Eltra Corporation v. Basic Incorporated

WEICK, Circuit Judge. The suit in the District Court was brought by Eltra Corporation (Eltra) against Basic Incorporated (Basic) seeking a declaratory judgment that Basic’s Reissue Patent No. Re. 27,111, entitled “Pitch-Bonded Refractory Composition” was invalid and noninfringed. Basie responded by counterclaiming against Eltra for infringement of its patent by North American Refractories Company (Narco), a division of Eltra. The subject matter of the patent in suit is a process for making pitch-bonded refractory bricks used to line furnaces in which steel is made by the basic oxygen process. Following a bench trial the District Court held that the patent in suit was valid and infringed. The Court further held that Basie was entitled to treble damages under 35 U.S.C. § 284 because the infringement was “deliberate and intentional.” The Court also held that Basic was entitled to attorney’s fees under 35 U.S.C. § 285 because Eltra’s claim of fraud on the patent office was a “smoke screen.” A later hearing on the amount of damages and fees to be awarded was to be held if the parties could not agree on the amounts. We disagree and reverse. In our opinion the patent in suit is invalid for obviousness under 35 U.S.C. § 103. Accordingly, Basic is not entitled to either damages or attorney’s fees, and it is not necessary for us to reach either the claim of fraud on the patent office or the issue of infringement. Additionally, even if the patent were valid, we can perceive no basis in the record of this case for awarding Basic either treble damages or attorney’s fees. I The parties in this case are competitors in the manufacture and sale of pitch-bonded refractory bricks. These bricks are used in the steel industry to line the inside of basic oxygen furnaces. The basic oxygen process for steelmaking became generally used and accepted in this country during the 1950’s. While refractory materials had always been used in steelmaking, producers found in the 1950’s and 1960’s that they needed an improved brick, one which would last longer under the adverse conditions present in the basic oxygen furnace. The essential process for making pitch-bonded refractory bricks is well known. A refractory material such as magnesia or dolomite is combined with a hydrocarbon binder (pitch). The mixture is compressed, formed into a brick, and baked. Basic’s claimed invention involves the addition of small amounts of carbon of specified varieties to the mixture. 1 According to the patent’s specifications, bricks made by this process will exhibit improved qualities as to oxidation, crushing strength and density. The invention of this formula was allegedly made in 1960 by a then employee of Basic, Roger E. Wilson. U. S. Letters Patent No. 3,236,664 were originally issued to him on his application filed in 1962. Thereafter, officers at Basic realized that the original patent might not be valid in light of an earlier, similar Canadian patent No. 614,742 issued to one Lisle Hodnett. 2 This patent had not been before the patent examiner. An application for a reissue patent 3 was filed in 1969. This application narrowed the original claims, deleted some, and apparently attempted to distinguish *748 them from Hodnett, although the Hodnett patent was again not cited to the patent office. In spite of this omission, the examiner located the prior art patent on his own and cited it in rejecting the reissue application. After amendment, the reissue patent was eventually granted in 1970 with the specific claims stated in the margin. 4 *750 The process described in the original and reissue patents has enjoyed some commercial success. In 1968 the Narco division of Eltra began producing bricks under a licensing agreement with Basic. The largest company in the business, Harbison-Walker, also acquired a license to use the process, as did a smaller firm. More recently, several other refractory producers have followed suit. In 1973, Narco became aware that Harbi-son-Walker had filed suit challenging the validity of the Wilson patent. Narco then consulted outside counsel and obtained an opinion that the Wilson patent was invalid. Narco thereafter ceased paying royalties to Basic, explaining that the patent was viewed as invalid. Basic and Narco attempted to renegotiate a mutually satisfactory licensing agreement. But when no agreement could be reached, the present suit was filed. In the interim, Basic and Harbison-Walker settled their lawsuit. Their agreement provided Harbison-Walker with a royalty-free license under Basic’s Wilson patent. In exchange for this Basic received a royalty-free license under a Harbison-Walker patent that was later held invalid in Dresser Indus., Inc. v. Eltra Corp., 432 F.Supp. 153 (N.D.Ohio 1977) (Dresser Indus, is the successor to Harbison-Walker). II Eltra’s principal contention is that the Wilson patent, as reissued, is invalid for obviousness under 35 U.S.C. § 103. The section provides: § 103. Conditions for patentability; non-obvious subject matter A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 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 negatived by the manner in which the invention was made. In analyzing the validity of any patent we must begin with the statutory presumption of validity that accompanies its issuance, 35 U.S.C. § 282. American Seating Co. v. National Seating Co., 586 F.2d 611, 615 (6th Cir. 1978). The presumption has no independent evidentiary significance, however, as it merely serves to allocate to the party claiming invalidity the burden of proving it. 5 Reynolds Metals Co. v. Acorn Bldg. Components, Inc., 548 F.2d 155, 160 (6th Cir. 1977); Dickstein v. Seventy Corp., 522 F.2d 1294,1296 (6th Cir. 1975), cert. denied, 423 U.S. 1055, 96 S.Ct. 787, 46 L.Ed.2d 644 (1976); Sperberg v. Goodyear Tire & Rubber Co., 519 F.2d 708, 713 (6th Cir.), cert. denied, 423 U.S. 987, 96 S.Ct. 395, 46 L.Ed.2d 303 (1975); Rains v. Niaqua, Inc., 406 F.2d 275, 278 (2d Cir.), cert. denied, 395 U.S. 909, 89 S.Ct. 1751, 23 L.Ed.2d 222 (1969). In the typical case such as this, where the bulk of the evidence of the prior art is contained in documents, the party claiming obviousness need only do so by a preponderance of the evidence. Dickstein, supra, 522 F.2d at 1295-97; cf. Campbell v. Spectrum Automation Co., 513 F.2d 932 (5th Cir. 1975). The District