiving fiber, but rather will shine off to the side. The pin and socket contacts in the electrical art eliminate these tolerances automatically.

    Fourth, axial tolerance relief is a problem associated solely with fiber optics. Due to inconsistencies involved in manufacturing optical fibers, it is likely that multiple fibers in a single connector will be different lengths. In mating these fibers of varying lengths, according to Mr. McCartney, “[Y]ou would apply force to the first one, which would increase until you hit the second one, until you hit the third one ... until you hit a stop and it would undoubtedly have crushed the fibers in that contact.” These complexities are not associated with electrical connections which simply entail placing a pin within a contact.

    Fifth, optic fibers are minuscule strands of glass, and as such are extremely fragile and susceptible to breakage. Thus, optic fibers, unlike electrical contacts, are subject to chipping when placed in abutment. So, too, the fragility of optical fibers precludes them from withstanding axial pulling, whereas a “typical” military electrical cable is capable of withstanding loads up to 400 pounds.

    Sixth, electrical wires may be fastened to a contact by crimping after first being soldered to the contact. Optic fibers crush if crimped, so they must be glued into contacts. Because the glue generally will contaminate the ends of the optic fibers, the fiber ends must be polished. Avoiding contamination is a problem unique to optic fibers. The presence of a dust particle on the end of a fiber or fiber bundle literally can block total transmission, according to Mr. McCartney.

    *329In short, the most significant difference between the technologies of fiber optics and electricity is the necessity in fiber optics to construct connectors in a manner which diminishes lateral, gap, and angular misalignment. These mechanical degrees of freedom are not part of the calculus employed in the construction of an electrical connector.1

    The problems distinct to fiber optics are numerous, and Mr. McCartney, in inventing the '182 patent, addressed many of them in the one device. Thus, it would appear that the electrical art is not analogous for purposes of proceeding with the obviousness determination related to the '182 patent. However, militating against this approach is the fact that the parties do not dispute, first, that fiber optics was intended to replace the corresponding electrical technology2 and, second, that those making the first efforts to produce fiber optic connectors utilized electrical connectors as a starting point. Thus, even if it can be said that the electrical devices cited as prior art were not in the field of the inventor’s endeavors, the references are “reasonably pertinent to the particular problem with which the inventor was involved.” In re Wood, 599 F.2d 1032, 1036 (C.C.P.A.1979); accord Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 1535 (Fed.Cir.1983).

    Mr. McCartney and Mrs. Landgreen testified concerning their initial efforts to modify an electrical connector so that a fiber optic contact could be installed in one. It was believed that a significant amount of money could be saved if the production of fiber optic connectors began with the use of a standard electrical connector. The two worked with an electrical connector styled as a “pv connector A-3,” a connector used commercially, industrially, and militarily. Mrs. Landgreen testified that the “pv” had to be “gutted” because the fiber optic contacts that were to be inserted would not fit. At times, parts from the electrical connectors had to be thrown out completely. New parts had to be manufactured and the inside of the “pv” had to be reworked. The ultimate device which developed from their work was the ’182 patent.

    Norbert L. Moulin, a Hughes employee who is the named inventor of both electrical and fiber optic patents, testified that one of the accused devices illustrated at PX-25 was originally designed in approximately 1968 from an electrical connector styled the “C-21 Environmental Connector.” Mr. Moulin stated, “We built the optical contact alignment system so that it would package into the cavities of the electrical C-21 connector.” He also testified that the C-21 utilized a part called a “contact retaining clip” and that this part is now included in the accused device. Testimony was not elicited as to whether the clip operated in the C-21 as it does in the Hughes device.

    The article Parfitt & Ellis, “Interconnections and switches for glassfiber [sic] optical links,” 13 Electronic Components 69 (1972) [hereinafter cited as “Parfitt”], describes a fiber optic connector. However, the fiber optic connector was developed as a result of the modification of a BNC coaxial or electrical connector. On direct examination George A. Kosmos, defendant’s principal expert on the defense of invalidity, stated that in order to make the Parfitt fiber optic connector, it was necessary only *330to taper the bore of the electrical connector’s termination pin, or the point where the optical fibers reside. Thus, Parfitt is a fiber optic connector modified from an electrical connector, and it is defendant’s contention that the parts that it contained as an electrical connector served to remedy the problems related to axial tolerance in fiber optics.3

    As a general matter, electrical art is an analogous art to the extent that electrical connectors, and to some extent the parts in those connectors, formed the starting point for connector production in the fiber optic art and arguably may have assisted in resolving problems in the new art. Electrical art is particularly analogous with respect to the ’797 patent, because, like the electrical art, the patent addresses the problem of achieving strain relief. Both electrical and fiber optic art are concerned with developing strain relief which is a means by which forces on a cable, whether from pulling or some other external factors, can be sustained by the cable housing alone, rather than the wire or optic fiber internal to the housing.. Strain relief is achieved by fixedly attaching a mechanism onto a connector that is capable of clamping the outer surface of the cable or stress member without simultaneously clamping the conducting wire or fiber held within it.

    Mr. Moulin of Hughes testified that there are two differences related to the clamping process for the two arts. One clamps a braided material to accomplish strain relief in the electrical art. In fiber optics this goal is accomplished by mechanically gripping a strength member normally fabricated of Kevlar. Second, in the electrical art, in contrast to fiber optic art, the clamping hardware is electrically conductive and performs a grounding function. For purposes of strain relief, according to Mr. Moulin, the principles are identical.

    Plaintiff elicited testimony from Mr. Moulin to the effect that the level of sophistication of the requirements for strain relief with respect to tensile strength is higher for fiber optics than for electrical cables because of the fragile characteristics of optical fibers. Plaintiff, however, did not show that the restraining mechanism itself would be changed for the sole purpose of accommodating this difference.

    Mr. Moulin explained the distinction between a clamping mechanism for securing optical cable and for clamping electrical cable: “[A] compression nut and a ferrule against the housing” was the basic mechanism used in both arts, but with fiber optics “the ferrule, basically was changed to cope with [the] ... much finer diameter of the Kevlar strands of ... fiber optic cable, in comparison to the rather heavy shielded conductors ... [in] electrical cable.” Despite this difference, the primary emphasis of Mr. Moulin’s testimony was that in 1975 “the basic clamping mechanisms [were] ... very similar in ... that they [had] ... almost the same quantity of components, and the same configuration of components.” It was not until after 1975 that mechanisms for strain relief were modified, although some of the strain relief mechanisms of today are the same parts as in 1975.

    Mrs. Landgreen worked with electrical connectors containing cable clamps and reduced the cable clamp mechanism of the ’797 patent to drawings. She conducted some experiments and witnessed others work with these cable restraint mechanisms. Mrs. Landgreen displayed an understanding of the function of cable clamps and stated that their application and function were important in both of the arts, but in a more “intensified” manner in fiber optics.

    Mr. Forman, employed by the Naval Ocean Systems Center (“NOSC” or the “Center”), stated that “[a] strength member is a strength member ... it doesn’t matter if it’s an electrical wire with strength members on it, hydraulic line with strength members, or a fiber optic connec*331tor with strength members____” The strength member always serves the same purpose—to sustain “loads.” However, Mr. Forman did agree that with wire and coaxial cable, “[t]ension load” may be placed on the wire itself, but with fiber optics such a load would break the fragile glass fiber.

    The ’145'patent is intended to resolve the problems in the fiber optic art related to angular misalignment and lateral alignment. Although these problems allegedly are distinct to the fiber optic art, the alignment sleeve of the ’145 patent supplements the efforts of the ’182 patent to resolve these problems. The ’145 patent resolves no other problems, and Mr. Moulin’s testimony indicates that parts were utilized in the electrical art which would carry over to resolve the alignment problems in the fiber optic art.

    Defendant relied on Mr. Moulin’s electrical connector, United States Patent No. 3,380,012, Apr. 23, 1968 (“Moulin I”), to illustrate the use of a slotted tubular sleeve for “[causing] intimate contact between [a] ... pin and ... socket finger members” sought to be connected. Mr. Moulin testified that the sleeve grips the contacts “[m]oderately tight for the purpose of making an electrical conductivity path” and that the sleeve would also be useful in fiber optics for grabbing the guide bushings (the body of the contact behind the chamfered head) of mated contacts. The specifications in Moulin I say that the resilient quality of the slotted sleeve accomplishes an alignment function. On this basis it can be said that the electrical art is at least minimally analogous with respect to the ’145 patent.

    B. Level of ordinary skill in the art

    “It should be clear that [the] hypothetical person [of 35 U.S.C. § 103] is not the inventor, but an imaginary being possessing ‘ordinary skill in the art’ created by Congress to provide a standard of patentability____” Kimberly-Clark Corp. v. Johnson & Johnson, 745 F.2d 1437, 1454 (Fed.Cir.1984) (emphasis in original); accord Standard Oil Co. v. American Cyanamid Co., 774 F.2d 448, 454 (Fed.Cir.1985) (“The actual inventor’s skill is irrelevant to the inquiry____”). If the inventor is not the person of ordinary skill, neither is the court, Litton Industrial Products, Inc. v. Solid State Systems Corp., 755 F.2d at 158 (Fed.Cir.1985). The Federal Circuit in Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693, 696 (Fed.Cir.1983), cert. denied, 464 U.S. 1043, 104 S.Ct. 709, 79 L.Ed.2d 173 (1984), established several criteria to guide this determination, including: (1) educational level of the inventor; (2) type of problems encountered in the art; (3) prior art solutions to those problems; (4) rapidity with which innovations are made; (5) sophistication of technology; and (6) educational level of active workers in the field.

    1. Educational level of inventors and active workers in the field (criteria 1 and 6)

    The court found during trial the level of ordinary skill was not high. This finding derived from the educational background of the witnesses, most of whom were inventors of fiber optic and electrical connectors and components. This analysis failed to take into consideration the other criteria identified by the Federal Circuit, dwelling too heavily on the education of inventors, and is supplanted by the following discussion. Mr. Borsuk received a Bachelor of Science Degree in Mechanical Engineering and has completed some graduate course work. Mr. McCartney has a undergraduate degree in Physics, a Master’s Degree in Organizational Management, and an MBA. Mrs. Landgreen is a high school graduate and attended some junior college classes. Mr. Ledbuska, another inventor who testified as a defense witness with respect to his own patents in fiber optics, has a Master’s Degree in Electrical Engineering. Donald Williams, an electronics engineer with NOSC who testified for defendant about the early work of the Naval Electronics Laboratories in fiber optics, has a degree in electrical engineering. David *332William de Champ, Sr., a Senior Products Design Engineer for ITT Cannon, had no college degree. Mr. Moulin, the Hughes employee who invented devices that allegedly infringe plaintiffs, does not have a college degree, although he attended college classes in machine design, as well as other subjects. He described himself as self-educated. Defendant’s principal expert witness, Mr. Kosmos, has a Master’s Degree in Electrical Engineering. Mr. For-man, another inventor, has a Bachelor of Science Degree in Mechanical Engineering. Morton Pomerantz, a former test engineer for the Army, has a degree in Electrical Engineering.

    The court in Stewart-Warner Corp. v. City of Pontiac, 767 F.2d 1563, 1570 (Fed. Cir.1985), would disagree with this court’s assessment that the level of ordinary skill was not high, considering the educational degrees of these individuals who are either inventors or active workers in the field. However, according to these witnesses, their knowledge and experience fairly span the field of electronics and fiber optic technology in the early 1970’s, and the median education was a four-year engineering degree, which, in almost no field is considered a prima facie indication of a level of high or extraordinary skill.

    2. Type of problems encountered in the art (criterion 2)

    The discussion of whether and to what extent electrical art is analogous art treats six problems encountered in the fiber optic art.

    3. Prior art solutions to the problems, rapidity with which inventions are made, and sophistication of technology (criteria 3-5)

    The level of technology in the field of fiber optics prior to October or November 1974 is pertinent for the ’182 and ’145 patents, respectively, and before June 1976 for the ’797 patent. The testimony at trial describing the tests conducted, contacts and connectors produced, literature published, and general discussion and interaction of active workers in the field are telling in resolving this issue.

    Defendant offered the testimony of several witnesses, including Messrs. Lebduska and Kosmos, to support its contention that during the relevant period the level of technology in the field of fiber optics was high. Mr. Lebduska stated that he began working with fiber optics in 1971 at NOSC, along with a small group of individuals. At this time several companies, including Corning Glass Works, were producing a simple fiber optic cable of bundle configuration encased in a plastic jacket. Mr. Leb-duska believed that other than at NOSC, as of 1971, no strong effort was being made to conduct research or experimentation in the new field. As of 1972 there were articles written on the topic of fiber optics, but it was not until 1974 or 1975 that outside individuals began making inquiries concerning the laboratory work conducted at NOSC and particularly the Center’s requirements for cables and connectors. Not until 1975 did NOSC have a significant number of individuals (25) working full-time in fiber optics.

    Between the end of 1972 and August or September 1973, Mr. Lebduska conducted two sets of tests on fiber optic cables of bundle and other configurations. Although these fibers would not be considered very low loss by today’s standards, they were so considered for that time period. Because it was intended that fiber optic cables ultimately might replace electrical cables, the tests were conducted according to the standards appropriate for electrical cables. For the first testing, Mr. Lebduska ordered from Coming Glass $1000 worth of two types of fiber optic cable. He believed that this may have been some of the first cable that Coming produced. For the second testing, four different types of cable were used, three of which were produced by three separate companies.

    The research that Mr. Lebduska undertook placed particular emphasis upon ascertaining the number of fibers that would break when wrapped around a mandrel, or *333curved surface, and the point at which such breakage occurred. The results of the first test issued in March 1973, and the results of the second test were documented in a report issued on December 11, 1973. The report was disseminated widely both throughout the United States and into foreign countries. The report did not discuss the termini that were used in the two tests. The termini were “crude” or “very simple,” and would bind and compress the individual fibers to make them efficient in their light-coupling properties. The fibers were then epoxied into the brass terminal ends. The termini had no moving parts, but, rather, were one complete piece. Prior to December 13, 1973, Mr. Lebduska and others also modified electrical connectors and used such connectors to test optical fibers. When these tests were performed, all terminals were either of this same nature or in the nature of “small coupling sleeves,” and utilized solely to facilitate cable testing. Connectors for actual practical application were not available commercially. The results of this testing were recorded in a technical note, TN 2367 (May 8, 1973), a publication of an informal nature. Because the note was considered informal and the information contained therein “somewhat proprietary,” dissemination of the note outside of NOSC was limited. The first public embodiment of the technical note was the 1973 report.

    During the period in which Mr. Lebduska was ordering cable from Corning Glass in 1973, he was aware that Coming was also conducting tests in the area of fiber optics. He recalls twice visiting Coming during this period and being shown test equipment related to light loss. He was also shown a “low loss hex arrangement” or connector with which Frank Theil of Coming was experimenting. Mr. Lebduska did not recall seeing any other connectors at Coming. Finally, as of 1974, Mr. Lebduska recalls attending electro-optical conferences presented by the Society of Photo-Optical Instrumentation Engineers (“the SPIE”). In January 1974, he made a presentation to the SPIE, entitled “Fiber Optic Cable Test Evaluation,” in which he discussed the results of the two tests that he had conducted.

    Ultimately, Mr. Theil of Coming Glass used his conhex connector as the subject of a 1974 conference on the concept of the low-light loss connector.4 The conference was also the subject of a paper representing Mr. Theil’s first disclosure of test results on the topic of light loss in the fiber optic junction.

    Mr. Kosmos, defendant’s second witness on this issue, worked under the supervision of Mr. Lebduska beginning in 1973 and assisted him in making valuations relating to the degradation of cable resulting from testing. In fact, Mr. Kosmos agreed that from 1973 to 1976 his primary work experience entailed administering the fiber optic cable program for the Navy and working with Mr. Lebduska testing cables. After 1973 Mr. Kosmos was assigned to administer a contract to develop glass on glass fiber optic cable which at that time was considered state of the art.

    Mr. Kosmos is the author of eight articles in the field of fiber optics, and although published, their dissemination to the public was limited. Defendant contends that these articles establish Mr. Kos-mos’ involvement in the design of fiber optic connectors during the relevant period, but no direct references to the content of these articles was elicited. The record indicates that only one of these articles was written during the relevant period and it pertained to a study of how liquid contaminates affect the radiation pattern within a connector housing. Mr. Kosmos did testify *334that he was familiar with connector design in order to specify cables, although he personally had never built an in-line fiber optic connector.

    In 1980 a handbook on electrical connectors was published containing an article by Mr. Kosmos on fiber optic connector. Mr. Kosmos had no knowledge of the extent to which this British publication was distributed. He stated that “packing fraction” referred to in the article (ratio of light transmitting core area of a fiber optic bundle to the area of the bundle orfice housing those fibers) was a known phenomenon that was considered significant during the relevant period in bundle technology.

    During the early 1970’s, a variety of scientific articles regarding fiber optics was in circulation. Mr. Kosmos stated as early as 1973, “[tjhere were numerous mechanisms for providing fiber optic infor-mation____” Course presentations and technical programs comprised a part of Mr. Kosmos’ work in fiber optics. He taught a three-week college course on the subject. Mr. Kosmos’ only referred to the content of his teachings with respect to one conference held in June 1976. This conference did not focus on fiber optic connectors, and Mr. Kosmos lectured on fiber optic cable design, operation, and light transmission. Another part of the conference was practical work entailing terminating fibers. Although he makes no specific reference to a particular connector or its level of sophistication, Mr. Kosmos contended that connectors were produced during the relevant period because he was specifying cables for them. Donald Williams, who was assigned in 1971 to design demonstration equipment for fiber optics for the Navy, testified similarly about the early activity in fiber optics, as did Mr. Pomerantz.

    For the relevant time period, none of defendant’s witnesses testified convincingly that the level of sophistication relating to fiber optics was anything but limited. Even as late as 1973, Mr. Lebduska was merely testing fiber optics for breakage and was using very crude one-piece terminals in doing so. He also testified that as of 1973 he had seen but one connector and that it was still in the experimental stage. Mr. Kosmos’ testimony suggested that sophistication existed in the art because he and others had a significant number of publications in the field that had received public dissemination. His testimony is not persuasive. He did indicate that an elevated level of technology existed during the relevant timeframe with respect to fiber optic cables. However, with respect to fiber optic connectors, those present in the field of fiber optics could boast nothing more than crude devices devoid of sophistication that did not solve the problems unique to fiber optics. Plaintiff’s witnesses, Mrs. Landgreen and Mr. McCartney, testified accordingly.

    Resuming her employment with ITT Cannon, Mrs. Landgreen began general drafting work in 1968. In approximately 1970 she was selected to work in the research and development/new products group. Mrs. Landgreen began working with Mr.McCartney in 1974, although she had no familiarity with fiber optics. Mrs. Landgreen’s first responsibility was to find an electrical connector that could somehow be modified to accommodate optic fibers. She testified that in performing her job she was acting as a pioneer in a field of new technology. This opinion, however, was based primarily upon what she had heard from Mr. McCartney. She was aware that there were companies producing optic fiber, but had no knowledge that connector housings were being produced. She testified that in working with Mr. McCartney, discussions took place concerning the problems associated with fiber optics, such as light loss, the fragile nature of the glass, and light continuities.

    Mrs. Landgreen also testified as to the type of fibers with which she worked, including “plastic fibers, glass single fibers, glass multiple fibers, [and]____whatever the cable manufacturers were able to make at the time, which was pretty crude stuff.” With respect to her efforts to produce fiber optic connectors from scratch, Mrs. Land-green testified that it was strictly a matter *335of trial and error and that her ideas frequently failed when put into practice.

    When Mr. McCartney was assigned to do work in fiber optics, he knew nothing about the field and found only one article on point. In response to the question whether he encountered problems in his early work in 1974, Mr. McCartney responded:

    Any problems? It was totally unknown technology. We had to first discover what it was that had to be done and then attempt to create mechanisms and ideas and devices that would handle it and then check those with what you might call optical and mechanical theory to see if they could be built. But it was all a paper study at that time.

    The notion of pioneering in the field was dampened somewhat by plaintiff’s Area Patent Counsel Thomas L. Peterson, who testified that in late spring 1984 he gave Mr. Arnold a stack of fiber optic patents, which apparently Mr. McCartney and Mrs. Landgreen did not review.

    Even at present Mr. McCartney would characterize the pace of development in the field of fiber optics as “slow.” To the best of his knowledge, no schools offer a degree in fiber optics. At the most the topic is covered as a subcourse to a normal engineering curriculum. Fiber optic engineering is a recognized speciality today, but few concentrate in it. “|T]here’s generally only one person in a large company that’s skilled in the art.”

    4. Person of ordinary skill in the art

    The hypothetical person of ordinary skill in the art in 1974-1976, based on the preceding analysis of the six criteria that define the individual’s education, knowledge, and experience, was someone with a Bachelor of Science degree in Electrical Engineering who had a limited knowledge of the problems associated with optic fibers, contacts, and fiber optic connectors.

    C. Interpretation of the claims at issue

    Claim interpretation is a matter of law ultimately for'resolution by the court based on factual underpinings adduced at trial. Structural Rubber Products Co. v. Park Rubber Co., 749 F.2d 707, 714 (Fed. Cir.1984). The court must consider all elements of the claimed invention and cannot compare the prior art with what one may consider as the gist of the invention. Perkin-Elmer Corp. v. Computervision Corp., 732 F.2d 888, 894 (Fed.Cir.), cert. denied, 469 U.S. 857, 105 S.Ct. 187, 83 L.Ed.2d 120 (1984); W.L. Gore & Associates v. Garlock, Inc., 721 F.2d 1540, 1548 (Fed.Cir.1983), cert. denied, 469 U.S. 851, 105 S.Ct. 172, 83 L.Ed.2d 107 (1984).

    1. The ’182 patent (See illustration 1 appended to this opinion.)

    After trial the parties briefed the issue of whether the language in the preambles to claims 1 and 2 of the ’182 patent should be interpreted as a claim limitation for purposes of obviousness. A complete analysis of claims 1 and 2 of the ’182 patent follows resolution of this issue.

    a. Claim 1

    The preamble to claim 1 states: “A termination pin assembly for adapting a fiber optic cable for insertion into a connector assembly, comprising____” The remainder of claim one recites the elements comprising the pin body itself. The preamble to claim 2 provides:

    In a fiber optic connector for providing a removable light signal connection between the ends of at least one pair of fiber optic cables, said pair consisting of first and second cables, each containing a fiber optic bundle comprising at least one light transmitting fiber, said connector having first and second mating connector shells and corresponding guide means for holding said first and second cables in substantially colinear positions with said ends of said first and second fiber bundles in substantial abutment when said shells are mated; the combination providing: .. .5

    *336Both parties agree that Kropa v. Robie, 187 F.2d 150 (C.C.P.A.1951), states the rules for construing the language of a preamble to constitute an actual claim limitation. They also cite Loctite Corp. v. Ultraseal, Ltd., 781 F.2d 861, and Perkin-Elmer Corp. v. Computervision Corp., 732 F.2d at 896, as recent decisions following the rule set forth in Kropa.

    In Kropa the Court of Customs and Patent Appeals held that the preamble will not constitute a claim limitation if the body of the claim at issue may be read independently of the preamble and does “not [depend] for completeness upon the introductory clause.” 187 F.2d at 152. If the body of the claim “completely [defines] the subject matter, and the preamble merely [states the] ... purpose or intended use of that subject matter,” then the preamble will not constitute a limitation. Id. However, where the preamble “[points] out the invention defined by the claim” or gives “life, meaning and vitality to the claims,” then the preamble will constitute a limitation. Id.

    Defendant argues that the body of claim 1 defines a complete structure, therefore making it unnecessary to refer to the preamble. Although the claim points out a complete structure, it is necessary to examine the preamble in order to ascertain the invention defined by claim 1. The body of the claim recites the various parts comprising the pin body, but only from the preamble may one determine that the combination of these parts constitutes a “termination pin assembly.” The preamble thus defines the device formed by the parts described in elements 1 through 4 of claim 1 and does not state merely the purpose or intended use of the subject matter defined in the body of the claim.

    Kropa also instructs that in the types of cases in which the preamble has been deemed to lend vitality to the body of the claim, “usually ... there is inhered in the article specified in the preamble a problem which transcends that before prior artisans and the solution which was not conceived or known by them.” However, this is not a requirement. Loctite and Perkin-Elmer are consistent with this view, in that these decisions refer only to the requirement that the preamble “[breathe] life,” Loctite, 781 F.2d at 866, or “give meaning”, Perkin-El-mer, 732 F.2d at 896, in order to constitute a claim limitation. Even so, the concept of the termination pin assembly, at least as it is defined by function in the elements of claim 1, was a problem before prior artisans and unresolved by them.

    Both parties agree that claim 2 is a Jep-son claim. As a result, defendant argues that the preamble constitutes prior art and cannot be a claim limitation. Plaintiff states that in the Jepson format, “the preamble sets forth a general description of the elements of the claimed- combination which are known in the art, while the body of the claim sets forth the new elements of the claimed combination____” Plf’s Brief, filed May 7, 1986, at 4 (empha*337sis in original). Placing particular emphasis on the reference in the preamble to the device’s capacity to provide “ ‘a removable light signal connection,’ ” plaintiff argues that the preamble in claim 2 constitutes a claim limitation because it illustrates the manner in which the “known connector shells ... [and] termination pin assemblies” interact with the termination pin assembly described in the body of claim 2. Id.

    Because plaintiff conceded that remova-bility is not a feature that it would consider novel to the ’182 patent, see supra note 5, and that the remainder of the features described in the preamble are prior art, there is no issue as to whether the preamble in claim 2 constitutes a claim limitation. The guide means feature of the preamble to claim 2 has been analyzed for completeness; its absence in the prior art has been deemed irrelevant to the conclusions on validity.

    b. Claim 2

    Through the testimony of Messrs. Borsuk and McCartney, plaintiff supplied a comprehensive analysis of the scope and content of claim 2, deeming it to be representative of claim 1.6

    The preamble relates to a connector means for holding cables in substantially colinear positions and maintaining fiber bundles7 in abutment. Thus, a cable is inserted into a termination pin. A “contact subassembly” is the means for maintaining the cables in colinear position. According to Mr. Borsuk, the subassembly is comprised of “a pin contact—-that’s a singular part, but all these parts that work in association with it, such as the ... resilient spring member and the ... slidable sleeve, are part of that assembly____” In “engineering [terms]” a subassembly is several pieces assembled together. Abutment of termination pins is maintained by the interaction of the slidable sleeve, resilient member, and retention clip fixed within a shoulder in the connector cavity. Although Mr. Borsuk accurately described plaintiff’s invention, consistent with the interpretation of the elements of claims 1 and 2, the preamble itself is not a claim limitation insofar as plaintiff conceded that it refers to a fiber optic connector as it existed in the prior art.