Court was thus in error in stating that: The basic rule is that a patent is presumed to be valid, and those who attack its validity, in order to succeed, must establish their case by clear and convincing evidence. [App. 471.] While this higher standard of proof may apply to the unusual case, such as where the *751 evidence may be of an inherently unreliable nature, see Dickstein, supra, 522 F.2d at 1296; Campbell, supra, or where fraud is alleged, see Schnadig Corp. v. Gains Mfg. Co., Inc., 494 F.2d 383, 392 (6th Cir. 1974), it is simply not the “basic rule” to be applied in cases involving alleged obviousness under 35 U.S.C. § 103. In the seminal case of Graham v. John Deere Co., 383 U.S. 1, 86 S.Ct. 684, 15 L.Ed.2d 545 (1966), the Supreme Court outlined the proper inquiry under section 103 as follows: [T]he scope and content of the prior art are to be determined; differences between the prior art and the claims at issue are to be ascertained; and the level of ordinary skill in the pertinent art resolved. Against this background, the obviousness or nonobviousness of the subject matter is determined. [Id. at 17, 86 S.Ct. at 694.] The Court continued: Such 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. [Id. at 17— 18, 86 S.Ct. at 694.] Essentially, nonobviousness is the statutory equivalent of the requirement of “invention,” derived from the Supreme Court’s decision in Hotchkiss v. Greenwood, 52 U.S. (11 How.) 248, 13 L.Ed. 683 (1851). Sakraida v. Ag Pro, Inc., 425 U.S. 273, 279, 96 S.Ct. 1532, 47 L.Ed.2d 784 (1976); Dann v. Johnston, 425 U.S. 219, 225-26, 96 S.Ct. 1393, 47 L.Ed.2d 692 (1976); Reynolds Metals, supra, 548 F.2d at 159; Monroe Auto Equip. Co. v. Heckethorn Mfg. & Supply Co., 332 F.2d 406, 410 (6th Cir.), cert. denied, 379 U.S. 888, 85 S.Ct. 160, 13 L.Ed.2d 93 (1964). This requirement explicates the constitutional mandate that inventions must “promote the useful arts,” Art. I, § 8, cl. 8, before a limited monopoly may be granted. Anderson’s Black-Rock v. Pavement Co., 396 U.S. 57, 61, 90 S.Ct. 305, 24 L.Ed.2d 258 (1969); see Graham, supra. Thus not every improvement in the art is patentable. To be nonobvious, some prior decisions held that the advancement must be the product of “inventive genius.” A. & P. Tea Co. v. Supermarket Corp., 340 U.S. 147, 154, 71 S.Ct. 127, 95 L.Ed. 162 (1950) (Douglas, J., concurring); Mantle Lamp Co. v. Aluminum Co., 301 U.S. 544, 546, 57 S.Ct. 837, 81 L.Ed. 1277 (1937); Concrete Appliances v. Gomery, 269 U.S. 177, 185, 46 S.Ct. 42, 70 L.Ed. 222 (1925); Reckendorfer v. Faber, 92 U.S. 347, 354, 23 L.Ed. 719 (1875). The standard of invention is a “demanding” one. Lear, Inc. v. Adkins, 395 U.S. 653, 676, 89 S.Ct. 1902, 23 L.Ed.2d 610 (1969). As stated in Dann, supra, the question is whether the difference between the prior art and the subject matter in question “is a difference sufficient to render the claimed subject matter unobvious to one skilled in the applicable art. . . . ” [ 425 U.S. at 228, 96 S.Ct. at 1398, quoting Application of Johnston, 502 F.2d 765, 772 (C.C.P.A.1974) (Markey, C. J., dissenting).] Although the subsidiary questions outlined in Graham, supra, are essentially factual, the ultimate issue is one of law. Sakraida, supra, 425 U.S. at 280, 96 S.Ct. 1532; Graham, supra, 383 U.S. at 17; American Seating, supra, 586 F.2d at 619; Nickola v. Peterson, 580 F.2d 898, 910-12 (6th Cir. 1978), cert. denied, - U.S. -, 99 S.Ct. 1504, 59 L.Ed.2d 774 (1979); Reynolds Metals, supra, 548 F.2d at 161; Kolene Corp. v. Motor City Metal Treating, Inc., 440 F.2d 77, 81 (6th Cir.), cert. denied, 404 U.S. 886, 92 S.Ct. 203, 30 L.Ed.2d 169 (1971). In our opinion the advancement disclosed in the Wilson reissue patent does not meet this standard in view of the prior art. A As set out in footnote 4, supra, the claims of the Wilson reissue patent relate to the process of adding small amounts (from .5% *752 to 10% by weight) of carbon black, 6 where at least one third of the carbon black is a variety known as thermal black, 7 to the refractory mixture. The preferred amount of carbon additive is 2% to 3% by weight. One specific variety of the invention calls for the use of a mixture of thermal black with any of several “high oil absorbing blacks” (a subclass of carbon blacks). Another calls for the use of all thermal black. The District Court concluded that the process described in the Wilson reissue patent was “something entirely new and different from prior art.” App. 475'. As a conclusion that the process was nonobvious it is patently incorrect. At least by 1944 it was known that the chemical resistance of refractory materials to wear and corrosion by slag during the steelmaking process could be improved by increasing the carbon content of the refractory mixture. This was taught in Australian patent No. 118,590 where it explained the utility of adding 2Vz% by weight of free carbon 8 in the form of flake graphite 9 to the refractory mixture. The Hodnett Canadian patent, No. 614,742, which was issued more than a year prior to Wilson’s original application, 10 also taught the value of carbon generally in making improved refractory materials. But the patent was also rather specific in teaching a process for making a better refractory brick for use in the steel industry. Hodnett called for the addition of from 0% to 7% by weight of “finely divided free carbon.” The patent specified both graphite and carbon black, although it expressed a preference for graphite. 11 Carbon black was said to be less desirable because of its tendency to lower density. 12 Basic places great emphasis on this last teaching because the core of its argument is that the carbon mixture prescribed by Wilson gave rise to such a surprising and unex *753 pected improvement in the strength and density of refractory bricks over those using graphite as to be patentable over the prior art, particularly Hodnett. 13 We believe that the proof fell far short of establishing the existence of such a difference. Much evidence was presented at trial concerning the way in which carbon acts to affect the strength and density of pitch-bonded refractory bricks. The refractory material (here magnesia or dolomite) exists in the form of particles of irregular shapes and sizes. Even when combined with a pitch binder, formed into a brick, and baked, pores remain between the particles. When those pores are filled with smaller particles, here carbon, the resulting brick exhibits improved crushing strength and density over bricks made simply with refractory and pitch. Using filler particles that are too large, however, can result in a weaker brick since the refractory particles will be forced apart. Hence it is better to err on the side of having filler particles which are too small rather than too large. This much is basic in the art. It was also taught in the Heuer patent, U.S. No. 1,851,-181 (1932). It should be noted that the improved chemical resistance of the bricks is achieved by the choice of carbon particles in order to increase the carbon content of the brick, as is explained above. Finally there was also some evidence that free carbon has some independent effect on the pitch binder which may cause additional improvement in the strength of the bricks, but not in their density. 14 In pressing its claim on reissue before the patent office, and in an effort to distinguish Hodnett after it had been cited by the patent examiner, Basic made the following representation: [Substantial evidence has been submitted, in the Collin Hyde affidavit, as to the unexpected and surprising results achieved by the present invention (the affidavit compares the presently claimed species with the preferred species of the Hodnett patent). The statement contained in the parentheses was false. The tests reported by Collin Hyde in his affidavit did not involve “the preferred species” of Hodnett. Hyde compared bricks made with a thermal black mixture to those made with a coarse variety of graphite (Joseph Dixon flake graphite 1101). Hodnett, however, had specified fine (or finely divided) graphite. Hyde reported that the use of coarse graphite actually decreased the density and compressive strength of the bricks over those containing no carbon additive at all. In contrast, Hyde found that the use of thermal black produced an increase in the density and strength of bricks over those containing no carbon *754 additive. By representing the coarse graphite as Hodnett’s “preferred species” the Hyde affidavit appeared to show two things: (1) it tended to disprove Hodnett by indicating that his process resulted in an inferior brick; (2) it tended to show that Wilson’s process was far superior to Hod-nett’s. In truth the experiments reported in the affidavit offered no such basis for comparison because they did not replicate the Hodnett invention. 15 It must be noted that Basic had in fact tested fine graphite, Hodnett’s “preferred species,” prior to this time. In 1960 Wilson reported tests in which he compared fine graphite to Joseph Dixon flake graphite 1101, the coarse graphite used in the Hyde experiments which were reported to the patent examiner. The tests showed that the fine graphite improved the density and compressive strength of refractory bricks whereas the coarse variety resulted in a poorer quality brick, even poorer than one containing no additive. Plaintiff’s Ex. 42. This report was not given to the patent examiner, even though it predated Hyde’s affidavit. More significantly, very little evidence was presented which directly compared fine graphite with carbon black generally, or with thermal black in particular. The Hyde affidavit purportedly did so and was presented both to the patent examiner and to the District Court. But as noted above, it did not offer a fair comparison of Wilson and Hodnett. The Wilson patent, either originally or as reissued, also does not furnish any report of a comparison test. The data provided only compares various carbon blacks. The specifications instruct that “pulverant carbon of non-cubic crystalline structure may also be used in practicing the invention. For example graphite may be used, but such carbons are not as efficacious as carbon blacks.” The patent offers neither explanation or data to support this claimed distinction. 16 Apparently, Basic’s only reported direct comparison between fine graphite and thermal black was contained in a report by Collin Hyde which was dated four months earlier than the affidavit submitted to the patent examiner. The report showed that the addition of either thermal black or fine graphite increased the density and compressive strength of pitch-bonded refractory bricks. Neither additive was found to be the clear superior of the other. 17 Again this report was not given to the patent examiner. 18 *755 To show that thermal black is the clear and surprising superior of fine graphite, Basic relies on the reports of one of Narco’s scientists. In 1973 he reported on experiments which were conducted in an effort to get “outside” the Wilson patent. He compared the pressed (or “green”) densities 19 of bricks made with thermal black with those made with a thermal black-graphite mixture and with a thermal black-furnace black mixture. He did not evaluate the coked density 20 of the bricks. Neither did the tests report on compressive strength, either green or coked. The most that this report shows is that the thermal-black-graphite mixture produced a reduction in pressed density. Notably, a mixture containing more graphite and less thermal black achieved an even smaller reduction in pressed density from bricks with pure thermal black. 21 Also, the report accompanying the data indicated that the reductions in density were at least in part due to a reduction in pitch content. In sum, the technical evidence disclosed the following relative to the prior art and the alleged advancement made by Wilson: As to the prior art, it was well known that the chemical resistance of bricks to wear and corrosion could be improved by increasing the carbon content of the refractory mixture; it was also known that the strength and density of refractory materials could be improved through the addition of smaller particles into the mixture; Hodnett specifically taught the use of carbon particles in pitch-bonded refractory bricks; while graphite was preferred by Hodnett, it was clear that both graphite and carbon black were useful for his invention. 