    The first element describes a hollow cable termination pin. Cables are to be terminated therein. The pin has an axial cavity or bore which extends the entire length of the pin. This cavity or bore accommodates the cable and, for a distance, exposed fibers. When the cable is inserted within the pin, the claim discloses that it must lie in a “colinear or coextensive” position with the terminating end of the pin.

    The second element requires that the optic fiber be adhesively bonded into the hollow portion of the terminating pin. Adhesion prevents light loss by immobilizing the fibers and preventing their axial movement.

    The third element refers to a means comprising a coaxial sleeve, slidably fit about the outside diameter of a central portion of the termination pin. This is a circular sleeve located in the middle of the pin. Mr. Borsuk testified that it “is capable of sliding with respect to” the pin. The sleeve is *338restrained by the clip that is “mounted within the connector cavity ... so that it cannot move aft. But the sleeve is slidable with respect to the pin____” Mr. Borsuk described the phenomenon further: “The pin may slide aft or to the left ... within the sleeve. And if I were to remove this contact from the connector as a subassem-bly [described in preamble analysis], we could see that they slide with respect to each other when not installed____”

    Element 4 refers to a means for restraining the axial translation of the sleeve in a direction away from the point of abutment. This defines the caesar or contact retention clip that is trapped within the connector. The clip has small “tangs” or “cantilever beams” that latch behind the shoulder on the slidable sleeve and restrain the sleeve from axial translation (or motion) away from the front face of the contact.

    The fifth element refers to a resilient member associated with the body of the termination pin and a coaxial sleeve associated with the pin and resilient member. The parts are associated in that they “function together.” The resilient member is an O-ring, Mr. Borsuk said, “but the key is that it’s resilient and resiliency means a spring such that when I compress it I generate a [resilient] force____” Mr. McCartney stated that the term “resilient member” is to be interpreted loosely. It could include “a common coiled spring, [Belle-ville] washers, snap rings ... and any other type of spring member ... [that would cause] an axial resilient distortion.” However, for effecting the purposes intended by the ’182 patent, an O-ring was most suitable.

    Mr. Borsuk described the purpose of the resilient member and its association with related parts:

    [The O-ring is] ... there to permit relatively small axial translation. Axial translation means movement of the pin within the sleeve, thereby to provide a compressive force at the point of abutment of the first and second fibers. Also providing relief for axial manufacturing tolerances. The reason for motion is that you must absorb manufacturing tolerances in all of the hardware as all of the contacts within the connector meet. The motion is to the left in this drawing, the sleeve, the pin moves, again, within the sleeve. And that’s this problem of relative motion. But it’s obvious that since the sleeve is restrained by the clip, that when it is pushed on from the front it must move to the left. And since the resilient member is trapped, if you will, in association between the faces of the sleeve and the shoulder of the pin, it must compress that resilient sleeve. When it does there will be an equal and opposite force to the right which will cause a force on the point of abutment of the two pins.

    Mr. McCartney testified that abutment is achieved in “a very controlled manner.” He also noted the device operates to achieve a “floating intermediate position between the pins.” “If one cable pulls back ... and pulls a contact away from this point where they are lined up, the other ... [contact] is still being pushed forward and it just fills the gap automatically. So, these things pump back and forth together____ They just automatically float together. Very essential in fiber optic technology.” Mr. Kosmos offered testimony consistent with this interpretation of the interactions or functioning of the parts in the ’182 patent.

    Claim 1 differs from claim 2 in two significant respects. The second element of claim 1 refers to the structure in which the resilient annular ring means rests. Specifically, claim 1 refers to a “circumferential groove about a central position of said pin body, said groove having first and second side walls.” In defining the groove, Mr. Borsuk began from the premise that a groove is a “space or volume” and stated that it is simply “the space the spring is in.”

    However, the claim language specifically states that first and second side walls within the pin body form the groove. Both Messrs. Borsuk and McCartney recognized that a portion of the groove was machined *339into the termination pin itself at its central portion. The first side wall, or the wall closest to the pin end, appears as a 45-de-gree sloping angle. Both also recognized that the side wall of the groove located furthest from the pin end also includes the front face of the slidable sleeve. Mr. McCartney stated that “[t]he groove was split because we had to [have it] act on the O-ring____” He pointed to the optical application action. The O-ring responds “to axial thrust of said pin in a direction away from said point of abutment by climbing said second groove wall thereby to provide a resilient reaction to said thrust to keep said point of abutment under compressive stress.” Because the front face of the shoulder of the slidable sleeve does not move, it serves to facilitate the movement of the O-ring up the forward groove wall.

    The second significant difference is the requirement in claim 2 that the device accommodate axial manufacturing tolerances.

    c. Insertion and Removability

    Although the features of insertion and removability of a termination pin assembly are not limitations to claims 1 and 2, the device in the ’182 patent nevertheless functions to achieve these results. As detailed in the analysis of claim 2, the present invention contains a termination pin assembly consisting of a termination pin surrounded by an O-ring and a slidable sleeve. A retention clip is fixed in the connector shell. The clip, according to Mr. McCartney, “is nothing more than a tubular piece of material, generally slit, that has hooks that come down.” They are so designed that in conjunction with the slidable sleeve insertion of the pin causes these hooks to slide down and slip into place behind a shoulder on the rear face of the slidable sleeve. The hooks then retain the contact in place. Plaintiff introduced a “contact insertion extraction tool.” The tool in inserted in the rear of the connector and will fit around the slidable sleeve. The tool is of such a diameter that it expands the tangs on the retention clip. When the tool is extracted, the released contact will slide out with it. Insertion is accomplished in the reverse manner.

    d. Figure 3 (See illustration 2.)

    Although plaintiff contends that claims 1 and 2 pertain only to figure 2 of the ’182 patent or the device depicted in illustration 1, defendant takes the position that the claims at issue also disclose the configuration in figure 3 of the ’182 patent or the device depicted in illustration 2.

    Mr. McCartney testified on cross-examination that the termination pin depicted in illustration 2 was capable of sliding rear-wardly through a nut in the rear of the device or what defendant would argue is the equivalent of a slidable sleeve. The patent specifications also refer to the capacity of the device to achieve a compression force at the point of abutment.

    Although there is some similarity in the final result achieved, the device in illustration 2 does not accomplish the result in a manner which is at all consistent with the claim interpretation of claims 1 and 2 and that corresponds to the device depicted in illustration 1. In fact, on cross-examination plaintiff elicited from Mr. Kosmos testimony to the effect that the device in illustration 2 does not utilize the slidable sleeve which was testified to as pertaining to claims 1 and 2. Referring to illustration 2, the patent specifications state: “Rather than utilizing a slidable sleeve between the rention clip and the O-ring ... in this embodiment of the invention the slidable sleeve is eliminated____” (Emphasis added.) Mr. Borsuk testified that the device in illustration 2 is an invention different from the one disclosed in the claims at issue. Unlike the invention depicted in claims 1 and 2, the device in illustration 2 does not contain a termination pin assembly as it is defined with respect to claim 1. Two or more parts are not assembled together on a termination pin. There is neither a slidable sleeve nor an O-ring mounted to the groove. Also, there are no groove side walls machined into the termination pin. Given the incongruity between the parts in the devices depicted as illustrations 1 and 2 *340and the difficulty of viewing the device in illustration 2 consistently with claims 1 and 2, the device in illustration 2 will not be viewed as an embodiment of the claims in issue of the ’182 patent.

    2. The ’H5 patent (See illustration 3.)

    The ’145 patent teaches a means for generating efficient light signal transfer between the ends of two fiber optic cables. Disclosed in this patent is an alignment sleeve which is intended to effect both lateral and rotational alignment of interfacing termination pins containing fiber optic cables. The alignment sleeve is narrow at the point at which the pins abut. The outside ends of the pin are flared and do not come into contact with the pin. It is only at the point at which the sleeve abuts the pins that lateral alignment is effected by inwardly directed radial forces. Where the sleeve flares, it does not come into contact with the pin and only supplements the alignment function. The sleeve contains “at least one axially extending slot,” which must overlap at least the point of abutment of the two termination pins.

    Mr. McCartney, the named inventor of the ’145 patent, testified that the alignment sleeve achieves lateral alignment regardless of the shape of the termination pin, but also achieves rotational alignment where the termination pin is of a polygon formation. Where the pin is a polygon, the alignment sleeve must have multiple slots. He testified how this requirement of claim 13 is satisfied: “[T]he comers of the termination pin engage the said grooves and ensure alignment in the rotational or twist mode. The grooves may be straight, that is to say parallel ... [to] the axis of the connector or may be spiraled or rifled to provide a random pick-up of the termination pin corners during insertion.” The drawings appended to the ’145 patent depict a six-point alignment system.

    Elements 2 and 4 of Claims 1 and 13, respectively, pertain to the capacity of the sleeve, which is fabricated of a resilient thin-walled metal, to generate inward directed radial forces and thereby achieve lateral alignment. Claim 13 supplements claim 1 by stating the manner in which alignment function referred to in the two claims is effected. Element 3 of claim 13 requires that the termination pins have a diameter slightly larger than that of the alignment sleeve. However, both claims recognize that insertion of the termination pins tends to spread the finger(s) represented by the metal strip(s) between the slot(s) in the sleeve. Accordingly, along the length of the sleeve, the finger or fingers will exert a radially inward force tending to prevent lateral offset at the point of abutment.

    The core of the controversy concerning claims 1 and 13 is whether they are premis-sive of an alignment sleeve containing both single and multiple slots. Also at issue is the question of the extent to which the slot or slots extend over the sleeve. Plaintiff contends that the claims teach an alignment sleeve with only a single slot, which is designed to receive a circular termination pin. The slot need only contact the termination pins over their point of abutment, but may extend the full length of the sleeve. Defendant contends that the claims teach multiple slots that do not extend the full length of the sleeve, but do extend into the sleeve’s flared portions.

    Mr. McCartney testified that in utilizing the ’145 patent to align the points of a termination pin of hexagonal shape containing fiber bundles, the sleeve needed multiple slots. Although he conceded that to maintain a sleeve manufactured in this configuration, the slots could not extend the entire length of the sleeve, Mr. McCartney insisted that one is not precluded from aligning tubular termination pins containing either single fibers or a bundle of fibers within a sleeve containing a single slot cut through the whole sleeve. According to Mr. McCartney, this is especially true because “we have purposely narrowed the center portion of that sleeve so that it acts as if it is a total cut through the whole thing.” It is Mr. McCartney’s belief that if the contacts are circular (whether containing single or bundle fibers), rather than hexagonal, only one slot, with what he re*341ferred to as a three-point alignment system, would have been used.

    Regardless of the number of slots in the sleeve, insertion of the termination pin assembly therein produces resilient lateral expansion. Mr. McCartney stated that the sleeve would spread if there were one slot or several slots, and the forces that are generated by this spreading will hold both pins in a lateral manner and keep them restrained together in alignment. Mr. McCartney went on to state that the sleeve could be of a tubular configuration—“a straight tube, with no connecting or hourglass figure. It could have had one slot all the way through it or the slot could have been partially through it, as long as it’s overlapping the portions where the two contacts come together.”

    Mr. Borsuk supported this testimony by stating that the central portion of the sleeve in the ’145 patent effects the lateral alignment function. The claims at issue, according to Mr. Borsuk, require the use of a sleeve containing only one slot which accepts a cylindrical termination pin. Or, more specifically, Mr. Borsuk testified, the claim language of elements 2 and 3 of claims 1 and 13, respectively, referring to “at least one axially extending slot,” is permissive of a sleeve with but one slot. Accordingly, Mr. Borsuk clarified that the drawing in figure 3 of the ’145 patent (see illustration 2) depicting a sleeve with multiple slots corresponded to those claims not in issue that pertain to the alignment of a hexagonally shaped termination pin.

    In response to an inquiry from the court, Mr. Borsuk agreed that the use of a rubber sleeve extending over the point of abutment of the termination pins would perform the same function as the sleeve in the ’145 patent with an allegedly single slot.

    On cross-examination Mr. Borsuk did not deviate from his testimony that the alignment function served by the sleeve occurred only at its central portion and that to accomplish the function allegedly claimed by the single slot, it was required to extend no further than over the central portion of the sleeve. However, on this point alone, Mr. Borsuk’s responses lacked conviction. In response to the question whether a termination pin satisfying the criteria of claim 13, i.e., a pin with a diameter larger than the sleeve, could be inserted within a sleeve containing a slot only over the central portion, Mr. Borsuk stated:

    We’re only talking about expanding this part a thousandths of an inch. I think I could press it into it, yes. I think I could press a pin, whose diameter is infinitally smaller than—I mean, is larger, just slightly larger—I could balance that and cause it to press in. The design of this product is a function of the relationship between the spray forces that are radially anchored on this sleeve, which cause friction as we enter, relative to the forward forces caused by the resilient member on the contacts, such as [Belleville] springs on Hughes, which exhibit a tremendous amount of force. Bellville springs are very stiff. Or, other types of springs. So, you balance that force so it’s greater than the frictional force to cause this larger pin to enter. With enough force I could press fit the first section even if it was a solid hoop.

    It is inferred from Mr. Borsuk’s testimony that an excessive amount of force would be required to insert the termination pin into a sleeve containing a slot only over its central portion. However, defendant offered no evidence to the effect that claims 1 and 13 do not contemplate the use of such force. Because Mr. McCartney supported Mr. Borsuk on this issue, and because both witnesses contend that the slot could extend not only into the flare of the sleeve, but also its entire length (in which case the sleeve could easily receive a pin of a larger diameter), Mr. Borsuk’s lack of certainty is of no moment. Finally, on cross-examination Mr. Borsuk recognized that were the sleeve slit from end to end, the forces exerted upon the termination pin would be distributed differently than if the sleeve were not slit to this extent.

    *342Defendant moved into evidence through Mr. Kosmos an abstract of a patent which issued to Mr. Arnold (the “Arnold patent”) in 1980 and was filed on March 1, 1976. Certain statements in the abstract made solely by ITT or its representatives were offered as admissions, although the whole document was received for context.

    By way of background, all of the claims in the Arnold patent were rejected on September 10, 1976. One basis for rejection was the determination that Mr. McCartney’s '145 patent taught that which the rejected claims in the Arnold patent disclosed. Mr. Peterson, Area Patent Counsel for plaintiff, submitted an amendment to the application, which responded to the rejection notice issued by the PTO:

    The Examiner has cited McCartney for the teaching of longitudinal slots in an alignment sleeve. However, the slots 32 do not extend to the opposite ends of the sleeve. The slots are utilized primarily for the purpose of guiding hexagonal termination pins. Since the slots do not extend to the ends of the sleeve, they do not function to provide a resilient mounting of the sleeve within the connector bore, as in Applicant’s invention. Therefore, there is no teaching or suggestion in either of the references cited by the Examiner of providing a longitudinal slot in an alignment sleeve which extends to the opposite ends of the sleeve for resilient mounting of the sleeve in a connector socket. (2606)

    Mr. Peterson submitted a second amendment to the Patent Office, as well. He reiterated that which he stated earlier and asserted further distinctions between the ’145 and the Arnold patent.

    The termination pins in McCartney are not cylindrical. They are hexagonal and the alignment sleeve 15 is constructed with oblique slots 32 for rotationally aligning the hexagonal pins. To make the pins in McCartney cylindrical, as in Applicant’s invention, would defeat the total purpose of the McCartney invention.
    ... The mid-region of the alignment sleeve in McCartney contains six longitudinally extending slots, rather than “not more than one slot” as required by claim 1 [of the Arnold patent]. It will be readily appreciated that the multi-slotted alignment sleeve of McCartney would not produce nearly so great a lateral alignment force upon a pair of termination pins, as does Applicant’s invention. Furthermore, to eliminate all the slots except one in the alignment sleeve of McCartney would be directly contrary to the teachings of that reference.8

    In the context of validity, Mr. Peterson’s statements are self-serving. If these statements were given any weight, they would compel an interpretation that the ’145 patent discloses only a multi-slotted sleeve of a hexagonal shape intended primarily for guiding hexagonal termination pins. If given weight over the testimony of Messrs. Borsuk and McCartney that the invention more closely resembles prior art, i.e., a circular slotted sleeve, the ’145 patent readily would be held valid. With respect to plaintiff’s claims of infringement, Mr. Peterson’s statements constitute admissions. However, this does not mean the same statements should serve to support plaintiff’s contentions with respect to validity.

    *3433. The ’797 patent (See illustrations 4 and 5.)

    Mr. Borsuk provided a straightforward analysis of what is disclosed by the four claims at issue in the ’797 patent.

    a. Claim 1

    The preamble to the claim refers to a connector for one or more single optical fibers encased within a cable containing a strength member. Thus, the connector is adaptable for either single or bundle fibers.

    The first element discloses a support member comprised of front and rear sections. Mr. Kosmos stated that a support member is equivalent to a yoke and that both can be defined as a frame. Contacts terminated at the ends of optical fibers are mounted in the forward section. The strength member of the cable is mounted in the rear section of the support member.

    The second element relates to means for “releasably mounting” a contact mounted in cavities on the forward section of the yoke. Two hemispherical clamps or caps are placed over the contacts. The contacts are deemed releasably mounted because they can be taken in and out by removing the clamp or cap.

    The third element refers to a means for achieving fixed attachment with a strength member. According to Mr. Borsuk, this element discloses “from an engineering point of view ... a rigid structure or a structure that will carry loads in a fixed manner.” Figure 4 (see illustration 5) depicts this element as an external sleeve that clamps the strength member of a cable between it and an internal sleeve either by crimping or another means for fixed attachment. Mr. Borsuk stated that “The key is the word fixed ... so that the two parts ... [the sleeves] rigidly clamp the strength member so that axial loads will be resisted and ... the material will not slip.” More specifically, he stated that such loads “will be restrained by the structure and thereby transmitted into the shell hardware as opposed to the optical fiber.” Mr. Kos-mos’ testimony was consistent with Mr. Borsuk’s.

    The third element is described as an attachment means and the fourth element discloses a means for “removably mounting” said attachment means. Part 44 depicted in figure 4 {see illustration 5) serves this purpose as it can “slide in and out of [a] ... slot” in the rear section of the support member and rests in a predetermined position relative to the forward section of the yoke. When the flange is inserted within the slot it can resist cable retention loads.

    b. Claim 6

    Claim 6 includes the connector defined in claim 1 and additionally defines the support member. The support member or yoke is a mounting structure that contains two longitudinal legs that extend from the forward section of the yoke to its rear. The legs are spaced apart and thereby define an open cavity in which the optical fibers of the cable may be spread apart and supported.

    c. Claim 9

    The elements of claim 1 form a subset of the elements in claim 9, except with respect to the first element in claim 9. Element 1 of claim 9 requires that the forward and rear section of the elongated yoke be connected by at least one longitudinally extending leg. This language is permissive of a connector containing one or more legs.

    d. Claim 12

    Element 1 of claim 12 refers to an outwardly extending flange affixed to the attachment means or sleeve that secures the strength member of the cable. Element 2 defines a mounting means for the attachment means. This mounting means consists of a longitudinal recess or opening between the two legs in the rear section of the yoke. The recess is open to one side and is intersected by a transverse slot. The flat sides of the outwardly extending flange slidedly engage the opposite flat sides of the transverse slot.

    *344The invention discloses a leg or legs which maintain the optical fibers in what Mr. Borsuk refers to as a “predetermined distance between the strength member retaining means and the contact.” The purpose for this requirement is to assure that the individual fibers do not bend in an uncontrolled fashion. If the fibers were to bend severely, or were curled, light or energy would be lost out the side of the fiber. Maintaining the fiber within a predetermined distance also controls the sharpness of the bend of the fiber as it enters the contact. If the bend is too severe at this point, the glass will be overstressed and may break.

    With respect to the relevant elements in claim 1 so far as they concern the longitudinal legs, defendant established during cross-examination of Mr. Borsuk that the requirement would be satisfied if a fully enclosed circular structure were used to separate the forward and rear sections of the support member or yoke. Defendant also established that when inserted, the outwardly extending flange required by claim 12 rests within a pocket containing two sides and to an extent a bottom. Mr. Borsuk testified that the flange fits tightly in the transverse slot and the walls of the slot are chamfered to assist in the insertion of the flange. Finally, although the attachment means in the particular embodiment depicted in the figures accompanying the ’797 patent utilizes inner- and outer-crimp sleeves or a press fit for securing the cable strength member, the teachings of the claims at issue are not restricted to this method. Claims 1 and 9 require securing the strength member in a non-removable manner, such that the strength members do not slip when the cable is pulled. Any means for achieving this objective is within the teachings of these claims.

    D. Evaluation of prior art

    In evaluating the prior art against each patent’s claims, considered as a whole, the court must scrutinize the prior art for “some teaching, suggestion, or inference

    ... [in the cited reference] or knowledge generally available to one of ordinary skill in the relevant art which would have led one skilled in the art to combine the relevant teachings of ... [the cited references].” Ashland Oil, Inc. v. Delta Resins & Refractories, Inc., 776 F.2d at 297 n. 24 (citation omitted); Litton Industrial Products, Inc. v. Solid State Systems Corp., 755 F.2d 158. “The combined teachings of the prior art as a whole must be con-sidered____” E.W.P. Corp. v. Reliance Universal, Inc., 755 F.2d 898, 907 (Fed.Cir. 1985); accord Lindemann Maschinenfa-brik GmbH v. American Hoist & Derrick Co., 730 F.2d 1452, 1462 (Fed.Cir.1984). The combination, if any, must have been obvious at the time of the invention and cannot be inferred through hindsight. W.L. Gore & Associates v. Garlock, Inc., 721 F.2d at 1553.

    The Federal Circuit has underscored the presumption that a patent statutorily is presumed valid by 35 U.S.C. § 282. E. g., Loctite Corp. v. Ultraseal, Ltd., 781 F. 2d at 872; In re Mark Industries, 751 F.2d 1219, 1225 (Fed.Cir.1984); W.L. Gore & Associates v. Garlock, Inc., 721 F.2d at 1553. Moreover, defendant, as the proponent of invalidity, must sustain its burden by clear and convincing evidence. E.g., Polaroid Corp. v. Eastman Kodak Co., 789 F.2d 1556, 1559 (Fed.Cir.1986); Loctite Corp., 781 F.2d at 872; Lindemann Mas-chinenfabrick GmbH v. American Hoist & Derrick Co., 730 F.2d at 1459.

    Prior to trial defendant submitted its claims chart as requested by the court. Every prior art reference cited in that chart has been analyzed for its teachings and disclosures. The references discussed in this section are those included in the claims chart and several references principally relied on by defendant at trial and in post-trial argument. See Medtronic, Inc. v. Daig Corp., 789 F.2d at 905 & n. 6. This opinion does not analyze specifically all the references cited by the PTO. These other references, however, have been reviewed and do *345not change the analysis of validity.9 For example, the references discussed in this opinion were sufficient to conclude that the claims of the '797 patent for releasable mounting of contacts read on the prior art without treating defendant’s additional references.

    In reviewing all of the references cited to the PTO, the court acknowledged the role of the examiner, “who with the deference ... [owed] government officials [is assumed to have] some expertise in interpreting the references and some familiajrity with the level of skill in the art----” Bausch & Lomb, Inc. v. Barnes Hind/Hydro-curve, Inc., 796 F.2d 443, 447, (Fed. Cir.1986) (citing American Hoist & Derrick Co. v. Sowa & Sons, Inc., 725 F.2d 1350, 1359 (Fed.Cir.), cert. denied, 469 U.S. 821, 105 S.Ct. 95, 83 L.Ed.2d 41 (1984)). Moreover, it has also been recognized in considering the same prior art before the court as before the PTO that “the burden on the party asserting invalidity is more difficult to meet.” Id., at 447 (citing American Hoist, 725 F.2d at 1359).

    As further preface to the following dis-cijssion, the court noted that in those instances when Mr. Kosmos was unable to state that the prior art contained a particular element appearing in the claims at issue, Mr. Borsuk’s testimony that analogues of the elements were not present was more consistent with the claim interpretation and teachings of the patents in suit. Although Mr. Kosmos discharged a Herculean task in a heroic fashion, his testimony fairly cannot be characterized as clear and convincing, nor can the exhibits upon which he relied.

    1. ’182 Patent

    a. Scope and content of the prior art

    Parfitt, published on January 28, 1972, teaches a fiber optic coupler with termination pins containing fiber optic bundles. A rubber pressure ring rests on one pin and, in conjunction with other parts on the coupler, serves to protect the optical surfaces from breakage. The coupler is comprised of five component parts. There are two hollowed termination pins in which the fiber optic bundles are secured with epoxy adhesive. The termination pins are contained within a connector housing consisting of two parts or halves threaded together. When the housing is coupled, the extreme ends of the termination pins extend through the extreme ends of the coupler housing. A rubber pressure ring is installed within a groove on a central portion of one of the termination pins. Of the two side walls of the groove, the wall nearest to the point of abutment of the termination pin is larger. When the termination pins are fully mated, the rubber pressure ring rests against the inside edge of a nut which is one half of a part of the connector housing and is located at the extreme end of the housing. The rubber pressure ring was incorporated in the coupling to avoid breakage for the optical surfaces if it were to be inadvertently over extended.

    United States Patent No. 3,904,269, Sept. 9, 1975 (filed Jan. 28, 1974) (“Lebduska”), teaches a means for providing a separable connection for fiber optic cable. The housing of the connector is intended to disconnect quickly. The housing contains a spring or resilient annular ring means which provides a positive force at the abutment or optical junction of the fiber optic cables and enables the device to withstand tensile loads generated from the rear. Compression of the spring avoids momentary discontinuance of communication.

    The connector is comprised of two identical hollow shells, or outer ferrules, within which which lie two termination pin bodies, or inner ferrules, and two springs. The fiber optic bundles are epoxied within the pins. When the connector shells are mated, the pin bodies move toward the back of the connector shell, thereby compressing *346the spring between the back or extreme end of the termination pin and internal shoulders and flanges which are located at the extreme ends of the connector shells, or outer ferrules. As the springs are compressed, a force is generated which forces the pin bodies into a forward bias position with respect to each other.