22 *756 As to the alleged advancement made by Wilson, Basic claims that a patentable improvement over the prior art was achieved through the use of various mixtures of carbon blacks, particularly thermal black. The Wilson patent’s specifications, however, also note that graphite may be used to practice the invention. Wilson simply expresses a preference for carbon black without documenting its superiority. Eltra came forward with evidence which tended to show that the claimed “surprising and unexpected” improvements over Hod-nett’s teachings were incorrect and that fine graphite was indeed the equivalent of thermal black. Basic responded with evidence in the form of incomplete test results which tended to show that as to only one property thermal black exhibited superiority over fine graphite. Thus in our view the technical evidence showed that Basic had developed a process which produced a brick that in most respects was the equivalent of a brick produced by the Hodnett process. All Basic had accomplished was the substitution of thermal black, together with other carbon blacks, for graphite. It did not even change the relative proportions. 23 B In addition to technical evidence, the parties offered evidence concerning certain “secondary considerations.” Graham, supra, 383 U.S. at 17, 86 S.Ct. 684. Basic pointed particularly to the commercial success of the thermal black process, noting that almost every producer of refractory bricks uses thermal black under license from Basic. These competitors were apparently unable to find a superior process. Of course, commercial success and the satisfaction of long-felt needs are alone not sufficient to establish that the product is the result of invention. Sakraida, supra, 425 U.S. at 278-79, 96 S.Ct. 1532; Anderson’s Black-Rock, supra, 396 U.S. at 61, 90 S.Ct. 305; Philips Indus. Inc. v. State Stove & Mfg. Co., 522 F.2d 1137, 1141-42 (6th Cir. 1975). Here, moreover, there was evidence that the consideration received from Harbi-son-Walker, the largest firm in the business, for its license was of questionable value. There was also evidence that thermal black is the cheapest of the free carbons. Indeed, a 1968 letter to patent counsel indicated that Basic wanted a patent specifying thermal black because of its comparatively low cost. Defendant’s Ex. FR at 4. The same letter also indicated that in 1965 when Basic reverted to using thermal black from graphite it did so purely for economic reasons. In our view these secondary facts can in no way “tip the scales” in Basic’s favor. American Seating, supra, 586 F.2d at 622. They tend, instead, to tip the scales in Eltra’s favor. In view of the fact that the prior art had already disclosed the usefulness of carbon generally, and carbon blacks in particular, and since the Wilson patent changed neither the quantities of carbon nor the basic process of particle stuffing from the Hodnett patent, and also given the fact that Wilson’s process was not shown to perform demonstrably better than Hod-nett’s, we believe that the selection of thermal black as a preferred additive is not a patentable difference. Cf. Lucerne Products, Inc. v. Cutler-Hammer, Inc., 568 F.2d 784, 798 (6th Cir. 1977). We are confirmed *757 in this view by the other evidence which tended to show that economics, not invention, may have been the real source of this patent. But a cost savings is not a patentable difference. It is not a substitute for invention. Sakraida, supra, 425 U.S. at 282-83, 96 S.Ct. 1532; Reynolds Metals, supra, 548 F.2d at 162. In addition, there can be no “exercise of the inventive faculty,” McClain v. Ortmayer, 141 U.S. 419, 427, 12 S.Ct. 76, 35 L.Ed. 800 (1891), where one merely substitutes one material for another. Graham, supra, 383 U.S. at 11, 86 S.Ct. 684; Hotchkiss v. Greenwood, 52 U.S. (11 How.) 248, 13 L.Ed. 683 (1851). In our opinion this case involves at most “improvement . [which is] the work of the skillful mechanic not that of the inventor.” Id. at 267. It was therefore obvious under 35 U.S.C. § 103. 24 III Although our holding that the patent is invalid necessarily eliminates the District Court’s award of treble damages and attorney’s fees, we believe those awards deserve specific comment because of the serious error committed below. In our opinion, even if the patent were valid there is no basis in the record of this case for either award. The statutory provision relating to damages is 35 U.S.C. § 284. It provides: § 284. Damages Upon finding for the claimant the court shall award the claimant damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer, together with interest and costs as fixed by the court. When the damages are not found by a jury, the court shall assess them. In either event the court may increase the damages up to three times the amount found or assessed. The court may receive expert testimony as an aid to the determination of damages or of what royalty would be reasonable under the circumstances. To support its extraordinary award the District Court stated: The evidence in this case leaves no doubt in the Court’s mind that the plaintiff’s infringement in this case was deliberate and intentional. To condone such misconduct would set an example that would seriously weaken the patent system. The Court will, therefore, when the amount of damages has been determined, either by its assessment or upon stipulation of the parties, increase it three times. [App. 477.] In order to support such an increased, punitive award there must be a finding that the infringement was wilful. H. K. Porter Co., Inc. v. Goodyear Tire & Rubber Co., 536 F.2d 1115, 1124 (6th Cir. 1976). The existence of honest doubt concerning the validity of a patent precludes a finding of wilfulness. Id.; General Electric Co. v. Sciaky Bros., Inc., 415 F.2d 1068, 1073 (6th Cir. 1969). As this Court stated in Enterprise Mfg. Co. v. Shakespeare Co., 141 F.2d 916, 921 (6th Cir. 1944): If honestly mistaken as to a reasonably debatable question of validity, an infringer should not be made to smart in punitive damages. Compensatory damages constitute adequate remuneration for invasion of a patentee’s property rights, unless the refusal of the infringer to bow to the presumptive validity of an issued patent is consciously wrongful. A court of equity, exercising patent jurisdiction, does not readily infer wrong motivation upon the part of those resisting the validity of patent claims. Patentees generally entertain suspicion that those who challenge their claims are deliberate malefactors. However bona fide, such suspicions produce no legal effect, unless sustained by evidence substantiating suspicion as truth. In this case Eltra’s Narco division stopped paying royalties after it learned that Harbison-Walker, the dominant firm in the business, had questioned *758 the validity of the patent. The opinion of outside counsel was also sought and it was to the effect that the patent was invalid. In this context Eltra’s Narco division ceased its royalty payments and attempted to renegotiate the license. Only after these negotiations broke down, and after it learned that Harbison-Walker had obtained a royalty-free license did Eltra bring suit. Its “infringement” was thus “intentional” in the sense that it was not inadvertent. But a licensee is not required to pay royalties when it successfully challenges a patent’s validity. Lear, Inc. v. Adkins, 395 U.S. 653, 671-74, 89 S.Ct. 1902, 23 L.Ed..2d 610 (1969). These facts simply do not make out a wilful infringement so that there was no basis whatever on which, to award damages in excess of those intended to compensate Basic for its loss. We can similarly find no plausible basis for the District Court’s award of attorney’s fees. The pertinent statute is 35 U.S.C. § 285. It provides simply: § 285. Attorney fees The court in exceptional cases may award reasonable attorney fees to the prevailing party. The District Court found this case to be exceptional stating: The defendant also asks for an allowance of attorneys fees. 35 U.S.C. § 285 provides that in exceptional cases the Court may award reasonable attorney fees. This case is made exceptional by the insistence of the plaintiff that the Wilson Patent and its re-issue were obtained by fraud and deceit practiced by the defendant upon the Patent Office. To make such a charge not only impugns the party charged, but also reflects on the Patent Office and its staff. It implies either lack of competence or lack of effort, or both, since it is ordinarily very difficult to practice fraud on one who is reasonably skilled in his business and is paying proper attention to it. It is elementary in law that one who seeks to establish fraud must do so by clear and convincing evidence. A person should not make damaging accusations without a very strong basis for believing them true. To make such an accusation as a smoke screen to divert attention from, or to attempt to confuse, the basic issues of patent validity or infringement ought not to be permitted. [App. 477.] In order to support an award of attorney’s fees in a patent case we have previously held that there must be a showing of conduct which is unfair, in bad faith, inequitable, or unconscionable. Deyerle v. Wright Mfg. Co., 496 F.2d 45, 54-55 (6th Cir. 1974); Uniflow Mfg. Co. v. King-Seeley Thermos Co., 428 F.2d 335, 341 (6th Cir.), cert. denied, 400 U.S. 943, 91 S.Ct. 245, 27 L.Ed.2d 248 (1970); Hoge Warren Zimmermann Co. v. Nourse & Co., 293 F.2d 779, 784 (1961). The District Court’s award was based on the fact that, in the Court’s view, Basic and the patent office had been “impugned” by an unproved allegation of fraud on the patent office. In our opinion only the most frivolous of allegations should give rise to an award of attorney’s fees under section 285. Normally awards under this provision are based on the conduct of the parties, not on the quality of their proof. E. g., Deyerle, supra. In this case, while we offer no opinion as to the merits of Eltra’s claim of fraud on the patent office, we cannot view the allegation as frivolous. As explained above, Basic misrepresented the thrust of the Hyde affidavit in its prosecution of the reissue application. In addition, there was some evidence that Basic “rigged” the Hyde results by intentionally using coarse graphite in its comparative tests, even though Basic’s officials were aware that fine graphite would yield different results. In the face of these facts the District Court erred in awarding attorney’s fees to Basic. The judgment of the District Court is reversed. Footnotes [1] The original patent included broad claims to the use of many forms of carbon including carbon black and graphite. (See note 6, infra, for an explanation of these terms.) The patent’s specifications, however, expressed a preference for carbon black. [2] The evidence showed that Basic had acquired a license under the Hodnett patent in 1961 for the nominal price of $1 per year. Basic’s officials did not rediscover the existence of the Hodnett patent until sometime later, possibly as late as 1968. [3] Under 35 U.S.C. § 251 a patent may be reissued under the following circumstances: § 251. Reissue of defective patents Whenever any patent is, through error without any deceptive intention, deemed wholly or partly inoperative or invalid, by reason of a defective specification or drawing, or by reason of the patentee claiming more or less than he had a right to claim in the patent, the Commissioner shall, on the surrender of such patent and the payment of the fee required by law, reissue the patent for *748 the invention disclosed in the original patent, and in accordance with a new and amended application, for the unexpired part of the term of the original patent. No new matter shall be introduced into the application for reissue. In Bolkcom v. Carborundum Co., 523 F.2d 492 (6th Cir. 1975), cert. denied, 425 U.S. 951, 96 S.Ct. 1725, 48 L.Ed.2d 194 (1976), we stated on page 502: The Supreme Court has made it clear that this section is intended only to permit the reissue of a new patent for the same invention as that disclosed in the original patent in order to permit the correction of an innocent inadvertent defect or omission. It is not intended to permit the patentee to broaden the claims of the original patent. (Emphasis in quotation). [4] The claims in the reissue patent provide as follows (material in roman was retained from the original patent; material in brackets was omitted upon reissue; material in italics was added upon reissue): 1. [omitted entirely] 2. In the method of forming a shaped, [green] grain refractory article by admixing dead-burned basic refractory particles with sufficient pitch capable of pyrolytic decomposition to bind said particles together and then shaping the admixture by pressure; the improvement which consists of adding approximately 0.5 to 10 percent- by weight, based on the weight of the total admixture, of finely divided carbon black to the admixture prior to such shaping[.] to produce such a shape of increased density and crushing