    The patent specifications define the intended function of the springs as

    [maintaining] the fiber optic cable faces at the optical junction in close contact with each other ... In addition, [the] spring[s] ... provide a corollary advantage in that any tension applied to the respective cable can be absorbed by [the] spring ... providing a certain ‘give’ without which the fragile cable or the connection may be accidently parted____

    The specifications go on to state that “[t]he resilient biasing of the optical junction assures not only a positive force to maintain the opposing fiber faces together with an accurate spacing therebetween, but enhances the physical integrity of the junction from shock, vibration and tensile loads that may be inadvertently applied to the cables____”

    United States Patent No. 3,408,610, Oct. 29, 1968 (filed Apr. 10, 1967) (“Clarkson”), teaches an electrical connector or coupler for coaxial cables. A resilient pressure ring and ball bearings are used to maintain contact pins in abutment and in a free-turning relationship. The invention discloses a grounding method capable of impeding interference. The connector is comprised of an elongated cylindrical barrel that is designed to connect with coaxial cables. A rotatable sleeve fits into one end of the barrel. Cocentric with the sleeve and surrounding it is a retaining collar which screws onto an externally threaded portion of the barrel end. At a central portion of the sleeve, on its outer surface, lies an annular channel which holds ball bearings. The upper surfaces of the channel are formed at the point at which the end of the barrel meets the retaining collar. An insulated plug is secured within the sleeve. Within the plug there lies a conductive post, the outer end of which has a connecting flat-headed pin. The connecting pin, along with the insulating plug, forms a socket for receiving a male connector of a coaxial cable. At the inner side of the plug, a contact face is attached to the connecting pin and protrudes slightly from the insulating plug. Axial movement of the connecting pin is resisted because the ball bearings prevent movement of the sleeve.

    A second insulated plug is secured in the opposite end of the barrel. A center pin is securely mounted in the second plug. The face of the center pin is rounded and is intended to make tangential contact with the flat face of the first connecting pin. The center pin has an enlarged central portion. Surrounding the pin, between its enlarged portion and the second plug, there is a resilient pressure ring, or O-ring. The O-ring is under slight compression when the assembly is connected, thus maintaining proper contact between the two pins. As finally assembled, the mechanics of the coupler permit the joined center pins to rotate freely.

    A cylindrical coaxial contact surrounds the enlarged portion on the second pin and rests within the coupler’s barrel. Resilient fingers of varying length extend from the contact to the point of abutment of the two pins and rest against the inner wall of the sleeve. Because the contact is secured into the second insulating plug, it cannot rotate. The coaxial contact serves as a grounding mechanism. The resilient fingers eliminate vibration and the resulting electrical noise generated.

    b. Differences between the prior art and claims in issue

    Parfitt—Claim 1 ’182 patent10

    Preamble

    The first part of the preamble to claim 1 of the ’182 patent refers to a termination *347pin assembly for fiber optic cable (“termination pin assembly”). Mr. Kosmos stated that the two hollowed termination pins in Parfitt come within this definition. Mr. Borsuk testified that the termination pin surrounded by the rubber pressure ring constitutes a pin assembly. The remainder of the preamble of the ’182 patent refers to a termination pin assembly for insertion into a connector assembly (“connector assembly”). Mr. Kosmos stated that the connector assembly in Parfitt consists of two hollow shells or nuts threaded together.

    Element 1

    The first element of claim 1 refers first to “a hollow elongated pin body having an axial bore of such a diameter over a predetermined axial dimension from a first end of said pin body so as to accommodate insertion of the optical fiber” (“hollow elongated pin bodies”). According to Mr. Kos-mos, the two pin bodies in Parfitt contain a tapered bore into which optical fibers may be inserted. Parfitt refers to the use of bundle fibers and states that an epoxy adhesive is applied to such fibers before they are mounted into the two termination pins. Mr. Borsuk’s testimony was consistent with Mr. Kosmos’.

    Element 2

    The first part of the second element refers to “a circumferential groove about a central portion of said pin body” (“circumferential groove”). Mr. Kosmos testified that there is a groove machined into the surface of one of the termination pins in Parfitt. Consistent with the claim interpretation, Mr. Kosmos defined a groove as a trough or configuration machined into a surface. On cross-examination he confirmed the presence of the circumferential groove.

    The second part of this element imposes requirements on the groove. It is to contain “first and second side walls, said first side wall being nearest said first end of said pin body” (“groove containing side walls”). Mr. Kosmos identified these side walls in Parfitt. The first side wall is slightly higher than the second side wall, and the channel formed by the two side walls is a circular groove. On cross-examination Mr. Kosmos agreed that the termination pin opposing the depicted pin does not contain a groove.

    Element 3

    The first part of this element refers to “an axially slidable sleeve mounted to slide over at least a portion of the outside surface of said pin body extending from said groove away from” the point of abutment of the pin bodies (“axially slidable sleeve”). Mr. Kosmos testified that the small flanged area or inner surface on the nut comprising one half of the connector shell represents the slidable sleeve in Parfitt. He considered the slidable sleeve the same as the end of the connector half. Because Mr. Kosmos testified that the termination pin body in Parfitt lies between the flanges and is capable of moving rearwardly in relation to the flange, he viewed the sleeve as mounted to slide over at least a portion of the outside surface of the pin body. As for the final requirement of element 3, Mr. Kosmos found that a slidable sleeve is present extending from the groove away from the point of abutment. On cross-examination he continued to testify that Par-fitt contains an axial slidable sleeve mounted to slide over a pin body, extending from the groove away from the first end of the pin body, but conceded that the sleeve is not part of the termination pin assembly in Parfitt.

    *348Mr. Kosmos also testified that the slida-ble sleeve in Parfitt has a different function than the slidable sleeve in the ’182 patent. The sleeve in the ’182 patent holds the retention clip against a ridge in the connector housing, and in Parfitt the slida-ble sleeve is “held in place by ... [a] nut and thread combination.” However, to the extent that the sleeve allegedly permits the termination pin to slide through it, the witness stated that there is a functional similarity between the sleeves in '182 and Par-fitt.

    Mr. Borsuk rejoined that a slidable sleeve is not present in Parfitt because there is no suggestion of a part mounted or “assembled to” the termination pin which allows the pin to slide through it. He could not look to Parfitt for a teaching or suggestion of a slidable sleeve that functions in the manner as in the ’182 patent. Based on the illustrations in Parfitt, it was Mr. Borsuk’s opinion that the cable housing, the extreme end of which allegedly functions as a slidable sleeve, could clamp onto the termination pin containing the pressure ring preventing the pin from moving at all. Thus, the only movement of the termination pin through the housing would occur when the housings were mated, and such pin would only slide through until it was tightened securely.

    Element 4

    The first part of element 4 relates to a “resilient annular ring means installed within ... [a] groove” (“resilient means”). According to Mr. Kosmos, there is a circular ring within a groove in Parfitt styled as a “rubber pressure ring.” Mr. Borsuk also agreed that a ring means was present in a groove.

    The second part of this element requires a ring means “being compressively engaged between said slidable sleeve and said first side wall of the groove” (“resilient means compressively engaged between a slidable sleeve and a groove side wall”). This requirement is not disclosed in Parfitt. Half of the rubber pressure ring is in the groove and therefore against the first side wall of the groove. The other side of the ring will be compressed against the alleged slidable sleeve or small flanged area at the end of the connector half. Mr. Borsuk testified that the rubber pressure ring was not engaged compressively between a slida-ble sleeve because a slidable sleeve is not taught in Parfitt. Mr. Kosmos’ unpersuasive testimony regarding the suggestion of a slidable sleeve in Parfitt militates in favor of the interpretation offered by Mr. Borsuk.

    The closing phrase of element 4 requires that the ring means must “[provide] a resilient force tending to resist axial movement of ... [the abutting] end of said pin body toward said slidable sleeve” (“resilient means that creates a resilient force for resisting axial movement of a pin body”). Mr. Kosmos stated that this requirement was satisfied in Parfitt, but did not identify the precise method or parts which achieved this requirement. The Parfitt publication itself states that the “small rubber washer was incorporated in the coupling to avoid damage to the optical surfaces if it were to be inadvertently overtightened____” Mr. Kosmos interpreted this language to mean that the function of the rubber washer is to release pressure from the optical face and for the ring itself to absorb the transfer of pressure.

    Mr. Kosmos opined, based upon his knowledge as of 1974, that he would have viewed Parfitt in 1974 as a connector for providing a “positive abutment of ... optical surfaces.” He also testified that the “subject matter of” of Parfitt, particularly its pressure ring, operated in the same way as the O-ring the ’182 patent.

    On cross-examination Mr. Kosmos testified that the reason the O-ring was placed in Parfitt was to avoid damage to the optical surfaces if they were inadvertently overtightened. As for the O-ring in the ’182 patent, it was incorporated to achieve axial tolerance relief. Nevertheless, it was Mr. Kosmos’ belief that the result achieved by the rings in Parfitt and the ’182 patent is identical. Both devices avoid damage to optical surfaces.

    *349Defendant sought to define axial tolerance relief on redirect. Mr. Kosmos stated: “It means there is a mechanism for absorbing the various tolerances that are built up between the parts so that when you’re finally in an assembled [position], if you provide axial tolerance relief, you should not have any [tolerance differences], or very minimal.” He established that the overtightening of, and consequent damage to, optic surfaces is a tolerance that Parfitt seeks to eliminate. Accordingly, Mr. Kos-mos concluded that the Parfitt and the ’182 patent “[accomplished] the same thing, using different words.”

    Mr. Borsuk’s testimony refutes Mr. Kos-mos’ statements that the pressure ring in Parfitt either satisfies the requirements of the closing clause of element 4 or achieves the same result as the O-ring in the ‘182 patent. Mr. Borsuk stated that the O-ring in the ’182 patent absorbs a “stack-up of tolerances” or manufacturing flaws. The ring “allows the contacts to always come together in ... [the] same position” regardless of the imprecision in the manufacturing of their parts. This process “normally requires the connectors to mate to a fixed stop,” i.e., the contacts come together in the same position each time.

    In Parfitt, Mr. Borsuk explained, the rubber pressure part is not round, but square.11 Accordingly, the rubber pressure part does not function with spring-type qualities. The pressure part is compressed between the housing and a wall of the pin on which it is located. Because it is not round, it will not deform and change shape and therefore exhibit a spring rate. The square ring is “infinitely stiff or exponentially stiff ... to absorb [overtighten-ing] forces.” As soon as compression is attempted, “[t]he forces generated become very large rapidly____” to prevent over-tightening. In coupling the Parfitt connector, without the pressure washer, the contacts would crush one another. With the ’182 patent, the connector halves join in a fixed position, and the O-ring relieves axial tolerances. It does not serve the function, as is the case with Parfitt, of preventing overtightening when the contacts are joined.

    Finally, Parfitt itself disclaims any attempt to align individual fibers and relied on fairly precise machining tolerances in mating parts for satisfactory alignment.

    Parfitt—Claim 2 ’182 patent

    Preamble

    The first part of the preamble to claim 2 refers to “a fiber optic connector for ... at least one pair of fiber optic cables” (“a connector for a pair of cables”). Mr. Kos-mos stated that this defined all parts of the connector in Parfitt, including cables entering the connector halves at their extreme ends. The second part of the preamble refers to “corresponding guide means for holding ... [the] cables in substantially colinear positions with ... [fibers], in substantial abutment when said shells are mated” (“guide means for holding cables in abutment”). Mr. Kosmos defined the guide means in Parfitt to include the two hollowed termination pins and the two halves comprising the connector housing which thread together. Mr. Borsuk countered on rebuttal that Parfitt contains no guide means within an insert in its housing for holding the termination pins.

    Element 1

    The first element of claim 2 requires “terminating pins corresponding ... to ... first and second cables, said pins having an axial cavity ... for a predetermined axial length, ... said pin end being substantially coextensive with said [cable ends]” (“termination pins with axial cavities”). Mr. Kos-mos referred to two termination pins which are hollow for accepting two fiber optic cables. Mr. Borsuk found that this requirement substantially was satisfied.

    Element 2

    The second element of this claim requires a “means for bonding said fiber [bundles] *350within at least a portion of said axial cavity [of each termination pin] to provide substantial resistance to axial movement of said bundles within said hollow pins” (“means for bonding fiber within the termination pins”). Mr. Kosmos referred to the Parfitt article which states that an epoxy adhesive is applied to the fiber bundles prior to their insertion into the axial cavity in order to maintain the bundles in position. Mr. Borsuk also agreed that Parfitt satisfied this requirement.

    Element 3

    The third element requires “a coaxial sleeve slidably fit about the outside diameter of a central portion of [a termination pin]” (“a slidable sleeve located about a central portion of a termination pin”). Discussing claim 1, Mr. Kosmos testified precisely that the sleeve in Parfitt is the small flanged area or inner surface on the nut comprising one half of the connector shell. In his analysis related to claim 2, he stated more broadly that the slidable sleeve in Parfitt consists of the halves comprising the connector housing or shell. Mr. Borsuk testified that there was no slidable sleeve for the reasons stated in his analysis of claim 1 and Parfitt.

    Element b

    The fourth element refers to a “means associated with the internal cavity of ... the ... connector shells for restraining axial translation of said coaxial sleeve in a direction away from the point of abutment of said cable ends” (“means in the connector housing for restraining movement of the coaxial sleeve away from the point of abutment”). In short, Mr. Kosmos stated that the element defines a method for holding the slidable sleeve in place. Mr. Kos-mos noted that the extreme ends of what he defined as the termination pin in Parfitt contain threads. This threaded portion “goes onto the nut” that he defined as the slidable sleeve in Parfitt. It is the combination of these two parts that restrains the sleeve’s movement. Mr. Borsuk took the position that the element was not satisfied because there are neither sleeves nor inserts in Parfitt, as discussed in the analysis of claim 1 and Parfitt.

    Element 5

    The fifth and final element of claim 2 requires a “means including a resilient member associated with ... [the termination pin body] and said coaxial sleeve to permit relatively small axial translation of said termination pin within said sleeve ... [in order to provide] a compression force at the point of said abutment ... [and] relief for axial manufacturing tolerances” (“means including a resilient member associated with the termination pin, that limits axial translation of the pin within the sleeve, provides a compression force at the point of abutment, and provides relief from axial manufacturing tolerances”). Mr. Kosmos was unclear in defining the parts of Parfitt that limit axial translation of the termination pin and in defining their function. He stated: “My understanding of that is its—the resilient member in Parfitt is the rubber pressure ring. And that second means would be the side wall of the groove____ It’s that action, that side wall and the [slidable] sleeve that contain the rubber pressure ring.” Mr. Kosmos referred to this same unclear testimony to illustrate that a compression force is present at the point of abutment. He was more exact in describing how Parfitt achieves relief from axial manufacturing tolerances. The combination of all the parts of the device perform this function. Upon mating the slidable sleeve [is forced] against the resilient member which forces the abutting pin bodies together. Mr. Kos-mos further stated that without the pressure ring, axial tolerances could not be relieved and the termination pin might not remain in abutment. Consistent with his analysis regarding the final element of claim 1 and Parfitt, Mr. Borsuk again stated that the pressure ring in Parfitt serves only to prevent overtightening and does not relieve axial manufacturing tolerances.

    Lebduska—Claim 1 ’182 patent

    Preamble

    Mr. Kosmos stated that the termination pin assembly in Lebduska includes the two *351pin bodies. Mr. Borsuk again defined a pin assembly as at least two or more parts mounted together. In Lebduska he found no pin assembly, as there are only pins to which there are no parts affixed or associated. Mr. Kosmos testified that the two identical hollow shells or outer ferrules and a precision alignment sleeve surrounding the pin bodies at their point of abutment form the invention’s connector assembly.12 He further stated that he would consider the two outer ferrules as synonymous with a connector assembly. Plaintiff elicited no testimony that would contradict Mr. Kos-mos’ views.

    Element 1

    There are two hollow elongated pin bodies in Lebduska; the bore in the pin is capable of accommodating insertion of an optical fiber. Mr. Borsuk corroborated this testimony on rebuttal.

    Element 2

    Having defined a groove as a trough machined into a surface, Mr. Kosmos was unable to find a circumferential groove in Lebduska. Presumably, the witness would find no groove containing side walls as required by element 2, because he did not find a groove. Mr. Borsuk agreed.

    Element 3

    According to Mr. Kosmos, the axially slidable sleeve of the ’182 patent has its analogue in the internal shoulders or flanges located at the extreme ends of the connector shells or outer ferrules. However, he stated that the sleeve in Lebduska is not mounted to slide over at least a portion of the outside surface of the pin body, as required by element 3. A resilient member or spring lies in between the internal shoulder or flange on the outer ferrule and the back or extreme end of the termination pin. Thus, when the connector is assembled, the fiber optic cable slides through the flange; the pin bodies do not slide through it. Because Mr. Kosmos did not find a groove in Lebduska,13 he determined that it would necessarily follow that the flange or analogous slidable sleeve could not extend from the groove away from the point of abutment. Mr. Kosmos conceded on cross-examination that the termination pin assembly, as he would define it, contains no axially slidable sleeve. Mr. Borsuk stated that there was nothing that he would consider a sleeve, “[Tjhere’s a shell in the view____ But the pin in no way passes through anything. It couldn’t even pass through the opening____” surrounded by the flanges or internal shoulders on the outer ferrule because the pin is bigger than the opening.

    Defendant offered testimony that although there was not a slidable sleeve in Lebduska which satisfied the characteristics of the ’182 patent, or was depicted in illustration 1, the slidable sleeve in Lebdus-ka nevertheless performed the same function. Through Mr. Borsuk plaintiff contended to the contrary that the alleged slidable sleeve, consisting of a flange and narrow section in the outer ferrule in Leb-duska, does not work to avoid overstrain on the optical fibers, as does the sleeve on the ’182 patent. The flange is narrow in width and so does not imply a sleeve. Because the alleged sleeve is just a narrow metal flange, it literally acts like a knife against the edge of the fiber passing through it. In the ’182 patent, the slidable sleeve does not come into contact with the fiber itself; it slides relative to the termination pin. The pin is capable of moving through the sleeve without allowing the fibers or cable to bend on the sleeve.

    *352 Element 4

    Mr. Kosmos stated that Lebduska contains a resilient means (a spring). The specifications in Lebduska state that “resilient material such as a compression coil spring is longitudinally positioned between an internal shoulder on the outer ferrule (28) and the back or extreme end (29) of terminal 30.” Although Mr. Kosmos found a resilient means in Lebduska, he did not find that such means is “installed within a groove,” as required by the first part of the final element of claim 1. Also, although he stated that Lebduska contains a resilient means, a resilient means is not part of the pin assembly as Mr. Kosmos himself defined it for Lebduska. Again, in Lebduska, the absence of a groove precluded Mr. Kosmos from pointing out a resilient means compressively engaged between a slidable sleeve and a groove side wall. He testified, “I can tell you where the spring is going to be compressed. I can define the two surfaces. I have trouble defining the groove.”

    Finally, Mr. Kosmos stated that Lebdus-ka contains a resilient means that creates a resilient force for resisting axial movement of a pin body. He neither identified the parts nor the interaction thereof which effected this function. On cross-examination, plaintiff reminded the witness that all of the elements of claim 1, and particularly this requirement, related to a termination pin assembly. Thus, because Mr. Kosmos defined the termination pin assembly in Lebduska as including only the termination pins and alignment sleeve, he stated that the termination pin assembly in Lebduska necessarily contained neither a groove nor a resilient annular ring means. It was from this premise that Mr. Kosmos was willing to concede that this part of the final element of claim 1 was not satisfied by Lebduska.

    Lebduska—Claim 2 ’182 patent

    Preamble

    According to Mr. Kosmos, Lebduska contains a connector for a pair of cables. He pointed out that only the cables. These cables enter the extreme ends of the entire device comprising the connector. Mr. Kos-mos testified that the outer ferrules are the parts in Lebduska which correspond to the first and second mating connector halves in the ’182 patent. Mr. Borsuk stated that Lebduska satisfied this requirement of the preamble.

    The corresponding guide means for holding cables in abutment includes the outer ferrules comprising the device’s connector shell and the hollow metal sleeve which surrounds the two termination pin bodies at their point of abutment. Thus, according to Mr. Kosmos, there are two guide means. Mr. Kosmos testified that the narrow part or shank of the outer ferrules located at the extreme ends thereof more precisely defined one guide means. He continued to maintain that the hollow sleeve was the other guide means. Mr. Kosmos stated, however, that neither guide means holds the cables in substantially coli-near position with fibers in substantial abutment, in the “same sense” and concluded that the guide means in Lebduska do not meet “all the requirements” set forth in the second part of the preamble to claim 2 of the ’182 patent.

    Mr. Borsuk testified that the guide means in the ’182 patent specifications are “inserts in the connectors that provide internal cavities that provide the removability feature for the contacts.” He also stated that if the contacts have no inserts to guide them, they do not function properly in “shake, rattle and roll type situations.” In Lebduska the device does not teach an insert supporting guide means. Therefore, the contacts and guides freely can move around.

    Element 1

    In Lebduska, the experts agreed, there are two termination pins with axial cavities, otherwise characterized as inner ferrules.

    Element 2

    The specifications in Lebduska refer to a means for bonding fiber within the termination pins: “[A]n epoxy compound or other suitable potting material is inserted ... [into the axial cavities of the inner ferrules *353and] forced in between the interstices of the free ends of the uncovered ... fibers.”

    Element 3

    Mr. Kosmos testified that the ferrule shank, or the narrow bore at the extreme end of the housing or outer ferrule, constitutes a slidable sleeve located about a central portion of a termination pin. Mr. Kos-mos stated that a coaxial slidable sleeve was not present. This testimony is consistent with his testimony regarding claim 1— that there is no slidable sleeve in the termination pin assembly—and Mr. Borsuk’s testimony pertaining to the slidable sleeve as it was referred to in claim 1.

    Element 4

    It is Mr. Kosmos’ position that there is present in Lebduska a means in the connector housing for restraining movement of the coaxial sleeve away from the point of abutment. The requirement is satisfied because the internal shoulder on the outer ferrule or alleged slidable sleeve is part of the outer ferrule itself. Because the outer ferrule is a stationary solid object, the internal shoulder is restrained from movement. On cross-examination Mr. Kosmos conceded that such means could not possibly be present because the invention does not even contain the “basic coaxial sleeve.” Mr. Borsuk stated, “[T]here’s no insert and there’s no restraining means____” for restraining translation of the sleeve.

    Element 5

    Mr. Kosmos testified that Lebduska contains a means including a resilient member associated with the termination pin that limits axial translation of the pin within the sleeve, that provides a compression force at the point of abutment, and that provides relief from axial manufacturing tolerances. He stated that the function of limiting axial translation by causing a compression force at the point of abutment is performed by the back or extreme end of the termination pin and the narrow portion or extreme end of the outer ferrule. These two parts are considered load-bearing surfaces for the resilient means encased between or within them. A compression force is generated as the spring member is compressed between these two parts. The presence of the compression force serves to accommodate axial manufacturing tolerances and thereby termination pin abutment.

    Again, Mr. Kosmos conceded that he had difficulty finding that Lebduska satisfied these requirements given that he had conceded earlier the absence of some of the features integral to this element. Mr. Bor-suk emphasized the absence of a coaxial sleeve and therefore the absence of an association between the spring and the sleeve as required by the claim.

    Clarkson

    Defendant points out that the rotary principles of the Clarkson device cause wear upon contact surfaces and a resilient means is incorporated into the device to achieve positive abutment and prevent such wear. Mr. Kosmos testified concerning the presence of the parts which function as a resilient loading device.

    There are two abutting contacts—one called a conductive post; the other, a center pin. Clarkson’s specifications state that the center pin “[c]ontact is maintained by a resilient pressure ring, ... such as a rubber O-ring or the like____” Mr. Kos-mos stated that the O-ring is compressed between the enlarged central portion of the center pin and the insulated plug which slides over the pin and is held in place in the housing by an internal rib protruding from the inner wall of the housing. He testified that the insulated plug functions in the same way that the slidable sleeve depicted in figure 2 of the '182 patent functions. He also stated that the internal rib on the housing functions as does the retaining clip depicted in figure 2 of the '182 patent.

    On cross-examination testimony was elicited from Mr. Kosmos that Clarkson does not contain the corresponding guide means required by the preamble to the ’182 patent. He also conceded that the pin body in Clarkson is not removable and that the pin bodies in Parfitt and Lebduska are also not removable in the same way as in the ’182 patent.

    *354Mr. Borsuk, on rebuttal, recognized Clarkson as a rotatable connector. As an example, he pointed to a circus carousel: “If I needed to bring electrical connections or electrical wiring out to the horses or the lights on a carousel that are revolving,” Clarkson is the appropriate device to use. Within the connector, the two contacts are in continual motion with respect to each other. He also testified that the real uniqueness of the device is its ability to utilize a coaxial contact with metal fingers of varying lengths to prevent an accidental disconnection caused by vibrations.

    Mr. Borsuk stressed that contact alignment is insignificant in Clarkson. The conductive post has a flat end and the center pin a round head. It is “understood in this type of connector that the [alignment] of the two contacts will not be good, they will not [align], that’s not important and therefore you make the end of the ... [center pin] round, ... so that if they’re off axis a little, slightly on an angle with each other, ... the corner of one would scrape the surface of the other.”

    Although recognizing the presence of a resilient pressure or O-ring in Clarkson which maintains the center pin, Mr. Borsuk said that Clarkson's specifications do not state whether the pin is capable of sliding with respect to the insulating plug. In his view Clarkson differs from both claims 1 and 2 with respect to every element, and he is correct, except insofar as Clarkson is a connector for a pair of cables.

    c. Obviousness of subject matter

    Claim 1 ’182 patent

    Only the preamble and elements 1 and 2 are taught by the prior art references. All the features described in elements 3 and 4 of claim 1 of the '182 patent are not suggested by these references. The record is clear that Parfitt, Lebudska, and Clarkson do not contain the features, functions, or teachings of elements 3 and 4.