strength, such carbon black having an average particle size of about 20 to about 500 millimicrons and a surface area of from about 5 to about 375 square meters per gram, at least one third of the carbon black being thermal black having an average particle diameter of about 180^170 millimicrons and a surface area of about 6-13 square meters per gram. 3. [omitted entirely] 4. In the method of bonding dead-burned basic refractory particles one to another by admixing such particles with about 4 percent to about 10 percent by weight of the admixture coal tar pitch and then heating to coke the admixture and form a bonded mass; the improvement which consists of incorporating approximately 0.5 to 10 percent by weight, based on the weight of the total admixture, of powdered carbon black in the admixture pri- or to heating to improve the useful life of the bonded mass at elevated temperatures[.], such carbon blacks having an average particle size of about 20 to about 500 millimicrons and a surface area of from about 5 to about 375 square meters per gram, at least one third of the carbon black being thermal black having an average particle diameter of about 180-470 millimicrons and a surface area of about 6-13 square meters per gram. 5. In the method of bonding refractory particles selected from the group consisting of dead-burned dolomite, dead-burned magnesia, and mixtures thereof by blending such particles with sufficient coal tar pitch to bind such particles together, shaping such blend, and then heating the resulting shape to a temperature sufficient to decompose pyro-lytically the pitch and form a carbon bond; the improvement which consists of adding to the blend prior to the heating from about 0.5 percent to about 10 percent by weight thereof finely divided carbon black[.] to increase the density and crushing strength of such shape, such carbon black having an average particle size of from about 20 to about 500 millimicrons and a surface area of about 5 to about 375 square meters per gram, at least one third of the carbon black being thermal black having an average particle diameter of about 180-470 millimicrons and a surface area of about 6-13 square meters per gram. 6. The method of claim 5 wherein such carbon black is selected from the group consisting of lamp blacks, channel blacks, furnace combustion blacks, thermal blacks and acetylene blacks. 7. The method of claim 5 wherein such carbon black has [properties within the following ranges: Average particle diameter ... 200 to 500 millimicrons. Surface area...........5 to 375 square meters per gram. Volatile content......Less than 14% by weight. Fixed carbon...........85 to 99.5% by weight.] a volatile content of less than 14 percent by weight and a fixed carbon content of 88 to 99.5 percent by weight. 8. In the method of bonding refractory particles selected from the group consisting of dead-burned dolomite, dead-burned magnesia and mixtures thereof by blending such particles with sufficient coal tar pitch to bind such particles together, shaping such blend, and then heating the resulting shape to a temperature sufficient to decompose pyro-lytically the pitch and form a carbon bond; the improvement which consists of adding to the blend prior to the heating from about 0.5 percent to about 10 percent by weight thereof finely divided carbon black [The method of *749 claim 5 wherein], such carbon black [consists] consisting essentially of a blend of high oil absorbing carbon black and a thermal carbon black. 9. In the method of bonding refractory particles selected from the group consisting of dead-burned dolomite, dead-burned magnesia and mixtures thereof by blending such particles with sufficient coal tar pitch to bind such particles together, shaping such blend, and then heating the resulting shape to a temperature sufficient to decompose pyro-lytically the pitch and form a carbon bond; the improvement which consists of adding to the blend prior to the heating from about 0.5 percent to about 10 percent by weight thereof finely divided carbon black [the method of claim 5 wherein], such carbon black [consists] consisting essentially of a blend of a high oil absorbing carbon black having an oil absorption of at least 85 pounds of oil per 100 pounds of black and a thermal carbon black, said carbon black being present within a weight ratio of 2:1 to 1:2, respectively. 10. In the method of bonding refractory particles selected from the group consisting of dead-burned dolomite, dead-burned magnesia and mixtures thereof by blending such particles with sufficient coal tar pitch to bind such particles together, shaping such blend, and then heating the resulting shape to a temperature sufficient to decompose pyro-lytically the pitch and form a carbon bond; the improvement which consists of adding to the blend prior to the heating from about 0.5 percent to about 10 percent by weight thereof finely divided carbon black [The method of claim 5 wherein], such carbon black [consists] consisting essentially of a blend of substantially equal parts by weight of a high oil absorbing carbon black selected from the group consisting of a conductive oil furnace carbon black and a long flow channel carbon black having an oil absorption of at least 85 pounds of oil per 100 pounds of black, and a fine thermal carbon black. 11. In the method of bonding refractory particles selected from the group consisting of dead-burned dolomite, dead-burned magnesia, and mixtures thereof by blending such particles with sufficient coal tar pitch to bind said particles together, shaping such blend under pressure, and then heating the resulting shape to a temperature sufficient to decompose pyrolytically the pitch and form a carbon bond, the improvement which consists of adding to the blend prior to shaping approximately 0.5 to 10 percent by weight, based on the weight of the total admixture, of powdered carbon black containing particles having a diameter within the range of from about 20 millimicrons to about 500 millimi-crons].] to increase the density and crushing strength of such carbon bonded shape, at least one third of such carbon black being thermal carbon black having an average particle diameter within the range of about 180 to about 470 millimicrons. 12. [omitted entirely] 13. [omitted entirely] 14. A refractory article of manufacture consisting essentially of basic refractory particles, carbon black and a pyrolytically decomposed carbonaceous material selected from the group consisting of pitch, coal tar and bituminous asphalts, approximately 0.5 to 10 percent Ijy weight, based on the weight of the total admixture, of said carbon black being present prior to such pyrolytic decomposition].], said refractory article having increased density and crushing strength and said carbon black having an average particle size of from about 20 to 500 millimicrons and a surface area of about 5 to about 375 square meters per gram, at least one third of such carbon black being thermal black having an average particle diameter of about 180-470 millimicrons and a surface area of about 6-13 square meters per gram. 15. The method of claim 11 in which such powdered carbon black consists essentially of a blend of a high oil absorbing carbon black and a thermal carbon black, said carbon black being present within the weight ratio of 2:1 to 1:2 respectively. 16. A pitch-bonded refractory having high strength and increased density comprising basic refractory particles, sufficient carbonaceous material selected from the group consisting of pitch, coal tar and bituminous asphalts, capable of pyro-lytic decomposition, to bind said particles together, and approximately 0.5 to 10% by weight, based on the weight of the total admixture, of finely divided carbon black of noncrystalline structure, said carbon black having an average particle size of from about 20 to 500 millimicrons and a surface area from about 5 to about 375 square meters per gram, at least on (sic one) third of the carbon black being thermal carbon black having an average particle diameter of from about 180 to about 470 millimicrons and a surface area of about 6-13 square meters per gram. 17. The refractory of claim 16 in which said finely divided carbon black consists essentially of a blend of a high oil absorbing carbon black and said thermal carbon black. 18. The refractory of claim 17 wherein the ratio of high oil absorbing carbon black to thermal carbon black is in the range of about 2:1 to about 1:2 respectively. 19. The refractory of claim 17 wherein said oil absorbing carbon black has an oil absorption of at least about 85 pounds of oil per 100 pounds of carbon black. 20. The refractory of claim 19 including about 4 to about 10% of carbonaceous material. *750 21. The refractory of claim 16 wherein the carbon black addition is substantially all thermal black. 22. The method of claim 5 wherein such carbon black comprises about 66 percent by weight to about 33 percent by weight thermal black. [Patent No. Re. 27,111, Defendant’s Ex. B.] [5] In addition we believe that the presumption was seriously weakened in this case because of Basic’s conduct before the patent office, where one document was misrepresented and another was withheld. See note 18, infra. [6] The term carbon black describes a family of so-called “free carbons,” where carbon exists in its elemental form not combined with any other chemical. Graphite is also a free carbon. The two forms differ, however, because carbon blacks are noncrystalline whereas graphite is a crystalline form of the element. [7] Different varieties of carbon black are made by heat treating various hydrocarbons (e. g., natural gas). Names are assigned (e. g., thermal black) according to the process which is used. They are also divided according to their physical properties, particularly particle diameter and surface area. [8] See note 6, supra. [9] See id. [10] As explained in footnote 3 above, a reissue patent under 35 U.S.C. § 281 is fundamentally only a correction of the original patent. Bolk-com, supra, 523 F.2d at 501-02. Thus in our opinion one must view the obviousness of the reissue claims to the same invention as of the time of invention, namely, the time of the original filing, cf. U. S. Expansion Bolt Co. v. Jordan Indus., Inc., 488 F.2d 566, 568 & n. 3 (3d Cir. 1973) (one looks to the time of the application in determining obviousness under 35 U.S.C. § 103). [11] This case is thus unlike United States v. Adams, 383 U.S. 39, 51-52, 86 S.Ct. 708, 15 L.Ed.2d 572 (1966), where the prior art specifically taught that the invention sought to be patented was impractical and that the use of certain chemicals specified in the patent would be detrimental to the process. [12] The District Court interpreted the Hodnett patent as follows: While certain broad language in both the Hodnett and Wilson patents refer to a wide spectrum of forms of finely divided carbon, what the Hodnett patent really covers is the use of certain forms of graphite. [App. 474.] This finding is at least curious. As it relates to obviousness under section 103, what is important is what Hodnett taught as a disclosure of the prior art, not what it may or may not have “covered.” As just noted in the text, the Hod-nett patent explained that both carbon black and graphite were useful in practicing the invention. That the patent expressed a preference for graphite does not negative the fact that it also taught that carbon black was useful. Indeed in 1968 Basic’s patent counsel advised that Hodnett disclosed the use of carbon black generally. In addition, as the finding may relate to the scope of the invention claimed by Hodnett, carbon black was plainly within the “coverage” of the patent because it was specified in the claims of the patent. Under American patent law, the scope of any invention is measured by its claims, Dickstein, supra, 522 F.2d at 1297, although these must be read in light of the patent’s specifications. United States v. Adams, supra, 383 U.S. at 49, 86 S.Ct. 708; Philips Indus. Inc. v. State Stove & Mfg. Co., 522 F.2d 1137, 1140 (6th Cir. 1975). [13] Basic’s carbon expert testified to his “amazement and surprise” when he first learned of the level of improvement that could be derived from the addition of only 2%-3% carbon black. Based on his own work in carbon bodies (not refractories), he would have expected only a negligible improvement in strength and density. We think that this testimony was of no particular persuasive value since Hodnett had already disclosed the quantity of free carbon necessary to effectuate a better brick. Wilson’s patent only changed the material added, it did not alter the prior teaching as to the quantity to be used. [14] The District Court made the rather cryptic