    As a matter of feature, function, and teaching, defendant has failed to sustain its burden that elements 3 and 4 of are suggested by Parfitt. First, defendant would define the slidable sleeve in Parfitt as the small flanged area or inner surface on the nut comprising one half of the connector shell. This is inconsistent with the interpretation of the claim language. The preamble refers to a termination pin assembly, and element 3 is intended to be a part of that assembly. Element 3 states that the sleeve is to be mounted to slide over the pin. Mr. Borsuk defined an assembly as a unified object containing two or more affixed parts. In Parfitt, as defendant has defined it, the sleeve is part of the device’s outer shell and is not mounted to slide over the termination pin. Notwithstanding these feature differences, defendant’s basic argument is that Parfitt functions just as the device in the '182 patent, but with fewer parts. Mr. Borsuk testified that the slidable sleeve in Parfitt does not operate as it is intended to in the ’182 patent. Mr. Kosmos himself testified that the extreme end of the housing which allegedly functions as a slidable sleeve is threaded into the corresponding termination pin. Accordingly, Mr. Borsuk recognized that the termination pin may not even slide once it is clamped into the connector housing. He also testified that the pressure washer in Parfitt was square, rather than round, and therefore did not have resilient qualities; instead, it had compressive qualities.

    As for element 4, the parties agreed that Parfitt contains the feature of a resilient means and agreed, as well, that it does not contain the feature of a means compres-sively engaged between a slidable sleeve and groove side wall. The parties were not in agreement concerning the last feature of element 4—resilient means that creates a resilient force for resisting axial movement of a pin body. Plaintiff arguably would agree that the resilient means does preclude axial movement of the pin body, but, according to the testimony, movement is not restricted for the same purpose as in the ’182 patent. The resilient means in the ’182 patent is intended to provide a positive abutment force, to dampen contact vibrations, and to initiate relief from axial manufacturing tolerances by interaction with a *355slidable sleeve. In Parfitt the resilient means serves only as a buffer when the cable housings are joined together to avoid overtightening. The resilient means does not allow the pin to rest, as Mr. McCartney described it, in “a floating intermediate position.” Mr. Kosmos even provided testimony to the effect that Parfitt functioned in a rigid manner when he stated that the alleged slidable sleeve was threaded into the housing.

    Because the prior art contains no teaching, reference, or implication of the features or functions of elements 3 and 4 of claim 1 of the ’182 patent, the subject matter as a whole of claim 1 would not have been obvious at the time the invention was made to a person having ordinary skill in the art.

    Defendant’s strongest argument was that Lebduska could be considered to perform the same functions described by the features in elements 3 and 4. Even assuming this factual predicate to be true, defendant still cannot sustain its burden. Lebduska does not contain a termination pin assembly, a circumferential groove, or a groove containing side walls and therefore does not suggest the features recited in the preamble and element 2 of claim 1. Parfitt discloses these features. However, there is no teaching or suggestion in either Lebduska or Parfitt which would have led one of ordinary skill in the art to combine the teachings of the two references to obtain the combination of elements in claim 1 of the ’182 patent, Ashland Oil, Inc. v. Delta Resins & Refractories, Inc., 776 F.2d at 297 n. 24, or to suggest the desirability—hence the obviousness—Fromson v. Advance Offset Plate, Inc., 755 F.2d 1549, 1556 (Fed.Cir.1985), of designing the termination pin assembly taught by claim 1.

    Although the features set forth in the preamble and element 2 are taught in Par-fitt, the features are not utilized in Parfitt to achieve the function accomplished by the ’182 patent, as a whole. The termination pin assembly and groove containing side walls in Parfitt represent only features that resemble those in the ’182 patent. Parfitt’s termination pin assembly with its groove containing side walls is present to receive a pressure washer, and this pressure washer functions to prevent overtight-ening or avoid damage to the optical faces. According to the testimony, Parfitt does not utilize the termination pin assembly with its groove and pressure ring to permit the termination pin to slide within a sleeve, thereby generating a compressive force to prevent axial translation and promote termination pin abutment. Because the features of the elements of the ’182 patent that are present in Parfitt do not in any way suggest the function or interaction of the elements in the ’182 patent, it is unlikely that the teachings of Parfitt and Lebdus-ka would be combined by one of ordinary skill in the art. See Orthopedic Equipment Co., Inc. v. United States, 707 F.2d 1376, 1382-83 (Fed.Cir.1983). Two other factors militate against the combination being taught: the low level of skill in the art and defendant’s failure to elicit testimony that either Lebduska or Parfitt would have taught or suggested to one of ordinary skill in the art to combine the elements present in the two references.

    Claim 2 ’182 patent

    The parties agreed that the items of significance in the preamble and elements 1 and 2 of claim 2 of the ’182 patent are present in both Parfitt and Lebduska.

    Elements 3 through 5 of claim 2 virtually are identical to elements 3 and 4 of claim 1 of the ’182 patent. The significant difference is that element 5 of claim 2 requires that the device operate to absorb axial manufacturing tolerances. Although this difference is present, the interaction of the features of elements 3 and 4 of claim 1 functions to achieve relief from axial manufacturing tolerances. Both element 4 of claim 1 and element 5 of claim 2 refer to features related to compression forces and axial movement of the pin, but element 5 goes on to state that these features “thereby ... [provide] relief from axial manufacturing tolerances.” Thus, the analysis pertaining to elements 3 and 4 of claim 1 and Parfitt compels the conclusion that Parfitt *356does not contain elements 3 through 5 of claim 2.

    Turning to Lebduska, defendant has not shown by clear and convincing evidence that the features of elements 3 through 5 are suggested in Lebduska. Faced with a record replete with cross-examination concessions by defendant’s expert that the features are not present in Lebduska, this conclusion is unavoidable. The testimony of Mr. Borsuk also showed significant distinctions between the features and teachings of Lebduska and the ’182 patent.

    Not only are the features different, but the functions related at least to elements 3 and 5 are different or not disclosed in Par-fitt. Mr. Kosmos initially testified that the slidable sleeve in Lebduska was the small flange at the extreme end of the device’s connector housing. Recognizing that such testimony would militate against the conclusion that the pin could slide through the sleeve or that the sleeve was “fit about the outside diameter” of the pin, Mr. Kosmos later testified that the sleeve included not only the small flange, but the shank located towards the pin’s extreme end, as well. Such vacillating testimony detracts from the witness’ credibility.

    Mr. Borsuk distinguished the alleged slidable sleeve in Lebduska from the sleeve in the ’182 patent. The optical cable in Lebduska is forced to slide directly over the narrow edge of the extreme end of the alleged sleeve and thereby overstrain the optical fibers. In the ’182 patent, the sleeve fits snugly over the center of the termination pin. The optical cable never comes into contact with the sleeve. Although the significance of this feature is not highlighted in the specifications to the ’182 patent, this feature appears in the claim language.

    Defendant offered the testimony of Messrs. Lebduska and Kosmos for the purpose of illustrating that Lebduska achieves the same functions as the ’182 patent. Because of the disparity in the configuration and number of parts between Lebduska and the ’182 patent, defendant was compelled to minimize, in terms of the function suggested, the disparity between the devices. Defendant tried to suggest that the functions taught in the ’182 patent were also performed in Lebduska, but was unable to muster from Mr. Borsuk the concession that the multitude of parts in the ’182 patent somehow could be consolidated into one and accomplish the relevant functions.

    Clarkson provides no teachings related to elements of claims 1 and 2 that are absent from the Parfitt and Lebduska patents. The ultimate goal to be achieved by the ’182 patent is proper gap and lateral alignment and avoidance of angular misalignment. Plaintiff’s uncontroverted testimony indicates that Clarkson is concerned with one teaching—maintaining abutment. Alignment is of no importance in Clarkson. The adjoining pin ends in Clarkson are round and flat. The patent discloses maintaining abutment and a continuous current while one pin is rotating. Defendant placed no emphasis on the device’s ability to accommodate manufacturing tolerances, absorb vibrations, or react to the imposition of tensile loads. Clarkson is seeking to accomplish problems distinct from those present in the fiber optic art. Alignment, avoidance of fiber breakage, and achieving relief from manufacturing tolerances are irrelevant in Clarkson. Even if Clarkson could be viewed as possessing the features or functions of claims 1 and 2, a person of ordinary skill in the art would not find a teaching in Clarkson that its features or functions could be utilized in Parfitt or Lebduska. Nor do the teachings of Parfitt or Lebduska suggest that the features or functions in Clarkson could be combined with them to satisfy claims 1 and 2 of the ’182 patent.

    d. The following lists are based upon the preceding discussion and designate the presence or absence in Parfitt, Lebduska, and Clarkson of the features, functions, or teachings of the elements set forth in claims 1 and 2 of the '182 patent:

    Parfitt—Claim 1 '182 patent

    Requirements present:

    Preamble

    termination pin assembly

    *357connector assembly

    Element 1

    hollow elongated pin bodies

    Element 2

    circumferential groove containing side walls

    Element 4

    resilient means

    Requirements absent:

    Element 3

    axially slidable sleeve

    Element 4

    resilient means compressively engaged between a slidable sleeve and a groove side wall

    resilient means that creates a resilient force for resisting axial movement of a pin body

    Lebduska—Claim 1 ’182 patent

    Requirements present:

    Preamble

    connector assembly

    Element 1

    hollow elongated pin bodies

    Requirements absent:

    Preamble

    termination pin assembly

    Element 2

    circumferential groove containing side walls

    Element 3

    axially slidable sleeve

    Element 4

    resilient means

    resilient means compressively engaged between a slidable sleeve and a groove side wall

    resilient means that creates a resilient force for resisting axial movement of a pin body

    Clarkson—Claim 1 ’182 patent

    Requirements present:

    Preamble

    termination pin assembly

    connector assembly

    Element 4

    resilient means

    Requirements absent:

    Element 1

    hollow elongated pin bodies

    Element 2

    circumferential groove containing side walls

    Element 3

    axially slidable sleeve

    Element 4

    resilient means compressively engaged between a slidable sleeve and a groove side wall

    resilient means that creates a resilient force for resisting axial movement of a pin body

    Parfitt—Claim 2 ’182 patent

    Requirements present:

    Preamble

    a connector for pair of cables

    Element 1

    termination pins with axial cavities

    Element 2

    means for bonding fiber within the termination pins

    Requirements absent:

    Preamble

    guide means for holding cables in abutment

    Element 3

    a slidable sleeve located about a central portion of a termination pin

    Element 4

    means in the connector housing for restraining movement of the coaxial sleeve away from the point of abutment

    Element 5

    means including a resilient member associated with the termination pin that limits axial translation of pin within the sleeve, that provides a compression force at point of abutment, and that provides relief from axial manufacturing tolerances

    Lebduska—Claim 2 ’182 patent

    Requirements present:

    Preamble

    a connector for pair of cables

    Element 1

    termination pins within axial cavities

    *358 Element 2

    means for bonding fiber within termination pins

    Requirements absent:

    Preamble

    guide means for holding cables in abutment

    Element 3

    a slidable sleeve located about a central portion of a termination pin

    Element 4

    means in the connector housing for restraining movement of the coaxial sleeve away from the point of abutment

    Element 5

    means including a resilient member associated with the termination pin that limits axial translation of pin within the sleeve, that provides a compression force at point of abutment, and that provides relief from axial manufacturing tolerances

    Clarkson—Claim 2 ’182 patent

    Only requirement present:

    Preamble

    connector for a pair of cables

    2. ’145 patent

    a. Scope and content of the prior art

    United States Patent No. 2,109,517, Mar. 1, 1938 (“Xenis”), teaches a connector, consisting of a resilient sleeve, for joining the ends of conducting or non-conducting cables or other parts in an intimate engagement. The connection process can be effected expeditiously if the cables are joined by a sleeve or hollow bore made of metal with inherent resiliency. The sleeve possesses a slot extending its entire length. The bore has a diameter that is smaller than the cable which it is to join. Thus, before the cables are inserted into the sleeve, it must be expanded. The sleeve is expanded during manufacture, and a removable separator device is affixed to it. When the separator is removed during field operation, the connector will contract and grip firmly the contiguous cable ends. The grip caused by the inherent resiliency of the connector effects a substantially permanent joint that is not hand-separable. The grip achieved is conducive to effecting high electrical conductivity. The patent contemplates instances in which the sleeve will come unexpanded and will require expansion by the user in the field.

    United States Patent No. 1,975,885, Oct. 9, 1934 (“Wellman”), discloses a connector consisting of a rubber tube over a resilient metal sleeve which together grip and join insulated electrical conductors. The connector may be readily assembled and disassembled by hand. A terminal will be fixedly attached to each wire to be joined. The terminals have a spherical head preceded by an annular groove. The terminals are inserted into the open ends of a sleeve possessing resilient qualities that is capable of holding the terminals together. Preferably, the sleeve is formed of copper and contains a slot extending its entire length. The sleeve is slid into a rubber insulating tube, which extends a distance over the sleeve ends to prevent it from slipping out of the tube. The overlapping tube teaches an insulating function. The rubber tube has an inside diameter that is smaller than the outside diameter of the sleeve. When the spherical terminal ends are inserted into the sleeve, the inwardly extending de-tents located at the extreme ends of the sleeve interact with the annular grooves in the terminals, thereby retaining them. Because of the size of the rubber tube, the sleeve is prevented from expanding resiliently outward. Thus, the terminals are maintained securely within the sleeve.

    United States Patent No. 3,314,044, Apr. 11, 1967 (“Powell”), teaches a female electrical contact designed to receive a corresponding male contact. The female contact contains expandable slots and twists in such manner that it readily receives a male member, but by a gripping action resists its withdrawal. The connection is still separable. It achieves superior axial alignment of electrical connectors. This invention consists of a female contact comprising a hollow tubular member fabricated from brass or a similar metal. Longitudinally extending slots are machined into the tube *359along a central portion. When the tube is twisted during manufacture, the slots take on a helical- or spiral-like configuration. Initially, the tube is twisted to the point that the central portion of the tube is of a diameter that is smaller than the male contact it receives. The extreme ends of the tube are of a diameter larger than the central portion or portion over which the longitudinal slots extend. To create an electrical connection, a male contact is slid into the chamfered end of the tubular member. The chamfer serves to align the male member for insertion when contact is to be made. During insertion the slots are forced open, and compression forces are exerted upon the male contact. When the male and female contact members are connected, the central portion of the female member expands to a diameter substantially equal to the diameter of the extreme ends of the tube. Removal of the male member is intended to be more difficult than insertion. When withdrawal is attempted, the tube’s central portion reduces in diameter, and by tortional deflection or twisting, the male member is secured tightly-

    United States Patent No. 3,734,594, May 22, 1973 (“Trambarulo”), teaches a rubber sleeve for aligning and securing the ends of two single optic fibers along a common axis within a connector assembly. The connector is also adaptable to receive fiber bundles. The connector consists of a pliable (silicone rubber) hollow core of a uniform diameter. The fiber(s) is inserted at the extreme ends of the core and abuts at its center. Two annular pressure plates are located on opposite sides of the core. Each plate has a center hole of a diameter slightly larger than the fiber(s) inserted through it. When a force is applied to the pressure plates, there is a radially inward displacement of the core material that has the effect of aligning the fibers in a firm, securely fixed position. Displacement of the core when depressed also forces the fibers together and causes a solid contact between the fiber ends.

    Lebduska was also offered by defendant as a prior art reference, and its scope and content are explained fully in the following discussion.

    b. Differences between the prior art and claims at issue

    Lebduska—Claim 1 ’145 patent

    The first part of the preamble to claim one of the '145 patent refers to a “connector for providing a removable ... connection between the ... ends of at least one pair of ... cables” (“connector with a removable connection” and “connector for a pair of cables”). According to Mr. Borsuk, a connector is two halves of a housing that may be mated and unmated in a removable fashion. Mr. Borsuk stated that the requirement related to mating connector members was satisfied by a plug and receptacle. Mr. Kosmos found this requirement was satisfied in Lebduska. The remainder of the preamble refers to cables having optical fibers and termination pins at the ends (“termination pins”). Mr. Kosmos noted termination pins in Lebduska.

    Element 1

    The first element requires a retaining means within said connector for retaining termination pins and therefore terminal ends of cables in axial abutment when the connectors are mated (“means for retaining” and “axial abutment”). Mr. Kosmos stated that this requirement is satisfied in Lebduska by the use of an alignment sleeve.

    Element 2

    The first part of the second element refers to a “means including an alignment sleeve” (“a sleeve”). The sleeve is fabricated “of resilient material ... for effecting lateral alignment of termination pins” (“a sleeve of resilient material” and “a sleeve for lateral alignment”). Mr. Kos-mos stated that there is a sleeve but it is not formed of resilient material. Mr. Bor-suk recognized that the sleeve performs a lateral alignment function, but contains no resilient properties. He explained that Lebduska utilizes a “precision bore ... [with] a diameter all the way through that *360is precisely the size of or a close fit to the contact.” Militating against a determination that the sleeve is resilient is the fact that Lebduska’s specifications state that the sleeve is preferably made of plastic or glass. Glass, according to Mr. Borsuk, has almost no elongation characteristics and therefore has limited resiliency.

    The second part of this element refers to the “sleeve [as] including at least one axially extending slot ... over an axial dimension overlapping the point of abutment” of the pins (“axial slot”). Mr. Kosmos testified that the sleeve in Lebduska contains no slot, although the termination pins do abut. On cross-examination plaintiff established that the sleeve is a solid precision bore-type sleeve. The sleeve contains no slot, but does contain a small hole at the point at which the pins interface. According to Lebduska’s specifications, the hole may be fil'sd with index-matching fluid and subsequently plugged. Mr. Borsuk testified that Lebduska actually teaches away from using a slot. Because the specifications in the patent state that an index-matching fluid may be inserted in the sleeve between the abutting pins, he contended that such fluid would leak if a slotted sleeve were used. On cross-examination he conceded that the specifications of Lebduska merely indicate that an index-matching fluid may be used. Thus, a dry connection is not precluded.

    The remainder of the second element requires that the sleeve be of a diameter over its axial dimension, so that insertion of pins produces resilient lateral expansion of the sleeve “to produce inwardly directed radial forces tending to hold” the pins in lateral alignment (“insertion causing resilient expansion and inward forces to effect lateral alignment"). Mr. Kosmos did not find that the Lebduska patent satisfied this requirement.

    Xenis—Claim 1 ’U5 patent

    Mr. Kosmos found that Xenis consists of a pair of electrical cables, adaptable to fit within a housing, although no housing is shown. Xenis does not have termination pins, but it does have a means for retaining. Mr. Kosmos noted that there is a sleeve of resilient material, containing an axial slot, within which the conductors do not come into contact or axially abut. The sleeve is opened and closed with a separator which is inserted into the sleeve's axial slot.

    During cross-examination plaintiff made the point that the device in Xenis predated fiber optics. Mr. Kosmos agreed that Xen-is does not teach the concept of the insertion of a contact into a sleeve and consequent radial expansion of that sleeve, i.e., the concept of a dynamic expansion. It was also shown that with the tubular member in Xenis, unlike the sleeve in the ’145 patent, a mandrel must be inserted into the member before contacts may be inserted.

    Mr. Kosmos recognized the process by which the sleeve is preconditioned for field use. First, the mandrel is inserted into the sleeve. The mandrel, according to Mr. Kosmos, is in the shape of a contact pin with a tapered end. After the mandrel is inserted, a separator device is placed in the slot of the sleeve. Finally, the mandrel is removed. Rehabilitating his earlier testimony, Mr. Kosmos agreed that insertion of the mandrel does involve a process of “dynamic expansion” or does create the resilient forces referred to in claim 1 of the ’145 patent.

    Defendant also referred Mr. Kosmos to Xenis to reflect upon the art predating Xenis:

    Mechanical means are necessary for expanding my connector as distinguished from the prior art devices in which the connections are adapted to be made and broken by hand operation. The connector may be made of various metals____
    The connector is adapted for use in joining copper or aluminum cables and in fact may be used for joining non-conducting parts as well.

    . Mr. Kosmos testified that viewing Xenis as of 1974, it “would be a possibility” that the device could align optical fibers mounted in termini. Through Mr. Kosmos, plaintiff distinguished Xenis from the ’145 patent in the sense that Xenis effects a substantially *361permanent joint from readily separable connections which are usually adapted to be made and broken by hand operation. Mr. Kosmos conceded that with fiber optics, “You want a separable connector.” To this extent he stated that Xenis teaches away from a non-permanent connection. Finally, Mr. Kosmos was unable to find in Xenis a reference stating that the device was intended to perform an alignment function.

    On rebuttal Mr. Borsuk confirmed that there was no reason to align. “This thing could be crooked, bent, straight, whatever. It’s not one of the factors involved in making electrical connections____” According to Mr. Borsuk, abutment also is not important as the sleeve itself acts as a conductor. Finally, Mr. Borsuk characterized the connection in Xenis as “semi-permanent.” He believed that by using a screwdriver the connection can be broken.

    Wellman—Claim 1 ’145 patent

    Mr. Kosmos testified that Wellman, like Xenis, is comprised of a pair of cables adaptable to fit within a housing. Wellman consists of termination pins and a sleeve of resilient material containing a slot within which the termination pins do not abut. According to Mr. Kosmos, the sleeve’s diameter is such that pin insertion allegedly causes resilient lateral expansion, and inwardly directed forces hold the pins in lateral alignment.

    On cross-examination Mr. Kosmos conceded that in Wellman there is taught a device intended to prevent contact removal. While he agreed that contact removal may not be intended by the device or referred to in the specifications, he took the position that removal is possible. The connection is not permanent since it is not soldered. During his redirect, Mr. Kosmos again was referred to Wellman’s specifications which state:

    When two wires are to be secured together the terminals are inserted in the open ends of the sleeve ... to positions so that the detents ... co-act with the annular groove ... [in the terminals] thereby resiliently securing the terminal in the sleeve____ The tube ... being also circumferentially expanded by the insertion of the terminals materially assists the sleeve in retaining the terminals therein____

    This language does not contain a persuasive suggestion of removability. Finally, Mr. Kosmos also conceded that Well-man’s specifications do not mention specifically that the device is intended to effect lateral alignment.

    Mr. Borsuk emphasized the important function the detents serve in assisting in preventing the terminals from coming out of the sleeve. He agreed with Mr. Kosmos that lateral alignment is not referred to in Wellman. The configuration of the parts even suggests the irrelevance of lateral alignment in an electrical connector. The ends of the terminals are spherical and suggest a pivoting function within the de-tent. During his cross-examination, Mr. Borsuk was directed to language from Wellman’s specifications that the “longitudinally split sleeve [is] formed from resilient metal so as to be readily expansi-ble____” Nevertheless, Mr. Borsuk continued to maintain that only the rubber tube has a resilient function.

    Powell—Claim 1

    Mr. Kosmos stated that Powell contains cables with termination pins and a sleeve of resilient material. The sleeve’s diameter is such that pin insertion causes resilient lateral expansion, and inwardly directed forces allegedly to hold the pins in lateral alignment.

    On cross-examination Mr. Kosmos agreed that Powell does not disclose that it should be used for fiber optics. He said that the device depicts a socket, rather than a sleeve, for accepting a male contact. Mr. Kosmos conceded that the socket is not intended to “effect lateral alignment of terminal ends of any kind of pins or cables.” Mr. Kosmos stated that, in his opinion, a socket is not used for alignment, but a sleeve is, although he stated that both devices have a bore. Regarding the alignment capabilities of the Powell device, Mr. Kosmos was asked:

    *362Q: Nor is there any reference to connecting [two] wires, [two] cables, or [two] fibers, correct, Sir?
    A: It’s a male/female contact..
    Nevertheless, Mr. Kosmos, in response to the court’s inquiry, said that axial or lateral alignment does occur in Powell to the extent that a socket functions to receive a conductor. Powell simply is not directed to the lateral alignment of terminal ends per se.

    Mr. Borsuk sought to distinguish the function of the socket in Powell from that of the sleeve in the ’145 patent. He stated that the male contact “retains itself by having ... [the] helical slots grip and bite in ... against the contact.” Due to the sleeve’s helical shape and deformed center section, the insertion of the male contact causes a resilient force to the right within the funneled-down section. At the same time, the female contact exerts a force to the left, which tends to spread open the receiving contact: “[W]hen the male member ... is being retracted ... from the female ..., the necked-down portion is under tension and hence this portion tends to reduce in diameter____”

    Mr. Borsuk testified that the manner in which Powell functions to grip would not be appropriate for the alignment of optical fibers. He stated that with this type of grip, the contact must be flexible or be able to pivot to be gripped. Without explanation, Mr. Borsuk stated that the tensile forces exerted by Powell could not cause alignment. He did concede that the central part of the sleeve depicted in figure 3 of the ’145 patent (see illustration 2), expands outwardly when a larger diameter pin is inserted; that the sleeve would then exert an inwardly radially directed compression force; and that the sleeve contains helical grooves. Mr. Borsuk, nevertheless, was unable to testify as to whether the sleeve in Powell, with its helical grooves, was manufactured in the same manner as the sleeve in the '145 patent.

    Trambarulo—Claim 1 ’145 patent

    Mr. Kosmos stated that the device contains a part of a housing; an alignment sleeve of resilient material; no terminations pins, but rather two bare optical fibers that contain a core and abut. The sleeve in Trambarulo contains no slot, nor is the use of a slotted sleeve taught by this patent.

    Mr. Borsuk explained that the sleeve is not resilient, as the patent specifications refer to “a deformable annular core” or “any stable pliable material.” The device operates through the application of pressure on plates surrounding the “deformable core.” Mr. Borsuk stated: “When you compress the two plates, it acts like a hydraulic cylinder. Hydraulic cylinders compress fluids and they expand and put equal pressure in all directions and by doing so it closes the inside diameter about the two fibers and compresses them until they [align]....”

    Finally, in Trambarulo, the pin or fiber has a diameter that is smaller than the sleeve. Thus, pin insertion does not cause resilient lateral expansion, and therefore inwardly directed forces, to hold the fibers in lateral alignment.

    Lebduska—Claim 13 ’145 patent

    Element 314

    The second part of element 3 of claim 13 is the only requirement or feature not re*363ferred to in claim 1. It states that the axial section of the sleeve must have a cross-section that is less than that of the pin to be inserted (“a sleeve with a cross-section smaller than the pin”). Mr. Kosmos did not cite Lebduska as satisfying this requirement. Mr. Borsuk explained that the sleeve is formed to fit precisely over the termination pins. Because Lebduska’s specifications suggest that the sleeve be fabricated of glass or plastic, it implied to him that the sleeve lacks resilient characteristics.

    Xenis—Claim 13 ’145 patent

    Mr. Kosmos stated that Xenis contains a sleeve fabricated of resilient spring material. He did not testify as to the presence of an alignment sleeve of resilient material for coupling cable. The sleeve, according to Mr. Kosmos, does contain a slot. Xenis does not refer to a pin per se, but does have a sleeve with a cross-section smaller than the part that the sleeve is intended to hold. Mr. Kosmos testified that Xenis could teach the use of pins. Finally, he testified that the sleeve in Xenis generates internally directed forces into "the conductor, whatever that conductor happens to be,” for the purpose of holding the conductors in place.

    Wellman—Claim 13 ’145 patent

    Mr. Kosmos determined that Wellman contains a sleeve of resilient spring material. He did not state that it was an alignment sleeve for coupling cable. In fact, what is sought to be connected are electric conductors. The sleeve is slotted with a cross-section smaller than the pin. The resilient sleeve in Wellman creates an inwardly directed force on the pin bodies or “cup-shaped terminal,” thereby holding the bodies securely in place.

    Powell—Claim 13 ’145 patent

    Mr. Kosmos testified that Powell contains a sleeve of resilient material. It was not referred to as an alignment sleeve for coupling cable, but, again, the sleeve contains a slot. Powell does contain a sleeve with a cross-section that is smaller then the pin it is intended to receive.

    Trambarulo—Claim 13 ’145 patent

    According to Mr. Kosmos, Trambarulo does not contain an alignment sleeve of resilient material for coupling cable. When fibers are inserted into the sleeve there is no radial expansion; the sleeve compresses only when pressure is placed upon it by plates. Mr. Borsuk determined the presence of a coupling assembly in Trambarulo, but he could not identify a sleeve containing either an axial slot in the sleeve wall or a cross-section smaller than the pin to be inserted. Like Mr. Kosmos, he found the sleeve to be non-resilient. Mr. Borsuk’s testimony supplied the reasoning behind his own, as well as Mr. Kosmos’ determinations. The sleeve in Trambarulo is constructed of a deformable or pliable material. As Mr. Borsuk stated with respect to claim 1, alignment is achieved through the application of hydraulic principles. External pressure on the sleeve is the catalyst in achieving alignment. Thus, Trambarulo does not teach a terminal or pin whose outside diameter is larger than the inside diameter of a resilient member that expands the resilient member resulting in equal and opposite forces to align fibers.

    c. Obviousness of the subject matter

    Even though it has been concluded that the electrical art is analogous for purposes of the ’145 patent based on Moulin I and Mr. Moulin’s testimony, Mr. Kosmos’ testimony undermined, to a significant extent, the pertinency of defendant’s references to the electrical art. It is true that objective factual evidence on obviousness is preferable to expert opinion, Ashland Oil, Inc. v. Delta Resins & Refractories, Inc., 776 F.2d at 294, but the court would be remiss *364in discounting Mr. Kosmos’ testimony insofar as it detracted from the objective evidence relied on by defendant.

    Viewing the prior art as a whole and constrained by considerable testimony unfavorable to defendant, it is concluded that the references cited by defendant teach or disclose the preamble and element 1 of claim 1 of the ’145 patent. It is not disputed that many of the features or functions disclosed by element 2 are present in the prior art. However, defendant did not develop consistent testimony for any of the devices in the prior art references to suggest insertion of a pin or other object into a sleeve causing resilient expansion and inward forces for the specific purpose of effecting lateral alignment.

    Lebduska utilized neither a sleeve of resilient material nor an axial slot to align optic fibers. Instead, Lebduska utilized a precision bore of a diameter that would afford a close fit between pin and sleeve. The parties agreed that the use of resilient forces for alignment was not taught by the device.

    With respect to Xenis, Mr. Kosmos’ testimony was inconsistent. He stated without conviction that there “would be a possibility” that the device could align optic fibers.

    Mr. Kosmos agreed that Wellman’s specifications do not refer to the performance of an alignment function. Defendant did not refute Mr. Borsuk’s testimony that lateral alignment is irrelevant to Wellman because it uses spherical contacts that pivot within the sleeve’s detents. At most, defendant established by clear and convincing evidence the presence of a tube which performs a gripping function, but the testimony and even Wellman’s specifications are not clear on whether the internal slotted metal sleeve itself performs a gripping function other than by the use of detents.

    With respect to Powell, Mr. Kosmos stated that the device utilizes resilient forces, but conceded that the device is a socket and, unlike a sleeve, is not intended to “effect lateral alignment of terminal ends of any kind of pins or cables.” However, Mr. Kosmos did agree that the central part of the sleeve in Powell exhibits the resilient forces referred to in element 2 of claim 1 of the ’145 patent. The patent specifications in Powell state that the female contact “has superior axial alignment characteristics.” Nevertheless, Mr. Borsuk testified, unrefuted by defendant, that the device is not appropriate for the alignment of optical fibers because the inserted contact must be flexible or pivot to be gripped.

    Finally, Trambarulo, the only other prior art reference in the fiber optic field, does not teach the requirements of element 2. Mr. Kosmos maintained no position as to whether the device utilizes resilient forces. He stated that it does contain a resilient sleeve, which is but one feature of element 2. The witness did not refute the more plausible testimony of Mr. Borsuk that the device functions to achieve alignment almost through hydraulic principles.

    It was defendant’s basic argument that Lebduska teaches an alignment sleeve for optic fibers, but not a resilient alignment sleeve exerting compression forces. Thus, it is contended that Trambarulo could be referred to, or would be the motivation, if one sought to modify the sleeve taught in Lebduska to achieve a resilient alignment sleeve exerting such compression forces. Defendant argues that between Lebduska and Trambarulo all of the functions of the sleeve are performed. The sleeve in Tram-barulo, however, is rubber. Thus, to achieve a resilient alignment sleeve from Lebduska, which itself is fabricated of metal, Wellman, Xenis, and Powell would be referred to. Defendant’s arguments rest on the premise that Trambarulo is a reference in the fiber optic art containing a sleeve which performs an alignment function by the exertion of resilient forces. It is from this premise that defendant urges a suggestion or implication in Trambarulo to refer to Wellman, Xenis, and Powell for the existence of a resilient slotted sleeve for optic fibers.

    The testimony supports the conclusion that defendant’s first premise is false and that its second premise therefore does not follow. The function performed in Tram-*365barulo is not equivalent to element 2 of claim 1. Defendant not only failed to present clear and convincing evidence that Trambarulo suggests the claim limitation of claim 1 of the ’145 patent, but also failed to refute plaintiffs arguments that Tram-barulo utilizes hydraulic forces on malleable rubber to achieve alignment. Mr. Bor-suk credibly explained that there is no termination pin insertion achieving resilient lateral expansion of a sleeve to produce inwardly directed forces. At most, the alignment device in Trambarulo produces only inwardly directed forces. Because Trambarulo actually discloses a hydraulic alignment method, plaintiff is without a prior art reference from which it can obtain a teaching of alignment through the use of outward and inward compression forces.

    Wellman, Xenis, and Powell superficially appear to satisfy the requirements of element 2. At most these devices suggest a gripping function, a function which fulfills but a part of that necessary to achieve a fiber optic coupling. These references do not teach the alignment function of the ’145 patent.

    Claim 13

    The preamble and elements of claim 1 are subsumed in the preamble and elements of claim 13 with one exception. Again, pin insertion must cause lateral expansion of a sleeve producing inwardly directed radial forces for the purpose of lateral alignment. Element 3 of claim 13 contains a requirement that facilitates the occurrence of resilient forces. In element 3 the termination pin is required to be of a diameter that is larger than the sleeve in which it is inserted. Element 3 is suggested by several of the prior art references. However, because element 4, as it relates to resilient forces for achieving alignment, is no different than element 2 of claim 1, the same testimony and conclusions that apply to the analysis of claim 1 also apply with respect to claim 13. The conclusion follows that defendant has not sustained its burden as to claim 13.

    d. The following lists are based upon the preceding discussion and designate the presence or absence in Lebduska, Xenis, Wellman, Powell, and Trambarulo of the features, functions, or teachings of the elements set forth in claims 1 and 13 of the ’145 patent:

    Lebduska—claim 1 ’145 patent

    Requirements present:

    Preamble

    connector with a removable connection

    connector for a pair of cables

    termination pins

    Element 1

    means for retaining

    axial abutment

    Element 2

    a sleeve

    a sleeve for lateral alignment

    Requirements absent:

    Element 2

    a sleeve of resilient material

    axial slot

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Xenis—claim 1 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    Element 1

    means for retaining

    Element 2

    a sleeve

    a sleeve of resilient material

    axial slot

    Requirements absent:

    Preamble

    connector with a removal connection

    termination pins

    Element 1

    axial abutment

    Element 2

    a sleeve for lateral alignment

    insertion causing resilient expansion and inward forces effect lateral alignment

    Wellman—claim 1 '145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    *366termination pins

    Element 1

    means for retaining

    Element 2

    a sleeve

    a sleeve of resilient material

    axial slot

    Requirements absent:

    Preamble

    connector with a removable connection

    Element 1

    axial abutment

    Element 2

    a sleeve for lateral alignment

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Powell—claim 1 ’145 patent

    Requirements present:

    Preamble

    termination pins

    Element 1

    means for retaining

    Element 2

    axial slot

    a sleeve

    a sleeve of resilient material

    Requirements absent:

    Preamble

    connector with a removable connection

    connector for a pair of cables

    Element 1

    axial abutment

    Element 2

    a sleeve for lateral alignment

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Trambarulo—claim 1 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    Element 1

    means for retaining

    axial abutment

    Element 2

    a sleeve

    a sleeve for lateral alignment

    Requirements absent:

    Preamble

    connector with a removable connection

    termination pins

    Element 2

    a sleeve of resilient material

    axial slot

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Lebduska—claim 13 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables (cable coupling assembly)

    Element 1

    a sleeve

    Element 2

    termination pins

    Element 4

    axial abutment

    Requirements absent:

    Element 1

    a sleeve of resilient material

    Element 3

    axial slot

    a sleeve with a cross-section smaller than the pin

    Element 4

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Xenis—claim 13 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    Element 1

    a sleeve

    a sleeve of resilient material

    Element 3

    a sleeve with a cross-section smaller than the pin

    axial slot

    Requirements absent:

    Element 2

    termination pins

    Element 4

    axial abutment

    *367insertion causing resilient expansion and inward forces to effect lateral alignment

    Wellman—claim 13 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    Element 1

    a sleeve

    a sleeve of resilient material

    Element 2

    termination pins

    Element 3

    a sleeve with a cross-section smaller than the pin

    axial slot

    Requirements absent:

    Element 4

    axial abutment

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Powell—claim 13 ’145 patent

    Requirements present:

    Element 1

    a sleeve

    a sleeve of resilient material

    Element 2

    termination pins

    Element 3

    a sleeve with a cross-section smaller than the pin

    axial slot

    Requirements absent:

    Preamble

    connector for a pair of cables

    Element 4

    axial abutment

    insertion causing resilient expansion and inward forces to effect lateral alignment

    Trambarulo—claim 13 ’145 patent

    Requirements present:

    Preamble

    connector for a pair of cables

    Element 1

    a sleeve

    Element 4

    axial abutment

    Requirements absent:

    Element 1

    a sleeve of resilient material

    Element 2

    termination pins

    Element 3

    a sleeve with a cross-section smaller than the pin

    axial slot

    Element 4

    insertion causing resilient expansion and inward forces to effect lateral alignment

    3. ’797 patent

    a. Scope and content of the prior art

    British Patent No. 112, 951, Peb. 7, 1918 (“British Patent”), teaches a means to secure electric cable by anchoring the sheathing or protective covering of the cable. The device is intended to support a load at the end of a cable in such a manner such that stress is exerted evenly only on the sheathing and not on the conductors or bore cable. The sheathing is secured in the following manner: The cable sheathing is slit and turned back over itself and then folded between a metal collar that surrounds the cable and an outer ring containing an inwardly extending flange. The sheathing is then clamped securely between the collar, ring, and a shrouding cap, which is threaded and thereby engages with threads located on the aforementioned ring. The stripped conductors are attached to contacts which are secured in an annular piece. The annular piece is threaded and thereby engages with a threaded section located on the portion of the ring furthest from the point on the ring where the clamping procedure occurs. Thus, the stripped contacts rest without stress in a recess within the ring. The specifications teach: “The device will generally be used where it is designed to support a lamp or the like object at the end of a cable and to distribute the strain exerted on the sheathing or protective surface of the cable only, leaving the conductors free.” The specifications *368state further: “The present invention relates to improved means whereby a load ... may be carried by an electrical cable without subjecting the conductors to stress.”

    United States Patent No. 1,152,005, Aug. 31, 1915 (“LeRoy Clark”), teaches a simple means of mechanically coupling two cable lengths together in a water-tight housing and electronically coupling conductors at the ends to such cables in a manner that avoids time-consuming splicing. Each coupling member, of which there are two, is permanently attached to wire strands forming the armor casing of the cable to be connected. Thus, when the two coupling members are attached, the armor wire is intended to sustain tensile strains, thereby preventing the coupling members from coming free from the cable. The invention consists of two coupling members that are hollow and bell shaped. The extreme end of the coupling member is narrow and forms a sleeve. The cable, possessing a plurality of conductors, passes through the sleeve of the coupling member. The conductors are then connected to connector terminals on a circular plate that is located at the face of the coupling member. Each terminal is rigidly mounted on the plate by a threaded portion that engages a similarly threaded aperture in the plate. The terminals are attached to the plate, which, in turn, is attached to the coupling member in a manner that assures a watertight seal. When the cable, containing conductors, is inserted into the extreme end or sleeve of the coupling member, the armor wires are bent back and passed through a plurality of openings in the exterior of the sleeve. The armor wires are then bound against and permanently connected to the sleeve by serving wire. The serving wire is capable of resisting strains equally as well, if not better, than the armor wire. Securing the armor wire renders it unlikely that the coupling member will free itself from the cable when tensile strain is impacted on the cable. The two faces of the conductors are joined in a pin and socket combination and then bolted together.

    German Patent No. 324,384, Aug. 16, 1919 (“Suspended Nipple”), discloses a method for clamping the strain cord of conducting wire betveen a clamp nipple and a screw cap possessing the form of a double shed insulator. Mr. Kosmos described the device basically as a means for hanging fixtures such as a lamp. By this clamping method, means are taught for tightly fastening strain cords while avoiding damage to the conducting wires. The device operates in the following manner: Conducting wire is pulled through a hole at the top of the screw cap. The strain cord surrounding the wire is bent around the flange located at the bottom of the screw cap. The stripped conducting wire is brought through the hollow nipple. The nipple is threaded at a point at which it is inserted into the screw cap and is fastened to threads therein. The clamping effect applied upon the strain cord occurs at the nipple seat—the point on the nipple that presses against the flange on the screw cap when the nipple is screwed into such cap.

    German Patent No. 217,595, Jan. 8, 1910 (“Tension Relief Nipple”), teaches a method for achieving tension relief for electric bulbs. According to Mr. Kosmos, a crimping procedure is adopted for fastening strain cords to a support nipple. Crimping is deemed simple and inexpensive and serves to eliminate the damage inflicted on strain cords secured by screws. The device operates in the following manner: A hollow bushing or nipple is slid around flexible cord. Strain or flex-cords that encase the flexible cord are then bent back over a small flange located at the extreme end of the bushing. The strain cords are fastened to the extreme end of the bushing by a hollow cap fabricated of sheet metal. Once the strain cords are tucked between the cap and flanged portion of the bushing, the cap is crimped around its edge.

    United States Patent No. 3,883,681, May 13, 1975 (“Campbell”), teaches a connector assembly for anchoring a braided shield surrounding electric, metallic, or fiber optic cable in such a manner that a uniform tensile stress is placed on such shield and not the chosen energy conductor. Because *369the braided shield is not required to be clamped in any way, localized tensile stress upon such shield is avoided. The invention consists of a connector mounted through a plain block analogous to a wall. Present is a cable containing at least one conductor of energy that is covered with insulation, packing material, a braided conductive shielding, and a flexible plastic or rubber covering. After the cable is passed through the plain block and lock-nut type object threaded into the block, the packing is removed from the cable leaving a void between the shielding and conductor. A narrow cylindrical fitting member is inserted into the void for a predetermined distance. Due to its size, the braided shield lightly grips the fitting member in a uniform manner. There is a radially extending rim at the end of the fitting member which prevents such member from travel-ling too far inside the braided shielding. A hollow cylindrically shaped retaining member is designed to fit over the braided shielding to the extent such shielding extends over the fitting member. The retaining member lies between the solid block and the extending rim of the fitting member. When a tensile force is exerted from the rear onto the end of the cable, the braided shielding grips uniformly the fitting member, thereby placing all load upon such member. This causes the fitting member to pull within the retaining member. Extensive pulling, however, is avoided because the retaining member rests against the radially extending rim of the fitting member. As the retaining member rests against the plain block, neither the retaining nor the fitting member is free to move.

    United States Patent No. 3,725,845, Apr. 3, 1973 (“Ruf-Nek”), also was offered by defendant as a prior art reference. Its scope and content are described in the following discussion by reference to relevant brochures on the device as it is commercially available. This approach is consistent with defendant’s presentation.

    b. Differences between the prior art and claims at issue

    British Patent—Claim 1 ’797 Patent

    The preamble of this claim refers to “[a] connector for one or more ... optical fibers of a fiber optic cable having a strength member” (“connector for cable having a strength member”). Although the British Patent does not pertain to fiber optics,15 the patent pertains to a connector for cable containing a strength member.

    Element 1

    Element 1 requires a support member having a forward and a rear section (“support member with forward and rear sections”). Mr. Kosmos said that the ring in the British Patent is the equivalent of a support member. He stated that the ring is but one piece, but it contains three integral sections—the ring, flange on the ring, and the piece’s upper body. Mr. Borsuk agreed that the device contains a support member.

    Element 2

    This element refers to a “means on said forward section for releasably mounting a contact terminated to a single optical fiber” (“means for releasably mounting a terminated contact”). According to Mr. Kos-mos, the conductors on this patent are terminated into contacts or terminals in the device’s forward section. He stated that the contacts or terminals are threaded into the forward section and are releasable. Mr. Kosmos maintained that this element was satisfied, but stated that releasability was present because a screw could be removed to release the conductor from the contact, not because the contact itself was releasable.

    Mr. Borsuk testified that the contacts in the British Patent are not releasable. The *370wires or conductors are placed into a hole in the side of the contacts and are then secured in place by a screw that is tightened against the wire. The wire may be released, but the contacts are molded into the front plate which is fabricated of “phenolic” plastic. According to Mr. Borsuk, these contacts are mounted permanently into the plastic. He also stated that in his experience with connectors of this sort, one was confronted with the difficult task of reaching through the back of the connector to get at the screws that fasten the conductors. Mr. Borsuk conceded that the plate in which the contacts are molded can be removed. However, if optic fibers were being used, removal in this manner would bend and wind the fibers, more than likely breaking them.

    Element 3

    This element refers to “a means for making fixed attachment to the strength member of a fiber optic cable” (“means for fixed attachment of the strength member of cable”). Mr. Kosmos stated that the sheathing which encases the cable entering the device constitutes a strength member. He explained the parts and the manner in which they accomplish a fixed attachment and consequently protect the conductors. The sheathing is folded back over a collar; when the threaded shrouding cap is mated with the ring, it forms a “press fit” or “friction fit.” The operable parts performing this clamping function are the threaded shrouding cap, collar, and ring, “and particularly the flange portion of the ring.” Mr. Borsuk conceded that element 3 was satisfied by the British patent.

    On cross-examination Mr. Kosmos stated that based on the patent specifications, he was unable to determine whether the conductors in the British Patent served to sustain tensile loads. Defendant later directed him to two provisions and one claim in the patent illustrating that the conductors are intended to remain free from stress.

    Element It

    Element 4 requires a “means for remov-ably mounting said attachment means on said rear section in a predetermined position relative to said forward section” (“means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section”). Mr. Kos-mos stated that this allegedly was present in the British Patent. Removable mount ing is achieved by unthreading the shrouding cap that is mated with the ring. He also found that the attachment means rests a fixed distance from the contacts. Mr. Borsuk stated that the attachment means was not removable because a part of the attachment means, the cap, is part of the entire device. Although the attachment means cannot be removed, he conceded that the strength member can be released, but it would have to be reclamped upon reinsertion.

    Claim 6 ’797 Patent

    Claim 6 refers to the support member of claim 1 and states that it comprises a yoke with two longitudinally extending legs laterally spaced to define an open cavity in which optical fibers of a cable may be spread apart (“a yoke with two legs creating a cavity in which cable may be spread”). Mr. Kosmos concluded that the upper body of the ring in the British Patent is a structure providing a cavity in which the cables may be spread. His testimony, however, did not refer to the presence of a longitudinally extending leg.

    Claim 9 ’797 Patent16

    Claim 9 differs from claim 1 of the ’797 patent in only one way. Element 1 of claim 9 requires “an elongated yoke having a forward section and a rear section connected by at least one longitudinally extending leg.” Element 1 of claim 1 refers to the yoke as a support member, but contains no *371reference to the longitudinally extending leg. Thus, claim 9 supplements claim 1 by requiring at least one longitudinally extending leg (“support member with at least one leg”). Defendant offered no specific testimony with respect to this claim.

    Claim 12 ’797 Patent

    Claim 12 encompasses claims 1 and 9 and describes in detail the attachment means and the mounting of the attachment means that is referred to in elements 3 and 4 of claim 1. Defendant did not refer specifically to claim 12 and elicited only generalized testimony with respect to its subject matter. The elements of this claim are particularized in the discussion of infringement.

    LeRoy Clark ’797 Patent

    Mr. Kosmos testified that LeRoy Clark contains means for releasably mounting a terminated contact. The device contains conductors or wires terminated into conductor socket set terminals. These terminals are male and female contacts. On the side of the device on which the attachment means lies, the conductors terminate into female contacts. The contacts contain threads and are threaded into an insulating block. According to Mr. Kosmos, the contacts are releasable by virtue of being threaded. He further defined the releasa-bility function by stating that the ends of the conductors are held between a washer and the conductor terminal, thereby facilitating releasability. He also stated that the insulating block in which the terminals lie is removably secured into the support member or shell.

    The device in LeRoy Clark is a connector with “armor wires” that constitute the strength member of a cable. The strength member is folded back over a coupling sleeve and then secured to the sleeve with a serving wire. Mr. Borsuk agreed that this constitutes a means for fixed attachment of the strength member of cable. The serving wire is wrapped about the coupling sleeve and strength member. Mr. Kosmos was uncertain as to how the serving wire was secured and therefore had no opinion as to the removability of the wire. Thus, he could not state that LeRoy Clark contains a means for removably mounting attachment means in the rear section. He did state that the attachment means is mounted in a predetermined position relative to the forward section.

    Mr. Kosmos found a support member with forward and rear sections. The device contains a housing called a coupling member. He disagreed that the part of the coupling member near the coupling sleeve is the rear section. He designated the part of the coupling member in which the insulator block rests as the forward section. Mr. Borsuk recognized that LeRoy Clark contains a support member.

    Mr. Kosmos conceded that there was a possibility that the conductors in LeRoy Clark assisted in sustaining tensile loads and that in fiber optics tensile load on optical fibers is undesirable. Defendant subsequently referred Mr. Kosmos to the patent specifications suggesting that the armor wires, rather than the conductors, are intended to sustain tensile strains.

    Mr. Borsuk testified concerning the re-movability of the attachment means in the LeRoy Clark. He explained that the coupling sleeve and the shell of the device form one piece. The armored wires are simply brought up through holes in the coupling sleeve, “looped back around the sleeve,” and secured with spring wire. Mr. Borsuk stated that this was a permanent attachment “in the sense that it’s not intended to be removed and reused.” By its nature spring wire is a high-tension material, and it is wound so tightly that “it is deformed into a certain shape” and cannot be taken apart without damaging it. The patent specifications also state that “[t]he armor wires ... are bound by a serving wire ... which gives a permanent connection capable of maintaining equal if not greater strains in the armor wire per se....”

    Mr. Kosmos also testified on the releasa-bility feature. He noted that the screw that threads into the insulator block holds or traps a conductor between a washer, around which it is wrapped, and the termi*372nal. The contact is a piece separate from the conductor and is threaded into the insulator block from the opposite side. The terminal and wire do not come out as one piece, so releasable mounting will not occur with a contact terminated to a fiber.

    Mr. Kosmos determined that LeRoy Clark contains a yoke or support member with at least one leg. The support member in the device is the housing or coupling member and is bell-shaped. He stated that this bell-shaped housing could be constructed in the same manner as a single leg. Mr. Borsuk agreed with this testimony.

    Suspended Nipple ’797 Patent

    Mr. Kosmos testified that the device contains conducting wire surrounded by a strength member defined in the patent as a strain cord. It is inferable that the device constitutes a connector for cable having a strength member.

    This device also contains “clamping surfaces that create a clamping on this strain cord.” The clamping surfaces are a part of larger parts of the device, but they represent the points at which the clamping action occurs. The illustration and testimony indicate that when the device’s screw cap is threaded to the nipple, the “screw cap ... presses the strength or strain cord between” an area on the nipple defined as a nipple seat and the end of the screw cap over which the strain cord is bent back. A secure clamp takes place when the screw cap is tightened or threaded onto the nipple. Defendant made no effort to relate the function of the Suspended Nipple to the language of the claim. The device does contain a means for fixed attachment of the strength member of cable, thereby satisfying element 3 of claim 1 and element 4 of claim 9. Mr. Borsuk agreed that the device contains means for fixed attachment of the strength member of cable. Neither the language of this patent nor its drawings relate to or disclose longitudinal legs or a complete support member.

    Mr. Kosmos conceded that the conductors or wire in this device, like the strain cord, carry some tensile load. Mr. Borsuk stated that the device does not contain a support member with forward and rear sections, or means for releasably mounting a terminated contact, or means for remov-ably mounting the attachment means.

    Tension Relief Nipple '797 Patent

    Although there was no testimony concerning any of the elements in claims 1, 6, 9, or 12, defendant’s evidence appeared to be directed generally to elements 1, 3, and 4 of claim 1 and elements 1, 4, and 5 of claim 9.

    The alleged means for fixed attachment of the strength member of cable functions in the following manner: A bushing containing a rim or flange is placed over the end of a cable containing conductors. Flex cords or the outer casing of the cable are then splayed over the rim or flange on the bushing. A hollow cap “or cup-shaped ring” is placed over the conductors protruding from the cord end. The cap is then crimped blocking the flex cord on the rim or flange of the bushing.