finding: There was evidence from which it can be concluded that some chemical reactions took place between the thermal black and the pitch which caused the latter to become very much harder and stiffer than normal. [App. 472-73.] To the extent that this is intended to relate to the independent effect that the free carbon might have on the pitch binder, which in turn, may affect the strength of the refractory bricks, it has some support in the record (particularly in the specifications of the Wilson patent, column 2, lines 34-39, and in the testimony of Stanislav Mrozoski, App. 265). To the extent that it attempts to explain the overall process, however, it is clearly erroneous. It is apparently based on the ambiguous comment of one witness who admitted that he had no empirical support for it. App. 370. This “theory” was directly rebutted by another expert witness and was overwhelmed by the other evidence of the process of “particle stuffing.” The evidence included documents provided by Basic to their patent attorneys which explained the process in terms of particle stuffing. Defendant’s Ex. FR at 2. [15] See note 18, infra. [16] The District Court made the following finding: The Wilson patent excludes the particular variety of graphite covered by the Hodnett patent, since Wilson’s experiments demonstrated that that variety of graphite decreased the density of the pitch-bonded refractory bricks, an undesirable characteristic. [App. 474-75.] This finding is clearly erroneous in two respects. First the Wilson patent nowhere “excludes the particular variety of graphite covered by the Hodnett patent,” because it nowhere discusses particular varieties of graphite. The most it does is express a preference for carbon black over graphite generally. Second, Wilson’s experiments, which are described in the text above, clearly do not show that Hodnett’s preferred variety of graphite decreases the density of refractory bricks. To the contrary, the report plainly shows that fine graphite, Hod-nett’s preference, markedly increased and improved the density and strength of the bricks over those with no carbon additive. [17] The test results showed the following: [18] Although Basic entered this litigation with the benefit of the statutory presumption of va *755 lidity under 35 U.S.C. § 282 noted above, it is axiomatic that its limited force can be weakened or destroyed where it is shown that the most relevant prior art was not disclosed to the patent examiner. American Seating, supra, 586 F.2d at 615; Reynolds Metals, supra, 548 F.2d at 160; Bolkcom, supra, 523 F.2d at 498; Tee-Pak Inc. v. St. Regis Paper Co., 491 F.2d 1193, 1196 (6th Cir. 1974). Similarly in this case we believe that the presumption was seriously weakened when Basic failed to disclose to the patent examiner the results of tests which came closest to comparing Hodnett and Wilson; and when Basic plainly misrepresented the nature of the tests reported in the Hyde affidavit. Because the examiner did not have the best and most accurate information before him we cannot entertain the usual presumption of the correctness of his conclusion. [19] The terms “pressed density” and “green density” refer to the density of a newly formed brick. Similarly, the pressed or green compressive strength of a brick related to the physical properties of a new brick. In contrast, the “coked” properties of a brick refer to its characteristics after it has been exposed to the steelmaking process. The expert testimony showed that coked properties are more meaningful because they relate to the conditions that the bricks actually experience in the basic oxygen furnace. [20] See note 19, supra. [21] The actual test results were as follows: [22] One other piece of prior art deserves mention. The Swallen patent, U.S. No. 2,527,595 (1950), disclosed a method for improving the strength of carbon electrodes used in electric furnaces. These electrodes are produced by combining carbon flour with a pitch binder and baking the mixture to form a carbon body. The patent disclosed that the electrode could be made more resistant to chemical attack, and also stronger, through the addition of thermal black. Again this process involved essentially void-filling. Thermal black was chosen because of its range of particle sizes. The precise process for making the electrodes differs from that used to make refractory bricks. The quantity of thermal black used also differs (up to 40% by weight can be used in the Swallen process). Despite these differences, the purpose of the additive is similar to that involved in the Wilson and Hodnett processes and the method of manufacture is analogous. Therefore we think that the Swallen patent, as prior art, would at least point one toward the use of thermal black. In fact, in his deposition, Wilson stated that he had read the Swallen patent and found it useful to his work in 1961. Plaintiffs Ex. 422 at 78-79. While Swallen alone *756 would not render Wilson’s disclosures obvious, we believe that it is plainly relevant as shedding light on the general state of the art at the time of the Wilson application. See Graham, supra, 383 U.S. at 17, 86 S.Ct. 684; Fed.R.Evid. 401. Thus we believe that the District Court erred in concluding as to the Swallen patent: [T]his is totally irrelevant, if for no other. reason than that electrodes use little or no refractory material such as dead-burned magnesia or dolomite, and are made by processes almost entirely different from those used in making the refractory bricks involved here. [App. 475.] [23] Although Basic does not discuss this point, we note that the purpose stated in the specifications to the Hodnett patent refers principally to the chemical improvement that can be obtained by adding free carbon. The patent does not specifically refer at that point to improved strength and density. Assuming, ar-guendo, that Hodnett was ignoring these important physical properties, we do not think that this is a basis on which to distinguish Wilson from ftodnett. The fact remains that the evidence tended to establish that the two processes achieve an essentially equivalent result and that the use of either carbon black or graphite was taught in the Hodnett patent. [24] This holding renders it unnecessary to discuss Eltra’s claim that Basic committed fraud on the patent office. Therefore we express no opinion on its merits.