    Element 4 of claim 1 and element 5 of claim 9 refer to a means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section. Mr. Kosmos made no reference to a means for removable mounting, but stated broadly that the attachment means lies in a predetermined position relative to the point at which the conductors are “finally attached.” He did not explain as to where the conductors ultimately attach, but stated that the device contains a “yoke or frame” that engages the nipple or bushing and crimped cap.

    Element 1 of both claims 1 and 9 refers to a support member or yoke with forward and rear sections, but no testimony was elicited by defendant as to the qualities of the support member or yoke allegedly present here. Mr. Borsuk conceded that the device contains a yoke with forward and rear sections. From the patent’s illustrations, it would appear that the device suggests a support member with at least one leg.

    *373The preamble to claim 1 refers to a connector for cable having a strength member. Although there was no testimony regarding a strength member per se, the flex cord would satisfy this requirement. Thus, it is inferable, consistent with the preamble to claims 1 and 9, that the Tension Relief Nipple is a connector for cable having a strength member.

    Defendant made no effort on direct to establish that the alleged attachment means of the Tension Relief Nipple was removably mounted. In response to plaintiffs questioning, Mr. Kosmos testified that the Tension Relief Nipple utilizes a crimping method to secure the strength member or flex cord and therefore precludes removability.17 Mr. Borsuk testified that removability is also precluded because the flange to which the strength member is secured is located on the inside of the device’s assembly. Mr. Kosmos stated on cross-examination that he did not know whether the conductors in this device sustained any of the tensile load.

    Mr. Borsuk testified that this device differs from the ’797 patent because the load is carried by a small loop of wire threaded to the device’s nipple. In the ’797 patent the clamping mechanism carries the entire load. He stated that the looped wire acts as a hanger. To illustrate that the flex cord in the Tension Relief Nipple carries the load, defendant referred to the patent’s specifications: “The advantage of this nipple over the existing tension-relief nipple is that... all parts of the cords encircling the conducting wire, including the braiding, drawn around the flange can be used for carrying.”

    Campbell—’797 patent

    Defendant drew out specific testimony regarding this device only as it pertained to elements 3 and 4 of claims 1 and 9, respectively—means for fixed attachment of strength member. Neither the language of this patent nor its illustration suggests that it pertains to any other elements of claims 1, 6, 9, and 12.

    Mr. Kosmos explained that the attachment was achieved in the following manner: A braided shield is moved over a portion of a fitting member or hollow sleeve-shaped object. Once this is accomplished, the fitting member is pressed into a retaining member. He stated that there is a “grabbing force, like a Chinese finger, between ... [the fitting member] and the braided shield.” These parts function to bring about a clamping effect. The braided shield is folded back over the retaining member where the retaining member meets a flange on the fitting member. When the fitting member is pressed into the retaining member, the shield is trapped. At the end of the fitting and retaining members, opposite to where the clamping action arises, lies a structural member serving to fully facilitate clamping.

    Ruf-Nek—Claim 1 ’797 patent

    Ruf-Nek discloses an electrical connector utilized for geophysical oil exploration applications. The device consists of a coupling ring; insert; insert retainer, consisting of two parts taped together; and an adapter. The insert supports and holds contacts; the insert retainer holds the insert in the shell permitting the insertion of the unit into the shell; and the adapter houses the strain relief components. After the adapter is assembled to the shell, it comprises the rear of the shell.

    The advertising brochure for the Ruf-Nek connector lists “the bee line for low level signal applications.” According to Mr. Moulin, this line is a “plastic circular [electrical] connector with a latch mechanism” featuring “quick disconnect, dead-face equipped, built-in strain relief.” He testified that Ruf-Nek utilizes strength members to facilitate strain relief. The electrical cable contains “a combination of strength members which would be comprised of the jacket of the cable, the conductors, [the] strength of the conductors themselves and in some cases stress members built into the cable.”

    *374Referring to the physical exhibit of the Ruf-Nek connector, Mr. Moulin testified that the strain relief utilized is known technically as “tied off or jacket wrapped” strain relief. He stated that the jacket of the cable is doubled back on a cylinder and wrapped with a waxed cord. The fixed wire is removable by removal of the waxed cord. As for the strength of the device, Mr. Moulin stated, “The normal requirement for this type of connector in a field application is about 450 pounds____”

    The device’s insert retainer or spacer lies between the insert block and strain relief and is the point through which wires or conductors travel. It maintains separation between these areas, forces the insert into a forward position in the shell, and provides an area in which excess wire is housed “so that cable stretching does not put tension on the contacts” ultimately fastened to the cables.

    The length of the insert retainer, or the area in which excess wire may lie, is controlled by its being positioned between the face of the adapter (strain relief) and the rear of the insert. The insert retainer is not used to cut or measure wire length, as this is predetermined by the specifications related to the device. Nor is size of the insert retainer determinative of the wire size. However, the wire is certainly longer than the insert retainer.

    When the device is fully assembled, the wires are finished off with a contact, and the contact will rest in the insert. The assembly and wiring instructions for the Ruf-Nek connector explain two types of contacts which may be utilized in the device: “a hermaphroditic type contact ... and a conventional pin socket type contact.” He stated that the inserts in which these contacts lie must be changed if the two types of contacts are interchanged and that the contacts are releasably mounted in the inserts. The assembly and wiring instructions provide that an insertion and removal tool may be utilized for contact insertion and removal. No testimony was elicited as to whether insertion and removal is possible with the contact terminated to the wire.

    Mr. Moulin designated the insert retainer as a device in the connector performing the function of a support member. He testified that the retainer supported the insert on one end and the strain relief assembly on its opposite end. On cross-examination Mr. Moulin stated that the Ruf-Nek connector carries 164 metallic copper wires. Testimony was obtained on whether the load was sustained by the copper wires or by the tied-back covering or jacket:

    Q. It’s your testimony that the jacket is carrying the load?
    A. It’s a combination of the jacket and the conductors.
    Q. Are they [the conductors] carrying the majority of that load?
    A. Again I would have to say if they are equally stressed they would carry the majority of the load.

    Plaintiff also established that the patent issued to Mr. Moulin covering the Ruf-Nek connector discloses a cable-clamping mechanism, but does not disclose any of the methods of strain relief discussed in the assembly and wiring instructions for the device. In particular, the patent contains no reference to the tied-off method. Mr. Moulin said that the relevant strain relief method was not disclosed in the patent because “at the time of this patent ... so many methods of strain relief were so well known and utilized in the electrical connector area that it was not important to make it part of that patent.” He agreed only that with the advent of fiber optics and the fragile nature of optic fibers, the requirements for strain relieving optic fiber cables became more sophisticated. According to Mr. Moulin, the requirements for clamping mechanisms changed with fiber optics only to the extent that they had to accommodate the very fine diameter Kevlar strands. Nevertheless, he stated that the basic clamping mechanism, even insofar «s the number of parts utilized, did not change with the progression to fiber optics.

    *375c. Obviousness of subject matter

    Claim 1—’797 patent

    The preamble, and elements 1 and 2 of claim 1 of the ’797 patent are disclosed in all of the prior art references. Element 3 is suggested only in one prior art reference —Ruf-Nek. The features described in element 4 of claim 1 of the ’797 patent do not appear in and are not taught by any of these references.

    As a matter of feature, function, or teachings, defendant has failed to sustain its burden that element 4 is present in any of the prior art. Element 4 pertains to a means for removably mounting the attachment means in the rear section of the support member in a predetermined position relative to the forward section. It is not disputed that the prior art references each contain an attachment means that lies in a predetermined position. At issue is the concept of removable mounting.

    Defendant offers a very different interpretation of the term removable mounting than is suggested by the claim interpretation, underlying patent specifications, and case law. Defendant suggests that the attachment means may be viewed as removable if the fastened strength member may be unfastened.

    The interpretation of claim 1 reveals that the attachment means in the ’797 patent is intended to be removed intact. It is not intended that upon removal the entire wire, including the strength member, should come free from the attachment means and thereby require refastening when mounting is again desired. The claim interpretation, consistent with the statement of the preferred embodiment, reveals that, when mounting is again desired, the attachment means is intended to contain a flange which is capable of sliding in and out of a well-defined slot in the rear section of the support member. In the ’797 patent, the strength member, once affixed, is not intended to be unfastened from the attachment means.

    The claims of the ’797 patent have been interpreted to equate removable mounting with releasable mounting. Moleculon Research Corp. v. CBS, Inc., 793 F.2d 1261, 1272-73 (Fed.Cir.1986), addressed the issue of how to interpret the term “releasably maintaining.” In Moleculon the patent related to a puzzle which utilized magnets to maintain the continuity of the puzzle. The Federal Circuit stated that magnetic pieces inherently were capable of a releasable disengagement. As in this case in which the specifications refer to a flange-in-slot slida-ble engagement, the preferred embodiment of the patent in issue in Moleculon referred to a “ ‘pop-in snap linkage, or a tongue-in-groove arrangement.’ ” The Federal Circuit upheld the lower court’s reliance upon the preferred embodiment to assist it in defining releasably maintaining.

    Here, the preferred embodiment supports an interpretation of the claim to mean removable and releasable. For every prior art reference, defendant either offered no testimony on the issue or contended that the requirement of removable mounting in element 4 of claim 1 of the ’797 patent is satisfied merely by the disassembly of the attachment means. Because the prior art contains no teaching, reference, or implication of the features or functions of element 4 of the ’797 patent for removable and“ releasable mounting, the subject matter of claim 1 as a whole would not have been obvious at the time the invention was made to a person having ordinary skill in the art.

    Claim 6—’797 patent

    Defendant elicited testimony with respect to one of the features of element 1 of claim 6—its requirement relating to an “open cavity.” There was no testimony about the requirement of two longitudinally extending legs. It appears, however, that the British patent suggests all of the features and functions of element 1, including two longitudinal legs.

    Although the requirements of element 1 are obvious from the prior art, the requirements of the preamble of claim 6 are not satisfied. The preamble refers to “[a] connector as set forth in claim 1.” Not all of the elements of claim 1 were obvious from the prior art. It follows, therefore, that *376the requirements of the preamble to claim 6 are not taught by the prior art.

    Claim 9—'797 patent

    The preamble and elements 1 through 4 are present in the prior art. Because element 5 of claim 9 of the ’797 patent virtually is identical to element 4 of claim 1, the same analysis of claim 1 also applies with respect to claim 9. That earlier analysis compels the conclusion that claim 9 of the '797 patent is not invalid.

    Claim 12—'797 patent

    Several factors militate against a conclusion that claim 12 of the '797 patent is invalid. First, the preamble to claim 12 refers to an “optical fiber connector and cable assembly as set forth in claim 9.” Not all of the elements of claim 9 were viewed as obvious from the prior art. It follows, therefore, that the requirements of the preamble to claim 12 cannot be viewed as suggested by the prior art.

    Element 1 requires an outwardly extending flange. Of the references presented, the British Patent and the Tension Relief Nipple come closest to suggesting one requirement of element 1, but fail to do so. In the British Patent the strength member is folded back over a collar, and several other parts serve to hold a sheathing or strength member in a “press fit” or “friction fit.” The flex cords or strength members are folded over a bushing, and a hollow cap is then crimped over the bushing to hold the strength member in place. Both devices utilize an attachment means that is similar to the one referred to in the claim interpretation of the ’797 patent. Although both devices contain a flange, the flange does not suggest features or functions which are similar to the flange referred to in the ’797 patent.

    In the '797 patent, the flange referred to in element 1 of claim 12 is affixed to the attachment means, but it performs no function related to securing the strength member. The flange in the ’797 patent is a small chamfered part that is intended to slidably fit into a slot that transverses the rear section of a support member. The claim interpretation is that the flange once inserted serves various functions: It receives cable retention loads and translates this tension to the support member or connector shell; it functions to slidably disengage the attachment means without unfastening the strength member; and it enables the attachment means to rest in the rear of the support member in a predetermined position relative to the device’s forward section.

    In the prior art references, the flange, if present, unlike the outwardly extending flange in the ’797 patent, is integrally associated with the attachment means. The prior art contains no part like the outwardly extending flange which performs or suggests the functions referred to in the ’797 claim interpretation.

    Element 2 of claim 12 of the '797 patent refers to the flange required in element 1 of claim 12 and focuses upon defining the slot into which the flange is inserted. The analysis related to element 1 of claim 12 gives a detailed explanation of the significance of the flange in slot relationship in the ’797 patent and shall be deemed to apply with respect to element 2.

    Element 2 of claim 12 states that the slot opens to one side. An examination of the prior art references reveals no feature resembling a slot, let alone teaching or suggesting a slot with an opening to either one or two sides.

    From a review of the scope and content of the prior art and the claim interpretation, it is concluded that defendant has failed to sustain its burden of proving by clear and convincing evidence that the '797 patent is invalid as to claim 12.

    d. The following lists are based upon the preceding discussion and designate the presence or absence in the British Patent, LeRoy Clerk, the Suspended Nipple, the Tension Relief Nipple, Campbell, and Ruf-Nek of the features, functions, or teachings of the elements set forth in Claims 1, 6, 9, and 12 of the ’797 patent:

    British Patent—Claim 1 '797 patent

    Requirements present:

    Preamble

    *377connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 3

    means for fixed attachment of the strength member of cable

    Requirements absent:

    Element 2

    means for releasably mounting a terminated contact

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    LeRoy Clark—Claim 1 ’797 patent

    Requirements present:

    Preamble

    connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 3

    means for fixed attachment of the strength member of cable

    Requirements absent:

    Element 2

    means for releasably mounting a terminated contact

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    Suspended Nipple—Claim 1 ’797 patent

    Requirements present:

    Preamble

    connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 3

    means for fixed attachment of the strength member of cable

    Requirements absent:

    Element 2

    means for releasably mounting a terminated contact

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    Tension Relief Nipple—Claim 1 ’797 patent

    Requirements present:

    Preamble

    connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 3

    means for fixed attachment of the strength member of cable

    Requirements absent:

    Element 2

    means for releasably mounting a terminated contact

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    Campbell—Claim 1 ’797 patent

    Requirements present:

    Preamble

    connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 3

    means for fixed attachment of the strength member of cable Requirements absent:

    Element 2

    means for releasably mounting a terminated contact

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    *378 Ruf-Nek—Claim 1—’797 patent

    Requirements present:

    Preamble

    connector for cable having a strength member

    Element 1

    support member with forward and rear sections

    Element 2

    means for releasably mounting a terminated contact

    Element 3

    means for fixed attachment of the strength member of cable

    Requirements absent:

    Element 4

    means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section

    British Patent—Claim 6 ’797patent18

    Requirements present:

    Element 1

    a yoke with two legs creating a cavity in which cable may be spread

    Requirements absent:

    Preamble

    LeRoy Clark—Claim 6 ’797 patent

    No requirements present

    Tension Relief Nipple Claim 6 ’797 patent

    No requirements present

    Ruf-Nek—Claim 6 ’797 patent

    No requirements present

    British Patent—Claim 9 ’797patent 19

    Requirements present:

    Preamble

    Element l

    support member with at least one leg

    Element 2

    Requirements absent:

    Element 3

    Element 5

    LeRoy Clark—Claim 9 ’797 patent

    Requirements present:

    Preamble

    Element 1

    support member with at least one leg

    Element 2

    Element 4

    Requirements absent:

    Element 3

    Element 5

    Suspended Nipple—Claim 9 ’797 patent

    Requirements present:

    Preamble

    Element 4

    Requirements absent:

    Element 1

    support member with at least one leg

    Element 2

    Element 3

    Element 5

    Tension Relief Nipple—Claim 9 ’797 patent

    Requirements present:

    Preamble

    Element 1

    support member with at least one leg

    Element 2

    Element 4

    Requirements absent:

    Element 3

    Element 5

    Campbell—Claim 9 ’797 patent

    Requirements present:

    Element 4

    Requirements absent:

    Preamble

    Element 1

    support member with at least one leg

    Element 2

    Element 3

    *379 Element 5

    Ruf-Nek—Claim 9 797patent

    Requirements present:

    Preamble

    Element 1

    Element 2

    Element 3

    Element 4

    Requirements absent:

    Element 5

    All Patents—Claim 12

    No requirements present

    E. Objective evidence of secondary considerations

    The parties made showings on the fourth Graham v. John Deere factor. Defendant offered several patents to show that the insights that led to the ’182 and ’145 patents were shared by subsequent patentees. The testimony of defendant’s witness Christopher J. Pleatsikas, an economist, amply supports a finding that plaintiff’s economic success in fiber optics is not striking, given the dominance of the commercial market by AT & T Technologies and AMP, which offer connectors that plaintiff does not manufacture. Although Hughes and plaintiff’s ITT Cannon are the major suppliers for the Government, the growth rate of this market in fiber optic sales is slower than the commercial market. James H. Davis, Director of the Fiber Optic Programs Office of the Naval Seas Systems Command, seconded Mr. Pleatsikas’ views; in addition, Mr. Davis described the Navy’s use of fiber optic devices as experimental and predicted that the state of the art in fiber optic connectors likely will be supplanted by fusion or splicing.

    To show commercial success, plaintiff introduced a license that has been placed under seal. This was solicited from plaintiff by a major conglomerate and entered into in early 1986. The license is nonexclusive, worldwide, and covers the manufacture by the licensee, by another company that is a part of the licensee, a division of another corporation which is associated with the licensee, and by three foreign fiber optic sublicensees. The license is directed primarily to the three patents in issue and covers components, as well as complete assemblies.

    Plaintiff concedes that the initial payment under the license was not significant, but claims that the percentage amount of the running royalty is highly desirable. The license acknowledges this litigation and gives plaintiff a premium if it prevails. Defendant asks the court to draw the reasonable inference that the license agreement was less expensive than defending a lawsuit. Notwithstanding the aptness of defendant’s position, it is noteworthy that one of the most highly touted features of plaintiff’s connectors is the “watch jewel alignment concept” or “jewel ferrule system,” a separately patented feature covering standard watch jewel bearings used for precision mounting a fiber in the end of a contact. Defendant introduced a substantial amount of catalogue and promotional literature which imply that the jewel ferrule system is the major component of its alignment system in plaintiff’s own estimation. That plaintiff’s license does not cover the jewel ferrule patent, however, indicates that the licensed patents themselves, apart from the jewel ferrule system, are perceived as valuable features of plaintiff’s connectors.

    Plaintiff also introduced evidence of present and projected sales of fiber optic connector products under the patents in suit, which indicated increasing sales between 1982-85. For example, between 1981 and 1985, plaintiff’s sales of its Fiber Optic Multi-Channel (“FOMC”) connectors rose from $94,000 to $735,000. In 1983 sales of its contacts were $1,500 and rose to $119,000 by 1985. Future production was presented in loftier terms. For example, over the past three years plaintiff sold $1.5 million of FOMC products to Magna-vox, but obtained an order in March 1986 (when trial began) for $530,000 alone.

    Overall, the evidence on secondary considerations was equivocal. In the court’s view that directed to the prior art was decisive that defendant could not sus*380tain its burden to demonstrate invalidity by clear and convincing evidence.

    II. Infringement

    The Federal Circuit in SRI International v. Matsushita Electric Corp., 775 F.2d 1107, 1122-24 (Fed.Cir.1985) (en banc), directed: “Infringement, literal or by equivalence, is determined by comparing an accused product not with a preferred embodiment described in the specification, or with a commercialized embodiment of the patentee, but with the properly and previously construed claims in suit----” 775 F.2d at 1121 (citations omitted). The claims have been construed without reference to the accused devices; and the task at hand is to apply those claims, as construed, to the accused devices to determine whether the patents are infringed literally or infringed under the doctrine of equivalents. See id. at 1118. The Federal Circuit also stated the familiar rule that the burden of proving infringement by a preponderance of the evidence rests on plaintiff. Id. at 1123 (citing cases). Only if the court determines that the burden has not been discharged with respect to literal infringement does the duty arise to consider whether the facts establish infringement under the doctrine of equivalents. “The doctrine of equivalents raises a fact question determinable upon inquiry into whether a product is substantially the same thing used substantially the same way to achieve substantially the same result.” Id. at 1124 (citing Graver Tank & Manufacturing Co. v. Linde Air Products Co., 339 U.S. 605, 609, 70 S.Ct. 854, 856, 94 L.Ed. 1097 (1950)).

    In post-trial argument, plaintiff took the position that the Hughes accused device depicted as PX-25 infringed claims 1 and 2 of the ’182 patent, those depicted as PX-26 and PX-26A infringed claims 1 and 13 of the ’145 patent, and that depicted as PX-27 infringed claims and 1 and 9 of the ’797 patent all on the basis of the doctrine of equivalents. Plaintiff also asserted that the Hughes devices depicted as PX-28 through 33 literally infringe either all or some combination of claims 1, 6, 9, and 12 of the ’797 patent.

    A. The ’182 patent

    1. Claim 1—PX-25

    Preamble

    The preamble to claim 1 refers to a termination pin assembly 20 for insertion into a connector assembly. Mr. Borsuk testified to the presence of both requirements. An assembly was defined by Mr. Borsuk, “in engineering terms,” as several pieces assembled together or which act in association with each other. In PX-25 a rear-end spacer and Belleville spring are affixed to the device’s termination pin. As for insertion of the pin assembly, the relevant feature of the device is a contact retention clip. He testified that the clip, which is common in the industry, is retained by a reduced shoulder in the connector cavity: “The contact is inserted from the rear in this case, as is in ours, and snaps into position____” Mr. Borsuk’s description of the features and functions of the Hughes device is consistent with the claim interpretation for claim 1 of the ’182 patent.

    Element 1

    This element comprises a hollow elongated pin body. The claim interpretation requires an axial bore which runs almost the full length or the full length of the pin body. Mr. Borsuk did not expressly state that this requirement was satisfied in PX-25, but it is obvious from the exhibit.

    Element 2

    Mr. Borsuk testified that PX-25 contains a circumferential groove about a central portion of said pin body. He stated that the groove in PX-25 was nonexistent: “Grooves don’t exist as physical parts____” He went on to state that the groove in PX-25 “is defined by the face of a spacer nearest the point of interface and a rear spring spacer which rests against a shoulder machined into the termination pin.” *381Thus, the face of the spacer and wall of the rear spring spacer maintain the Belleville washers in place on the pin in PX-25; in the '182 patent the groove maintains the 0-ring in position. Mr. Borsuk testified that the groove in PX-25 is analogous to the groove machined into the termination pin in the ’182 patent that holds the resilient 0-ring. Consistent with the remainder of element 2, one of the walls of PX-25, the rear end of the spacer, is closest to the point of abutment.

    The groove in PX-25 is not precisely consistent with the claim interpretation. Under the claim interpretation, the groove in the ’182 patent is machined partially into the termination pin and a part of its rear wall consists of a shoulder on the slidable sleeve. In PX-25 a groove is not machined into the termination pin, but, rather, is made up of various pieces. Thus, the wall of the rear spring spacer, or alleged slida-ble sleeve, is the only part that the ’182 patent and the device in PX-25 share in common in terms of a groove. Nevertheless, the groove for the Belleville washers in PX-25 is created by parts which are fastened to the termination pin. Another similarity is that the groove in PX-25 does not rely in any respect on the connector housing for formation. More importantly, Mr. Borsuk’s ultimate testimony on the functioning of the device discloses that this groove, although devoid of machined-in qualities, still is substantially the same as the groove in the ’182 patent, used substantially the same way, and achieves substantially the same result. Defendant offered no testimony to controvert this finding.

    Element 3

    Mr. Borsuk testified to the presence of an axially slidable sleeve in PX-25 containing the features specified in element 3. He stated that the rear spring spacer is mounted, as required, to slide over the outside of the pin body. The sleeve “has the ability to slide with respect to it____ [I]f not everything would jam and destroy itself [during mating]____” Mr. Borsuk determined by examination of the relevant physical exhibit that the rear spring spacer has the features and performs the functions of the slidable sleeve described in the claim interpretation of the ’182 patent.

    Element 4

    The first part of element 4 refers to a resilient means, and Mr. Borsuk testified to its presence in PX-25 in the form of Belle-ville springs which are compressible springs of a resilent character. As the Belleville springs lie between the rear spacer or alleged slidable sleeve and the face of the spacer, Mr. Borsuk stated that there is in PX-25, in conformity with the second part of this final element, resilent means compressively engaged between a slidable sleeve and a groove side wall.

    Although the claim interpretation disclosed that an 0-ring is the preferred resi-lent means, Belleville washers were still viewed as acceptable. Since both the groove and slidable sleeve in PX-25 have been found to contain features consistent with the claim interpretation, the compression of the Belleville washers within the groove and in association with the sleeve also must be viewed as consistent with claim 1 of the ’182 patent.

    Mr. Borsuk provided the following testimony concerning the functions performed by the Hughes device depicted in PX-25 as they correspond to the functions referred to in claim 1 of the ’182 patent:

    [I]f I mate these parts in two halves of a connector, ... I believe I testified ... that parts all have tolerances and they must have a means to accommodate those tolerances or they would jam. And we also want them in substantial optical abutment so as the mating contact comes into position, this pin, ... moves to the left and the only way it can do that is to compress these resilient springs because the ring spacer will not move because of the clip. As you can see they are rigidly jammed together into the cavity. And when it does move—when the pin does move toward that end, toward the slida-ble sleeve, a resilient force or the compression on the springs will have an equal and opposite reaction tending to resist that movement. And the intent *382there is to keep the end faces in abut-ment____

    This testimony, coupled with PX-25 and the physical exhibit, supports a finding of infringement by the doctrine of equivalents.

    Claim 2—PX-25

    Preamble

    According to Mr. Borsuk, PX-25 represents a connector for a pair of cables because there is depicted the standard contact retention clip trapped within the reduced diameter of a contact cavity. He also found a guide means for holding cable in abutment. The termination pin has a central bore for containing the cable. The part of the termination pin closest to the point of interface, the spring mechanism and the retention clip of the connector, will cause the cables to be in colinear position with cable ends in substantial abutment.

    Element 1

    This element requires the device to have termination pins with axial cavities. Mr. Borsuk testified that the termination pin in PX-25 depicts a central bore of a predetermined length to accommodate cable. He also stated that the fibers in the device are terminated at the end of the pin at the point of interface.

    Element 2

    Mr. Borsuk testified that element 2 is satisfied by PX-25. There are means for bonding fiber within the termination pins. He stated:

    [T]he Hughes drawings do not show the fibers. But having examined the parts and as I earlier said, I have purchased their parts, and when I purchased them ____I found epoxy throughout the cavity of the contact [adhesively] bonding the fiber so that it would not move axially or be displaced from that co-extensive end.

    Mr. Borsuk’s testimony concerning the presence in PX-25 of the requirements set forth in the preamble and elements 1 and 2 of the 182 patent is consistent with the claim interpretation. This defendant has made no effort to refute.

    Element 3

    Element 3 of claim 2 virtually is the same as element 3 of claim 1, and it refers to the presence of an axially slidable sleeve. The analysis related to element 3 of claim 1 illustrates that the rear spring spacer is substantially the same part performing substantially the same function in substantially the same way as the slidable sleeve described in the claim interpretation related of the '182 patent. Again, Mr. Bor-suk testified that PX-25 satisfied this requirement.

    Element 4

    Element 4 refers to a means in the connector housing for restraining movement of the coaxial sleeve away from the point of abutment. In PX-25 the contact retention clip in the cavity of the connector is restrained from rearward movement by the reduced diameter or shoulder within the connector. The clip has outwardly flared ends. Because the clip is secured against the connector and the rear spacer rests against the clip, the clip would restrain the spacer or alleged sleeve from movement. Mr. Borsuk explained: “[I]f during mating the pin attempts to move to the left everything will jam up, if you will, will be stacked up solid____”

    Based on the claim interpretation of the '182 patent and the weight of evidence, the Hughes device uses the same number of parts with virtually the same configuration as the parts described in the ’182 patent to achieve the effect of restraining movement of the slidable sleeve.

    Element 5

    This element requires a means, including a resilient member associated with the termination pin, that limits axial translation of the pin within the sleeve, provides a compression force at the point of abutment, and provides relief from axial manufacturing tolerances.

    The Belleville springs are the resilient member in PX-25. According to Mr. Bor-suk:

    They are in cross-section the shape of a saucer, like a cup and saucer. And they *383have a hole in the middle such that they fit as annular rings around this pin.
    If I squeeze them, that saucer will flatten. To flatten, it must be resilient. And, therefore, these resilient members associated with both the body of the pin and the sleeve. That means, that they are trapped between the shoulder, the first side wall of the groove, and the second side wall, the face of the [slida-ble] sleeve or rear spring spacer, as Hughes calls it, in a manner that they will give, compress and cause an equal and opposite reaction or force called here a compressive force at the point of abutment of the first and second bundles.

    During its direct examination of Mr. Moulin, defendant attempted to show that the rear spring spacer in the Hughes device could not be viewed as analogous to the slidable sleeve in the ’182 patent. Mr. Moulin testified as to the similarity between the contact pin body of the device in his own patent, United States Patent No. 3,792,416, Feb. 12, 1974 (filed Apr. 3, 1972) ("Moulin II”), and the optic pin body depicted in PX-25. Mr. Moulin referred to a “connector block pin body” in the device and described it as a “plastic housing with a contact cavity molded into it.” Simply stated, it is the outer part of the connector.

    A catalogue containing a subsequent version of the connector in Moulin II illustrates that its connector block is capable of accepting power contacts, coaxial contacts, and optical contacts. The configuration of the fiber optic pin and socket body and alignment tube have been designed to fit into the cavity of the connector block. Mr. Moulin noted, however, that a fiber optic contact is distinguishable from an electrical contact when used in this connector. An electrical contact does not contain Belleville springs. Mr. Moulin also conceded that Moulin II does not contemplate the use of Belleville springs.

    Mr. Moulin testified that, closest to the point of abutment, the pin in Moulin II contains an “actual spherical radius” and its body forms towards its extreme end. At the pin’s center is a shoulder or an “enlarged diameter” (“shoulder”). Mr. Moulin stated: “This enlarged diameter is the front end of the contact retention area____ [A] chamfer [on the shoulder] reacts on the contact retention clip in the same manner that the rear spacer reacts on the contact retention clip____” in PX-25. He stated that the clip and shoulder in Moulin I serve the same function as the clip and rear spacer in PX-25—“to prevent the contact from pushing back through the connector block when force is applied” to the pin from any direction. Finally, Mr. Moulin compared the shoulder in Moulin II with the rear spacer in PX-25. The shoulder is “integrally formed with the pin body,” while the rear spring spacer in PX-25 is a separate piece. (1769).

    Defendant sought to establish, through Mr. Moulin, that the relationship between the shoulder and the retention clip would not differ, so far as function, if the shoulder were a separate piece in contrast to a part molded to the contact:

    [In Moulin II] the force on the end of the pin will react, will cause this shoulder to react against the retention clip and the contact will not move.
    ... [In PX-25] the same force on the end of the contact will allow the entire contact to move to the rear, compressing the stack of Belleville springs between the front spacer and the front face of the rear spacer.

    Mr. Moulin recognized that although the interaction between the shoulder or rear spacer and retention clip does not change in the two devices, there is a difference with respect to the action frontally on the Belle-ville springs.21

    Although the precise purpose for eliciting this testimony was not clear, as a general matter defendant failed to distinguish the rear spacer in the Hughes device from the slidable sleeve in the ’182 patent. *384There is some similarity between the parts in Moulin II and PX-25, but the two devices differ in that the shoulder in the Moulin II is part of the pin, whereas the rear spring spacer in PX-25 slides on the sleeve.

    It is not denied that the rear spacer in PX-25, like the shoulder in Moulin II, interacts with the retention clip to prevent the pin from coming out of the back of the connector. However, because the shoulder in the Moulin II is nonmovable and the rear spacer in PX-25 is movable, the device in PX-25 is capable of performing a function beyond simply preventing pin removal. Because the device in PX-25 contains a movable rear spacer, Mr. Borsuk could testify that the rear spacer in PX-25 possessed features consistent with the claim interpretation for the slidable sleeve of the ’182 patent. Even Mr. Moulin’s testimony illustrates that the presence of the movable spacer in PX-25 fosters the compression of Belleville washers consistent with the functioning of the slidable sleeve in the ’182 patent.

    The testimony and exhibits establish that the Hughes connector depicted in PX-25 is substantially the same device as illustrated in the ’182 patent and performs substantially the same function in substantially the same way.

    B. The ’145 patent

    1. Claim 1—PX-26 and 26A (“PX-26”) 22

    The preamble to claim 1 refers to a connector with a removable connection and a connector for a pair of cables and requires that those cables contain optical fibers terminated to termination pins. Mr. Borsuk stated that the components of PX-26 implied that the Hughes alignment sleeve is a connector satisfying all of these requirements. He referred to the plug and receptacle illustrated in the Hughes device as its mating connector members and testified that cables are contained- in the mating pair. However, he conceded that the device relates to single, rather than bundle, fibers.

    Element 1

    This element comprises a means for retaining termination pins in axial abutment in a connector member. Mr. Borsuk testified that the means for retaining is the retaining clip referred to in the discussion related to PX-25 and is depicted in PX-26. The clip “works in conjunction or in association with the contact [and] connector” and holds the contact in the connector. Finally, he stated that the sleeve in PX-26 axially aligns the contacts as they abut.23

    Element 2

    This element refers to a sleeve of resilient material and a sleeve for lateral alignment. It is required to contain an axial slot and is designed to receive opposing termination pins. Upon insertion of the pin into the sleeve, the sleeve exerts reactionary forces which affect the pin. Thus, in PX-26 there is insertion causing resilient expansion and inward forces to effect lateral alignment.

    Based predominantly upon his inspection of a physical sample of the Hughes alignment sleeve, Mr. Borsuk testified to the presence of a resilient alignment sleeve for effecting lateral alignment in PX-26. He testified that the pin’s diameter is greater than the diameter of the guide sleeve. Because the sleeve contains a slot, the larger pin expands the smaller guide sleeve. *385Therefore, expansion is a manifestation of the sleeve’s resiliency. He stated that the sleeve effects lateral alignment because “[t]hat’s what a resilient alignment sleeve does____” He identified the point of abutment of the termination pins in PX-26 and illustrated that the sleeve overlaps this point and extends outwardly from both sides of this point. The sleeve is also depicted as containing a slot extending the entire length of its wall. The slot extends axially or runs parallel with the axis of the part. Based upon a demonstration conducted with a sample of the Hughes “pin or socket pin and the guide sleeve,” Mr. Borsuk testified that the feature of element 2 related to insertion and resilient reactive forces was satisfied: “[T]he diameter of the pin was larger and it caused a resilient expansion of the guide sleeve and produced forces inwardly upon it to achieve the alignment.”24

    In cross-examining Mr. Moulin, defendant sought to undermine Mr. Borsuk’s testimony by showing that the Hughes device, PX-26, fails to “produce inwardly directly radial forces” on termination pins in conformity with elements 2 and 4 of claims 1 and 13, respectively. (Emphasis added.) Mr. Moulin testified as to what occurs to the configuration of Hughes alignment sleeve, which is sold commercially, when a termination pin is inserted in it. He stated that pin insertion causes the sleeve “to expand its inside diameter but not to do it concentrically. There is a method in which you could build an alignment tube that would expand concentric, but it is not this configuration.” A tube or sleeve expands radially if it “[moves] from a central point equally in all directions.”

    The Hughes sleeve does not expand equally along each radius. In fact, upon pin insertion, the radial point on the sleeve directly opposite the slot does not expand at all. From this point and from the edges of the slot only, the sleeve exerts force on the pin. In the areas on the sleeve between these three points, no force is exerted on the guide bushing. Where no force is exerted by the sleeve, gaps or air space are created between the pin and sleeve.25 Mr. Moulin referred to a physical exhibit to illustrate his testimony that the forces exerted by the Hughes alignment sleeve on the pin and socket contacts were not equal in all directions, i.e., what he viewed as radial.

    The physical exhibit that Mr. Moulin used was an all-aluminum enlarged model of the Hughes alignment system consisting of the guide bushing (stainless steel in the Hughes device) and the alignment tube (beryllium copper in the Hughes device). The guide bushing is the body of the contact behind the chamfered end. Thus, the model demonstrated the forces exerted by the guide bushing and the gaps created by the alignment system.

    Prior to trial defendant had listed photographs of this model, which themselves were enlarged, and Mr. Moulin used them to illustrate his testimony further, since examination of the model itself would not reveal how the forces were distributed. Plaintiff claimed that prior to trial it understood that the photographs depicted the actual Hughes alignment sleeve. At trial plaintiff sought to establish through Mr. *386Moulin that although the Hughes sleeve has three points of force, the force exerted is distributed to both sides of the points. Mr. Moulin could not testify that forces were distributed in this manner.

    On rebuttal plaintiff offered a series of microscopic photographs of the actual Hughes alignment sleeve. In the interval between the conclusion of trial in California at which Mr. Moulin testified and resumption of proceedings in Washington, D.C., Mr. Borsuk supervised the preparation of these photographs of the actual Hughes device. Mr. Borsuk made several cuts through a sleeve and used a zoom microscope to photograph the different cuts at 32 times actual size. Objection was made on the ground that these photographs were ex parte tests, most recently condemned by the Federal Circuit in In re Newman, 782 F.2d 971, 974 (Fed.Cir.1986) (petition for writ of mandamus granted) (citing Wagoner v. Barger, 463 F.2d 1377, 1382 (C.C.P.A. 1972), and Congoleum Industries, Inc. v. Armstrong Cork Co., 319 F.Supp. 714, 716 (E.D.Pa.1970)); see Rosemount, Inc. v. Beckman Instruments, Inc., 727 F.2d 1540, 1549 (Fed.Cir.1984) (ex parte tests excluded when defendant failed to list expert). Wagoner and Congoleum counsel that ex parte tests should be admitted, but accorded scant weight. Plaintiff rejoins that its photographs are no different than Mr. Moulin’s and that impeachment evidence under Appendix G to the rules of this court need not be disclosed before trial.

    The Borsuk photographs have been admitted, but accorded minimal weight for three reasons. First, they were derived ex parte. The photographs purport to represent how the actual Hughes device works and, as such, they constitute test results. Second, consonant with traditions of trial advocacy, Appendix G does not require the pretrial disclosure of impeachment evidence. The photographs technically are rebuttal evidence that would not impeach Mr. Moulin, but would undercut the probativeness of his testimony and rebut the exhibits on which he relied. If plaintiff's view were accepted, all evidence adverse to a line of testimony would be viewed as impeachment. Even if the Borsuk photographs were deemed impeachment evidence, the procedural rules of Appendix G do not supplant substantive law dictating the requisites for the admissibility of evidence. The Federal Circuit has announced a preference for collaborative testing in patent litigation. This direction takes precedence over Appendix G. Third, and most important, the Borsuk photographs are not probative. Having been examined carefully against Mr. Borsuk’s advocacy for what he saw in them, the court finds that the photographs do not show that the alignment sleeve is exerting radially expanding forces on the guide bushing. Mr. McCartney’s testimony on point was equally unpersuasive.

    Mr. Peterson’s statements to the PTO 26 constitute an admission that the ’145 patent teaches or discloses away from a resilient alignment sleeve that contains one slot, that extends the full length of the sleeve, that receives a cylindrical pin, and that is effective for achieving significant lateral alignment. Applying Mr. Borsuk’s testimony about the requirements of the Hughes alignment device depicted in PX-26 to Mr. Peterson’s claim interpretation of the ’145 patent would support a finding that the Hughes device not only fails to satisfy the doctrine of equivalents, but pertains to teachings diametrically contrary to the teachings of the ’145 patent. It is not necessary to go this far.

    Plaintiff’s own interpretation of the relevant claims of the ’145 patent, in conjunction with the testimony of Mr. Moulin, showed that pin insertion into plaintiff’s device results in the exertion of equal forces from every radius on the sleeve, with the exception of the slot. Thus, both the sleeve in the ’145 patent and the Hughes sleeve accomplish alignment, but do so in a different manner with a different device. Mr. Moulin’s testimony was credible, and plaintiff was unable to counter it *387effectively. Even disregarding Mr. Peterson’s damaging statements, the other evidence when viewed in light of plaintiff's claim interpretation through Mr. Borsuk, which has been adopted as the interpretation of claims 1 and 13, does not support a finding that claims 1 and 13 of the '145 patent have been infringed by the doctrine of equivalents.

    C. The ’797 Patent

    1. Claim 1—PX-27

    The preamble to claim 1 refers to a connector for cable having a strength member. Mr. Borsuk pointed out a part of the cable in the illustration which appears to be folded back and integrated with several parts in the rear of the device.27 The strength member is located inside the cable jacket.

    Element 1

    This element comprises a support member with forward and rear sections. Mr. Borsuk described the support member in PX-27 as load brackets which are separable into a forward section, consisting of a mounted contact in an insert, and a rear section in which the strength member of the cable is integrated with various parts described in the element 3 analysis. The forward section threads into the rear section by means of a retaining shaft. Mr. Borsuk said that the rear section would be defined as a load bracket because it carries the load of the cable when the cable retention load is applied.28

    Element 2

    Element 2 requires a means for releas-ably mounting a terminated contact. Although it is not clearly visible in the illustration, Mr. Borsuk stated that there is a spring-loaded contact or pin mounted on a fiber in the forward section of the support member. Thus, a contact is terminated to an optical fiber. He suggested that the contact in the Hughes device is releasably mounted: The forward section of the support member is the insert of the connector and it contains a reduced diameter or internal shoulder that works in conjunction with a contact cavity. The contact cavity has the same retention clip that facilitates re-movability in the '182 patent.29

    Element 3

    Element 3 requires a means for fixed attachment of the strength member of a fiber optic cable. Mr. Borsuk found this element was satisfied in PX-27 based upon his examination of a Hughes drawing styled by Hughes as a “Conn Ass’y, InLine Plug Hermaphroditic, 6 Position.” He explained that the exhibit depicts fixed attachment as occurring in the following manner: The stress member exits a cable and goes away from the cable between a jam nut and locking sleeve and is then wrapped around the locking sleeve. It is then brought underneath the strain relief 0-ring and back up again between the locking sleeve and an inner sleeve. The jam nut is threaded to the inner sleeve and when it is tightened, the jam nut, locking sleeve, and inner sleeve are brought together and thereby fix the Kevlar in place.30

    *388 Element 4

    This element comprises a means for re-movably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section. Mr. Borsuk stated that the rear section of the load bracket contains a separable top load bracket clamp that is bolted into position over the inner sleeve and into the lower portion of the rear section. The bolts may be removed, thereby freeing the entire inner sleeve, including the attachment means intact. The inner sleeve has two flanges. One flange has a larger diameter than the rear section of the load bracket, including the load bracket clamp. The other flange on the inner sleeve is the jam nut and it has a larger diameter than a hole or space in the load bracket. The presence of these two flanges combined with the effect achieved by the bolted load bracket clamp thus prevents the attachment means from moving relative to the load bracket. The attachment means lies in a fixed, predetermined position relative to the front insert. Fiber length thereby is controlled between the point at which it exits the cable to where it enters the contact in the front insert. Mr. Borsuk further testified that “[t]he cable is positioned to control stresses and attenuation of the fiber relative to where the contact is releasably mounted.”31

    Claim 9

    As stated in the discussion of validity, claim 9 differs from claim 1 of the ’797 patent in only one way. Element 1 of claim 9, unlike element 1 of claim 1, requires a yoke with at least one leg. Element 1 of claim 1 refers to the yolk as a support member, and the parties agree that the terms are synonymous, but the element contains no reference to the longitudinally extending leg.32

    Mr. Borsuk testified that PX-27 depicts the Hughes device as a side view. From this reference point, one continuous metal body consisting of two parts threaded together is present. The continuous part extends from the forward section of the support member in which the contacts are mounted to the rear section or load bracket in which the strength member is fastened and mounted. Thus, Mr. Borsuk concluded that element 1 of claim 9 may be read onto the Hughes device in PX-27.

    From Mr. Borsuk’s description, it is found that the Hughes device in PX-27 does not infringe claims 1 and 9 by the doctrine of equivalents. The preamble and elements 1 through 3 of claim 1 and the preamble and elements 1 through 4 of claim 9 of the '797 patent may be read onto the Hughes device. However, the Hughes device contains features that are not consistent with the claim interpretation related to element 4 of claim 1 and element 5 of claim 9 of the ’797 patent. Both of the relevant elements were viewed as synonymous and pertain to a means for remov-ably mounting attachment means. Removable mounting in PX-27 is accomplished by bolting and unbolting. The interpretation of element 4 of claim 1 and element 5 of claim 9 of the ’797 patent is that removable mounting is accomplished in a far less permanent manner. In the ’797 patent, a flange affixed to the attachment means is capable of sliding in and out of a slot.

    The specifications emphasize the capacity of the flange to “slidably engage” into the sides of a transverse slot. Unbolting is not contemplated by the ’797 patent as the means for removably mounting the attach*389ment means, and it cannot be viewed as the use of substantially the same parts as in the ’797 patent, i.e., flange in slot, in substantially the same manner to accomplish substantially the same function.33

    2. Claim 1 PX-28

    From Mr. Borsuk’s description, it is found that the Hughes device in PX-28 infringes the ’797 patent as to claims 1, 6, and 9 by the doctrine of equivalents, but does not infringe as to claim 12. The preamble and elements 1 through 3 of claim 1 and therefore the preamble and elements 2 through 4 of claim 9 are presumed to be present in PX-28 because of the similarity in structure of PX-27 and 28. The Hughes device in PX-28 contains features that are consistent with the claim interpretation of the '797 patent for the requirements of claim 6 and element 1 of claim 9 of longitudinally extending leg(s). Because of the distinctions between the mechanisms in PX-27 and 28 for removable mounting of the attachment means, element 4 of claim 1 and therefore its counterpart in claim 9 are discussed more particularly.

    Element 4

    Mr. Borsuk testified that PX-28 depicts means for removably mounting attachment means in the rear section of the support member in a predetermined position relative to the forward section. Mr. Borsuk stated that PX-28 refers to a continuous structure which therefore defines or fixes the distance between the point at which the optical fiber exits the jacket and the point at which it is terminated to the contact. He distinguished the means for removable mounting in PX-28 from such means in PX-27. According to Mr. Borsuk, the mechanism for facilitating the releasable mounting is located at the point where a jam nut, the rear of which assists in maintaining the strength member, and the rear section of the support member meet. The rear section has flanges that fit in a groove or slot in the jam nut or rear section of the support structure. Mr. Borsuk stated: “[W]e see a detent mechanism, a little pin, that snaps into ... place as you rotationally slide the two pieces together____” Mr. Borsuk explained his testimony by the use of a physical exhibit and described the flange-in-slot connection as a bayonet coupling effect.

    As discussed in the analysis related to PX-27, removable mounting in the ’797 patent, as defined by the claim interpretation and consistent with the specifications, requires the slidable engagement of a flange associated with an attachment means into a slot. Unlike PX-27, PX-28 uses an actual flange-in-slot means for mounting the attachment means. The parts for effecting engagement in PX-28 do not resemble the flange and slot described in the claim interpretation. However, the flange in PX-28 is part of the attachment means and mounts in the rear section in a predetermined position relative to the forward section of the support member. Thus, the means for removable mounting in PX-28 is substantially the same thing as in the ’797 patent, and it performs substantially the function in substantially the same way.

    Claim 6

    Claim 6 of the ’797 patent requires a yoke with two legs creating a cavity in which cable may be spread. Referring to PX-28, Mr. Borsuk testified that the Hughes device depicted contains two legs, as did the physical exhibit, which extend longitudinally or lengthwise between the forward section of the yolk or support member and the rear section of it—the portion maintaining the attachment means. The two longitudinal legs are “off axis” or spread apart. As a result of the separation of the longitudinal legs, a space or cavity is created in which the optic fibers may be spread apart and supported within.

    Claim 9

    Mr. Borsuk testified that the Hughes device in PX-28 contains all the require-*390mente of claim 1. Because he stated with respect to claim 6 that the device contained two longitudinally extending legs, he also testified that the device, consistent with the requirements of claim 9, necessarily contains a support member with at least one leg.

    Claim 12

    Claim 12 is analyzed in detail for the first time in this discussion because defendant did not develop the record on point in its invalidity case.

    Preamble

    Claim 12 requires “an optical fiber connector and cable assembly” (“a connector and cable assembly”). Mr. Borsuk testified that PX-28 depicts an assembly of connector components attached to a cable.

    Element 1

    This element comprises the attachment means including a sleeve surrounding a cable and possessing an outwardly extending flange (“an attachment means including a sleeve and an outwardly extending flange”). Mr. Borsuk testified that the jam nut in PX-28 has a flange on it which extends outward towards the forward section of the support member from the area in which the attachment of the strength member occurs. He did not define a sleeve in PX-28 which surrounds a cable.

    Element 2

    Element 2 pertains to mounting means for the attachment means comprising a longitudinally extending recess in the rear section which opens to one side of the rear section (“mounting means, in the rear section of the support member, for attachment means consisting of a longitudinal recess that opens to one side”). Mr. Borsuk found this requirement satisfied by PX-28. He stated that the longitudinal‘recess is the area between the flanges protruding from the jam nut that serves to accept the flanges located on the extreme end of the rear section of the support member. Mr. Borsuk also testified that at the point at which the flanges are coupled together, there is present a side which is out of view in PX-28. The side which is within view contains the required opening.

    The second part of element 2 requires a transversely extending slot in the rear section of the support member that intersects the longitudinal recess and opens to one side. The sleeve with the flange is mounted in the recess with the flange inserted in the transverse slot (“a transversely extending slot, intersecting the recess, opening to one side for the acceptance of the sleeve and flange attached thereon”). Mr. Bor-suk’s testimony concerning the presence of this feature was ambiguous and not persuasive. He stated:

    I believe the slot transverses the recess; recess being I guess you would say from left to right in the drawing. The slot again in the rear section [of the support member] would be the slot that the ... flanges ... [on the jam nut] are installed in. That slot really is one slot extending at 90 [degrees]—to the cable____ The slot, as we saw, could be considered— made in two portions or there across the recess in the ends of the two legs____

    Without elaboration Mr. Borsuk testified that the transversely extending slot opens to “at least one side.” Mr. Borsuk did describe the parts which he contends may be viewed as a sleeve mounted in a recess and a flange mounted in a slot. Although he had not described the sleeve when it first was referred to in element 1, Mr. Borsuk attempted to describe it for this element and exhibited difficulty in doing so. He stated: “The sleeve again is the strength member means—attaching means or fixing means sleeve, cylindrical part around it, the blue nut as Hughes calls it____” Mr. Borsuk’s testimony on point was rendered less credible when, referring to a physical exhibit embodying PX-28, he stated that the sleeve surrounding the cable was not the jam nut, as he had stated, but rather a part of the extreme end of the support member. Again, he referred to the flange on the jam nut as fitting into the extreme rear of the support member or its alleged transverse slots as satisfying the described flange mounting process.

    *391The engagement of the alleged flange and slot in PX-28 cannot be viewed as occurring in the same way with substantially the same parts as occurs in the ’797 patent. In PX-28 the engagement is defined as bayonet coupling, because the flange on the jam nut enter the slot by rotational action. In the ’797 patent, the flange simply is placed into a well-defined slot that transverses the longitudinal recess in the rear of the support member.

    Claims 1, 6, 9, 12—PX-29

    The parties are in agreement that PX-29 differs from PX-28 in only one respect. For PX-29 the manner in which the alleged attachment means is mounted and whether it is removably mounted remain in issue. Elements 4 and 5 of claims 1 and 9, respectively, and all elements of claim 12 encompass this issue.34 The description of and findings on PX-28 with respect to claim 6 apply to PX-29.

    Mr. Moulin explained the manner in which the rear end of the support member is attached to the area of the attachment means. Three shafts or rods lie between the front end of the support member and the attachment area. The ends of the shaft fit into a strain relief compression nut located at the rear end of the support member. The compression nut contains three slots or indentations. Approaching the end of each shaft there is a reduced diameter portion which slidably fits into these three slots. At their extreme end, the diameter of the rods again increases.

    Mr. Moulin explained the phenomenon by which the rods are mounted into the compression nut:

    The shafts are on the same center line as the three slots; they are sprung open distorted to a position out here, and then released. The spring load that is built into the shaft causes it to move to a central position and it is locked into the slot due to the large diameters on the forward end and the rear end of the shaft.

    With respect to removability, Mr. Moulin stated: “If you have to disassemble the insert and rod assembly from the strain relief, you spring the shafts open____” and the reduced diameter of the rod slides out of the slot.

    The preamble and elements 1 through 8 of claim 1 and the preamble and elements 1 through 4 of claim 9 have been deemed present in PX-29. From Mr. Moulin’s testimony, it is found with respect to the disputed feature that the Hughes device depicted at PX-29 satisfies the requirements of element 4 of claim 1 and element 5 of claim 9 of the ’797 patent. It contains a means for removably mounting attachment means in the rear section of the support member. Because the parts for effecting engagement in PX-29 do not resemble the flange-in-slot arrangement described in the claim interpretation, the Hughes device in PX-29 does not literally infringe claims 1 and 9 of the ’797 patent, but does infringe them on the basis of the doctrine of equivalents.

    Plaintiff made no specific effort to read claim 12 on to the Hughes device in PX-29. However, doing so reveals the dissimilarity between PX-29 and the ’797 patent so far as parts and the manner in which they function. Without direction one can only speculate as to the presence of a sleeve which surrounds a cable. As with the analysis related to PX-28, it is difficult to locate a flange, a longitudinal recess that opens to one side, a transversely extending slot, or a slot in which there is mounted the *392sleeve and flange that reasonably could be considered to correspond with the claim interpretation discussing the parts in the '797 patent that relate to claim 12.

    As far as function and the manner in which that function is performed, there is little similarity between PX-29 and the ’797 patent. The sleeve with its attached flange in the ’797 patent has been construed in the claim interpretation simply to slide into a well-defined slot in the rear section of the support member. So, too, it will slide out consistent with the removability requirement. In PX-29 Mr. Moulin did not describe a separate part serving as a flange; rather, he implied that a reduced diameter of a shaft or the rear section of the support member was tantamount to the flange in ’797. Moreover, the alleged flange in PX-29 is not inserted simply into a well-defined slot. Instead, the shaft on which the alleged flange lies is inserted and retained in the alleged slot by a spring or distortion effect inherent in the rods. Thus, the Hughes device in PX-29 is not substantially the same thing as is depicted in claim 12 of the '797 patent and does not perform substantially the same function in substantially the same way.

    Claim 1—PX-30

    Element 3

    Mr. Borsuk testified that PX-30 contains a means for fixed attachment of strength member of cable. He stated that there is an inner sleeve threaded externally with a nut that surrounds it and is threaded to it. The Kevlar, or strength member, is flowered back and wrapped under an O-ring. The O-ring holds the Kevlar in place. An outer sleeve serves fixedly to attach the Kevlar between the rear shoulder of the flange on the inner sleeve and the outer sleeve itself when the nut is tightened against the assembly.

    Element 4

    According to Mr. Borsuk, PX-30 contains a means for removably mounting the attachment means in the rear section of the support member. A support block with a semi-cylindrical opening which forms a flange is the mounting means. The flange on the block fits into a groove in the attachment means for the strength member. The support block is secured over the attachment means and to the rear section of the support member by bolts. Because the bolts may be removed, Mr. Borsuk testified that the attachment means is removably mounted.

    Claim 9

    Mr. Borsuk stated that the Hughes device in PX-30 contains a support member with at least one longitudinal leg. The leg in PX-30 is a metal plate on which the insert holding the contacts and support block lies. The insert apparently is bolted to the forward section of the leg or yoke and the support block maintaining the attachment means is bolted to its rear section. He stated that the leg is longitudinal because it is in axial alignment with the cable that is spread between the forward and rear sections of the yoke.

    Prom Mr. Borsuk’s description, it is concluded that the Hughes device in PX-30 does not infringe claims 1 and 9 either literally or based upon the doctrine of equivalents. The preamble and elements 1 and 2 of claim 1 and the preamble and elements 2 and 3 of claim 9 were presumed to be present in PX-30 because of the similarity in structure of PX-27 and PX-30. The Hughes device in PX-30 contains features that are consistent with the claim interpretation of the ’797 patent for the relevant requirement in element 1 of claim 9.

    Neither element 3 of claim 1 nor element 4 of claim 9 can be read onto the device in PX-30. The claim interpretation concerning the parts in the ’797 patent that comprise the attachment means is not consistent with the parts which comprise the attachment means in PX-30. The attachment means in PX-30 achieves fixed attachment, as in the ’797 patent, by the clamping of a strength member between various parts and meets these elements under the doctrine of equivalents. However, the claim language and features and functions described in element 4 of claim 1 *393and its counterpart, element 5 of claim 9 of the '797 patent, do not appear in PX-30, so the device does not infringe these claims either literally or by the doctrine of equivalents. Just as in PX-27, the attachment means in PX-30 is bolted down. Accordingly, the analysis pertaining to PX-27 and the issue of whether bolting may be considered removable mounting is applicable equally to PX-30. In PX-30 the permanency of the mounting is even more evident than it is in PX-27 because the illustration depicts two plates covering the attachment means. The uppermost plate even extends to the forward section of the device and it, too, is bolted in and serves to maintain both the forward insert and the attachment means in place. To remove the attachment means, all four bolts must be removed. However, removal of all four bolts not only frees the attachment means, but it also causes the major parts of the entire connector to become disassembled.

    Claim 1—PX-31

    Element 3

    Mr. Borsuk testified that PX-31 contains a means for fixed attachment of strength member of cable. He explained that the device contains an inner and outer sleeve which both surround the cable. The Kevlar strength member lies between the two sleeves and is fastened between them. Mr. Borsuk stated:

    I’m aware from the Hughes drawings that the outer sleeve is crimped to the inner sleeve. Crimping ... [requires a] crimp tool. And you can visualize a pair of [pliers]. If you put it around the outer sleeve and just squeeze, then you deform it such that the two are tightly together, which will grip the kevlar so that it is fixably attached to the members.

    Element 4

    According to Mr. Borsuk, PX-31 contains a means for releasably mounting an attachment means in the rear section of the support member. Mr. Borsuk stated that the removable mounting means operates by principles similar to the releasable mounting means for contacts, such as is described in the ’182 patent and as depicted in all of the Hughes devices. There is a clip which surrounds the inner sleeve of the attachment means. The “rear section of the support member has an internal cavity with a reduced diameter or shoulder that this clip ... is able to latch to---- I can see by the [Hughes] drawings that it is a spring member clip, which is released by moving it radially inward to release____” the attachment means. Finally, because the attachment means is fixed into the rear section, it lies in a predetermined position relative to the forward section of the support member.

    Claim 9

    Referring to the Hughes drawing depicting a multi-channel fiber optic connector that he used in preparing the device depicted in PX-31, Mr. Borsuk stated that the device in PX-31 contains the feature in dispute of element 1 of claim 9. The Hughes drawing depicts a yoke or support member with forward and rear sections containing two longitudinal legs. Therefore, Mr. Borsuk testified, “at least one leg” is present in PX-31.

    The analysis related to element 3 of claim 1 and therefore its counterpart, element 4 of claim 9, supports a finding that these elements read on the device in PX-31. Just as testified to by Mr. Borsuk concerning PX-31, the claim interpretation of the ’797 patent states that the strength member in the ’797 patent is clamped between an external sleeve and an internal sleeve by crimping or some similar method. The Hughes device in PX-31 also contains features that are consistent with the claim interpretation of the ’797 patent for the relevant requirements in element 1 of claim 9.

    Element 4 of claim 1 and element 5 of claim 9 refer broadly to the requirement of removable mounting. The claim interpretation states that a flange-in-slot arrangement is the maimer in which removable mounting is achieved by the device in the ’797 patent. The retention clip method *394of removable mounting in PX-31 cannot be viewed as a method of removal similar to the flange-in-slot arrangement. Mr. Bor-suk even testified that the method of removal of the attachment means in PX-31 virtually is identical to the means by which contacts are releasably mounted in the ’182 patent. Therefore, claims 1 and 9 do not read literally on the Hughes device.

    Nevertheless, the features and functions of the device in PX-31 may be viewed as suggesting a removable mounting means. Similar to the manner in which the flange in the '797 patent slidably engages a slot without the permanency of bolting, the attachment means in PX-31, with its retention clip, slidably engages a shoulder in the rear section of the support member in a non-permanent manner. Accordingly, the removable mounting of the attachment means in PX-31 is the same thing as the invention in the ’797 patent, performing substantially the same function in substantially the same manner.35

    Claims 1 and 9—PX-33

    Although Mr. Borsuk testified that all of the elements in claims 1 and 9 are present in PX-33, the requirements set forth in elements 1, 2, and 4 of claim 1, and the duplicative elements of claim 9 are not satisfied in PX-33.

    Claim 1—Elements 1 and 2

    With the exception of PX-33, all of the Hughes devices contain a terminated contact which lies on the same axis on which the optic fiber exits the retaining means in the rear section of the support member. Also, Mr. Borsuk consistently testified with respect to these Hughes devices that the area in which the contacts lie is the forward section of the support member. However, with respect to PX-33, Mr. Bor-suk testified that a “horizontal line,” in which the contacts lie on an axis that is vertical to the plane on which the fibers exit from the attachment means, is the forward section in the Hughes device. In short, although contrary to testimony with respect to all other Hughes devices, plaintiff now contends with respect to PX-33 that the forward section of the support member may be located on the top of the support member’s frame.

    It is more reasonable to interpret the forward section of the support member as axially aligned with the rear section, rather than on a vertical axis to it. This interpretation is consistent with plaintiff’s own claim interpretation and its emphasis upon maintaining the optical fibers in a predetermined position to control cable bending. Such an interpretation also is consistent with the specifications and drawings in the ’797 patent which imply and explicitly illustrate a support member forward section which carries contacts in axial alignment with the fibers exiting the attachment means in the rear section. Plaintiff elicited no evidence which would support a finding that the location of the forward section of the support member, and consequently the position of the contacts, is insignificant. Moreover, plaintiff shall be held to this more reasonable interpretation in that it adopted such an interpretation with respect to all of the Hughes devices except the one depicted in PX-33.

    *395It is implied in element 2 that the terminated contacts are required to lie in the forward section of the support member. Holding plaintiff to the more reasonable interpretation of the location of the forward section of the support member, it is found that PX-33 satisfies neither the requirements of element 1 nor element 2. PX-33 does not relate to a support member with a forward section or contacts terminated within the forward section within the meaning of the ’797 patent.

    Claim 1—Element 4

    Mr. Borsuk testified that the attachment means in the device in PX-33 is removably mounted because screws hold it into place and may be unthreaded to facilitate removal. For the same reasons discussed with respect to PX-27 and PX-30, the use of screws for dismounting cannot be viewed as substantially the same device used in the ’797 patent for removal, used in substantially the same way to achieve substantially the same result.

    Elements 1, 2, and 4 cannot be read onto PX-33, and plaintiff cannot be viewed as having supplied evidence that the features in PX-33 are the same thing as in the ’797 patent, performing substantially the same function in substantially the same way. Accordingly, plaintiff has not sustained its burden of proving by a preponderance of the evidence that PX-33 infringes claims 1 and 9 either literally or by the doctrine of equivalents.36

    CONCLUSION

    Based on the foregoing, it is found and concluded that the accused device depicted as PX-25 infringes the ’182 patent and that those devices depicted as PX-28, 29, 31, and 32 infringe the ’797 patent. Plaintiff has not proved that the ’145 patent is infringed by the devices depicted as PX-26 and 26A or that those devices depicted as PX-27, 30, and 33 infringe the ’797 patent. It is concluded that defendant has not sustained its defense that the subject patents are invalid.

    IT IS SO ORDERED.

    *396[[Image here]]

    *397[[Image here]]

    *398[[Image here]]

    *399[[Image here]]

    *400[[Image here]]

    . Testifying for defendant, Robert L. Lebduska, inventor of several patents for fiber optic connectors, conceded that the lateral alignment requirement of fiber optic art was one of the reasons for his statement in the specifications of his patent on fiber optic connectors: "Although the electrical connector art may, at first blush, appear to be similar ... electrical type connectors tire not able to provide satisfactory performance for optical applications."

    . Mr. Lebduska testified that in testing fiber optic cables, the standards and tests for electrical cables originally were utilized. There were, however, tests distinct to fiber optics. In testing fibers for breakage as a result of stress, the testimony was not clear as to what standards were applied. He did state, however, that it was conceivable that a copper conductor may not have been able to survive some of the tests for optic fibers. The Electronics Industry Association has developed separate standards for the fiber optic field, as well as a glossary of terminology for fiber optics.

    . Defendant also introduced United States Patent No. 3,904,269, Sept. 9, 1975 (filed Jan. 28, 1974) ("Lebduska"), for which Mr. Lebduska was the inventor. Lebduska is another example of an electrical connector modified into a fiber optic connector.

    . TN 2367 refers to a "CONHEX connector" made by Sealectro Corporation. The device Mr. Thiel demonstrated, whether it was Coming’s or Sealectro’s, consisted of three parts—two terminal sections containing a fiber which was threaded on the opposite ends of a center union section. In conducting tests of this device, Mr. Lebduska was aware that the union must occur in a controlled fashion. The fibers had to be “as close as possible” without touching for efficient light coupling. Mr. Lebduska testified that the importance of achieving efficient light coupling was known well before 1973.

    . The preamble to claim 1 states that the termination pin assembly is to be inserted into the *336connector assembly. The preamble to claim 2 states that the connector assembly provides for a removable connection between the ends of fiber optic cables. At trial plaintiff asserted that the insertion and removability function of the T82 patent are its most noteworthy features. These features enable one to remove and replace both fiber optic contacts and electrical contacts in existing connectors. Defendant contended that these features should not be touted as novel because they are not mentioned anywhere in the patent specifications.

    In post-trial argument, defendant argued that the concepts of insertion and removability were addressed in the prior art. Specifically, defendant referred to United States Patent No. 3,790,-791, Feb. 5, 1974 (filed July 20, 1971) ("Anderson"). Throughout its specifications Anderson contains references related to rear removability. Figure 4 of Anderson depicts a clip that virtually is identical to the retention clip in the ’182 patent. The specifications describing figure 4 refer to the clip as a “rear release clip” and its function with respect to insertion and remova-bility in a manner that is consistent with Mr. Borsuk's testimony concerning insertion and re-movability in the ’182 patent. Confronted by this evidence, plaintiff took the position that insertion and removability as referred to in .the preambles for claims 1 and 2, respectively, were not features unique to the ’182 patent. However, plaintiff continued to maintain that the reference in claim 1 to "a termination pin assembly” rendered the preamble a claim limitation. Only the pin assembly feature therefore is discussed.

    . To the extent that analysis of claim 2 does not fully interpret the preamble and elements of claim 1, an analysis of claim 1 will follow. Mr. Borsuk's interpretation of claim 2 corresponds to illustration 1 appended to this opinion.

    . Claims 1 and 2 of the '182 patent refer to both the termination of an “optical fiber" and the termination of "fiber optic bundles." Much testimony was elicited that greater precision was required in mating two single fibers as opposed to bundles. There was also testimony that the earliest connectors were intended to join bundles of fiber, whereas later connectors were utilized to join single fibers. However, even assuming that the '182 patent was intended solely for the termination of bundle fibers, as defendant contends, there was only generalized testimony that the '182 patent or a similar fiber optic connector would require modification in order to accommodate single fiber termination. In view of the limited evidence depicting the physical and mechanical differences between connectors for terminating a single fiber versus connectors for terminating bundle fibers, the record provides no basis for distinguishing between these connectors.

    . During the prosecution of another ITT patent, United States Patent No. 3,990,779, filed July 24, 1975, Mr. Peterson distinguished Lebduska which was concerned with coupling bundle fibers as "not even concerned with the same problem” as the invention in question. He said, "(T]he problem of coupling single optical fibers ... is far more complex and difficult than the coupling of optical fibers____” Mr. Peterson's view suggests nothing about the manner in which an optical fiber connector would be modified for single fiber coupling. Therefore, it does not mean that the '182 patent could not accommodate single fibers if its claim limitations so allowed. As for the '145 patent, no distinction shall be raised concerning the coupling of single versus bundle fibers. The specifications of the '145 patent refer to the capacity of device, regardless of its configuration, to align both single and bundle fibers. See supra note 7.

    . The parties were advised before trial that any exhibit admitted but not discussed with a witness or referred to in argument would be excluded from consideration. Most of the references cited by the PTO fall into this category.

    . This part of the discussion is organized to correspond with the presentation of evidence at trial. With respect to the ’182 patent, defendant, in its case in chief, utilized the testimony of *347Mr. Kosmos to provide a detailed comparison between the claims at issue and Parfitt and Lebduska. Defendant proceeded generally with respect to the Clarkson reference. Plaintiff cross-examined Mr. Kosmos regarding Parfitt only as to claim 1, but as to both claims 1 and 2 with respect to Lebduska. Mr. Borsuk compared Parfitt, Lebduska, and Clarkson to both claims of the ’182 patent. Plaintiff’s failure to cross-examine Mr. Kosmos with respect to the comparison of claim 2 and Parfitt is of no consequence. Claim 2 of the ’182 patent differs from claim 1 in only one significant respect. Element 5 of claim 2 refers to the device’s ability to relieve “axial manufacturing tolerances," and plaintiff, in cross-examining Mr. Kos-mos with respect to claim 1 and Parfitt, treats this feature or function of claim 2. Also, Mr. Kosmos gave detailed rebuttal testimony on the topic during the claim 1 analysis.

    . The Parfitt article on its final page refers to the rubber part as a "small rubber washer,” although figure 4 on the penultimate page describes the piece as a "rubber pressure ring". Mr. Borsuk stated that a pressure washer is square as opposed to an O-ring, which is round.

    . Defendant contends that whether one defines the connector assembly as comprising all parts including the termination pin or just the outer ferrule is "a distinction without a difference."

    . In response to a defense question, Mr. Kos-mos refused to concede that a groove could be "an absence of a structure." With respect to Lebduska specifically, he stated: "[W]hen I took the resilient member out, part 33, the pin body could slide back to the flange 28 ... and the groove would disappear____" He also stated that a groove could not have a moving wall. Mr. Kosmos did agree that the groove in the ’182 patent could be defined as serving the function of holding a resilient member. Consistent with this definition, he was willing to state that the space on Lebduska between the back of the termination pin and the internal shoulders of the outer ferrule is a groove.

    . With the exception of the notion of a removable connection, the preamble, elements 1 and 2, and the first part of element 3 of claim 13 duplicate all of the requirements or features stated in the preamble and elements of claim 1. Therefore, with one exception, the conclusions reached with respect to claim 1 and the prior art references apply in full to the claim 13 analysis and will not be discussed in detail.

    The preamble and elements 1 and 2 of claim 13 refer to an alignment sleeve of resilient spring material for coupling cables containing fibers terminated by a termination pin. These requirements correspond to the preamble and element 2 of claim 1. The first part of element 3 of claim 13 requires "one axially extending slot through the wall of an axial section of said sleeve." Similarly, the second part of the second element of claim 1 of the 145 patent contains virtually the same requirements relating to “an axially extending slot.”

    *363Finally, the first part of element 4 of claim 13 requires "pins ... mounted in the ... ends of said sleeve and abutting each other in said axial section" of the sleeve. This requirement of abutment is referred to in both elements 1 and 2 of claim 1. The second part of the fourth element of claim 1 contains virtually the same terminology present in element 2 of claim 1 of the 145 patent, le., "resilient lateral expansion of said sleeve to produce inwardly directed radial forces" tending to hold said terminal ends or pins in lateral alignment.

    . Mr. Borsuk failed to find the requirements of the preamble satisfied simply because the device does not pertain to fiber optics. A device predating fiber optics likely would not satisfy the language of a preamble or elements in a patent using terminology descriptive of a fiber optic device. Thus, unless it disturbs the intent of the patent as it relates to fiber optics, the terminology unique to fiber optics shall be disregarded.

    . A part of element 1 and all of elements 2-5 of claim 9 virtually are equivalent to all of the elements in claim 1 of the 797 patent. For a prior art reference to satisfy claim 9, the reference therefore must satisfy all of the elements of claim 1 and the one difference between claims 1 and 9 as discussed in the text accompanying this note. Counsel agreed at trial that claims 1 and 9 were similar.

    . Although the '797 patent utilizes crimping or a similar method for securing a strength member, the entire attachment means slides out of a slot freely and is therefore removable.

    . Defendant elicited no testimony that the Suspended Nipple or Campbell contained the features set forth in claim 6. Nor is there a representation by defendant in its claim chart, which was part of its pretrial submission, that these prior art references disclose the features of claim 6. Therefore, with respect to claim 6 these references were not considered as part of defendant’s validity case.

    . Because a part of element 1 and all of elements 2-5 of claim 9 virtually are equivalent to all of the elements in claim 1 of the 797 patent, it was not necessary to develop abbreviations for the elements of the two claims that were similar. Thus, an abbreviation is used only for element 1 of claim 9.

    .' The abbreviations of the features of the elements that were identified with respect to the discussion of validity also apply in the infringement analysis.

    . To demonstrate further the similarity between the Moulin II and the device in PX-25, defendant established that the pin body in Mr. Moulin’s invention is designed to be inserted in a cavity.

    . PX-26A differs from PX-26 only insofar as ' in PX-26A the sleeve is shown with mating pin and socket contacts, whereas PX-26 shows only a pin contact. PX-26A will be referred to specifically only when it assists the explanation.

    . With respect to validity, Mr. Kosmos implied that this element calling for a retaining means does not pertain to the retention clip. In discussing every prior art reference, he testified that element 1 was satisfied solely by the alignment sleeve. During defendant’s case on validity, Mr. Borsuk never testified that element 1 could not be satisfied by an alignment sleeve, and in particular he did not testify that the prior art references lacked a retaining means because they did not contain a retention clip. Consequently, it is determined that the requirement of a retaining means is satisfied by a device which contains at least an alignment sleeve.

    . Both claims 1 and 13 require the functioning of resilient forces to effect alignment. However, element 3 of claim 13, with its requirement related to the diameter of the pin and sleeve, provides the integral means by which these forces are created. Mr. Borsuk’s testimony regarding resilient forces in his analysis of claim 1 illustrates that element 3 of claim 13 is also satisfied in the Hughes alignment sleeve. Resilient forces are generated in PX-26 because the sleeve allegedly has a diameter that is smaller than the outside diameter of the termination pin. Plaintiff did elicit testimony regarding all elements of claim 13 purporting to show the presence of all features in these elements. As stated with respect to validity, see supra note 14, the requirement of the first part of element 3 of claim 13 is the only difference between claims 1 and 13. This difference was discussed with respect to element 2 of claim 1 and therefore, along with the other elements of claim 13, need not be discussed separately.

    . Corroborating that forces are exerted on the' pin in only three spots, Mr. Moulin stated that the sleeve causes wear upon the guide bushing at the points of contact.

    . See supra text accompanying note 8.

    . With the exception of the Hughes device depicted in PX-32, each illustration offered by plaintiff of a Hughes device relevant to its infringement claims appears to show a connector for cable having a strength member as it is depicted in PX-27. Therefore, it is unnecessary to pursue discussion of this element with respect to PX-28 through 31 and PX-33.

    . Each illustration offered by plaintiff of a Hughes device relevant to its infringement claims depicts a support member with a forward and rear section depicted similarly in PX-27. Therefore, it is unnecessary to pursue discussion of this element with respect to PX-28 through 33.

    . In each illustration offered by plaintiff of Hughes devices relevant to the '797 infringement analysis, there appear a terminated contact and a means for releasably mounting the contact in the maimer testified to with respect to PX-27. Therefore, it is unnecessary to pursue discussion of this element with respect to PX-28 through 32. For these references, the element will be presumed present.

    . The illustrations depicted as PX-28 and 29 and offered by plaintiff in support of its infringement claims show means for fixed attachment of a strength member of cable in the manner testified to with respect to PX-27. Therefore, it is unnecessary to pursue discus*388sion of this element with respect to PX-28 and 29.

    . Bach illustration offered by plaintiff of a Hughes device relevant to its infringement claims depicts an attachment means in the rear section of the support member in a predetermined position relative to the forward section similar to the manner in which it is depicted in PX-27.

    . A part of element 1 and all of elements 2-4 of claim 9 virtually are equivalent to all of the elements in claim 1 of the ’797 patent. For a Hughes device to infringe claim 9, the reference therefore must satisfy all of the elements of claim 1 and the one difference between claims 1 and 9, as discussed in the text accompanying this note. Counsel agreed at trial that claims 1 and 9 were similar.

    . Were the claim interpreted without regard to the specifications to reach any form of remova-bility, defendant’s case on invalidity would be strengthened. See discussion of Moleculon Research Corp. v. CBS, Inc., 793 F.2d 1261, supra p. 375.

    . Mr. Borsuk explained that the PX-29 does not adequately depict the manner in which the attachment means is mounted into the Hughes device. Accordingly, Mr. Moulin, who has had considerable exposure to the connector which served as the model for PX-29, testified in detail concerning the mounting of the attachment means in the device. Element 4 of claim 1 and element 5 of claim 9 refer to the broad requirement relating to a means for removably mounting, without reference to the specific parts that are intended to perform this function. Claim 12 is much more specific and refers to the means for mounting and sets forth the parts and the manner in which they interrelate. Given the similarity among elements 4 and 5 of claims 1 and 9, respectively, and claim 12, Mr. Moulin’s detailed testimony is entitled to decisive weight with respect to element 4 of claim 1 and element 5 of claim 9, as well as claim 12.

    . Mr. Borsuk’s testimony supports the conclusion that there is virtually no difference between the Hughes devices in PX-31 and PX-32. From the illustration of PX-32 it would appear that the attachment means and means for removable mounting in the two devices is different. However, Mr. Borsuk stated that PX-31 merely demonstrates what PX-32 looks like internally. He also stated that he had conducted research regarding the Hughes drawings which revealed that the parts in PX-31 and 32 representing the attachment means and means for removable mounting share the same part numbers. As for plaintiff’s infringement claim related to claim 6, Mr. Borsuk’s testimony was based upon the Hughes drawing from which PX-32 was generated and recognized the presence of two rods or longitudinal legs which join the forward and rear sections of the yoke. Because two legs have been deemed present, the Hughes device necessarily contains at least one leg consistent with the relevant part of element 1 of claim 9. The similarity between PX-31 and 32 in relation to element 1 compels the finding that PX-32 infringes claim 1 of the 797 patent by the doctrine of equivalents. Claim 6 is dependent on claim 1, and claim 9 substantially duplicates claim 1; therefore, the device in PX-32 infringes claims 1, 6, and 9 by the doctrine of equivalents.

    . Because elements 1, 2, and 4 of claim 1 are duplicated in claim 9, PX-33 does not infringe claim 9 for the same reasons stated in the analysis related to claim 1.