NORTON COMPANY v. Carborundum Company

OPINION 1 ALDRICH, Senior Circuit Judge. * In this action for patent infringement defendant asserts the usual defenses of invalidity and non-infringement, and adds a fillip—fraud in obtaining the patent and use of the patent in restraint of trade. The patent is No. 3,181,939, Marshall & Roschuk, Fused Aluminar Zirconia Abrasives, applied for January 27, 1961, and issued May 5, 1965. The purported invention is a formula for an abrasive product of high impact strength, suitable for heavy work such as snagging (removal of projections from) stainless steel. The word formula is appropriate because the process described, apart from the mix, was well known. The originally purported inventors were two employees of plaintiff, Norton Company, of Worcester, Massachusetts, assignee. Subsequent to the commencement of this action plaintiff moved to add the name of one Thibault, another of its employees, as a joint inventor. This motion is opposed. Plaintiff, and defendant Carborundum Company, Carbo, are internationally known manufacturers of abrasives, and apart from marked courtesy and cooperation of counsel, e. g., a 120-page stipulation, are at substantial loggerheads. *641 Claim 1, the principal claim of the patent, somewhat abbreviated, 2 was particularized by Norton at the pretrial as providing, a. A fused abrasive material, b. “consisting essentially” of alpha-alumina (A1203) and zirconia (Zr02), c. containing less than 0.1 percent soda (Na20) by weight, d. containing from about 10 to 60 percent zirconia by weight, e. containing an alumina-zirconia eutectic 3 surrounding primary crystals of alpha-alumina (or crystals of zirconia if zirconia above about 45 percent was used in the mix, f. the average size of the primary crystals (part e) to be less than 300 microns, g. the material to be one having a high impact strength, h. the material to be suitable for snagging stainless steel. The material, once cooled, is crushed to produce grains, or grits, of the desired size (each containing a number of crystals), which are subsequently bonded into grinding wheels, or which may be affixed to a flexible backing, hereafter sandpaper. When the raw mix has been melted in a refractory furnace the patent calls for iron mold, and hence rapid, cooling of the mix. It has long been known that rapid cooling inhibits crystal growth. See, e. g., Encyclopedia Britannica, 11th ed., 1910, Crystallization. It was also known that smaller crystals in fused alumina increase its abrasive strength. See, e. g., Tone, Crystalline Fused Alumina and the Manufacture Thereof, Patent No. 1,192,709, July 25, 1916; Kalmus, Method of Making Abrasive Material, Patent No. 1,226,892, May 22, 1917. The cited patents themselves taught that the process utilized by the ’939 patent could produce crystals of the sizes it specified. I attach no significance in this being for alumina alone, and read the crystal sizes in the patent as limitations, and not in any respect inventive. On the other hand, I reject as factually unpersuasive Carborundum’s claim that these crystal sizes, even in the early 1960’s, could not be measured, and were insufficiently described, so that the patent should fall for inadequate disclosure. Some other matters may be disposed of at the outset. As to items a and b, ante, it was not invention to report that mixing alumina and zirconia, in some proportion, produced a fused abrasive of some desirable characteristics. Nor was it inventive to report (item e) that such a fusion would contain a eutectic. Although it was known that a low soda content, unless corrected, was needed to achieve a strong abrasive suitable for snagging, and item (c) was essentially a limitation, it was somewhat inventive to realize that the circumstances required so small an amount. Finally, items g and h were descriptive, but accurately so. In sum, I find that the invention, if there was one, related principally to the high zirconia content, (item d). 4 The event of primary historical importance was the issuance to Saunders *642 and White, jointly, Norton’s assignors, of two patents, No. 1,240,490, Composition Containing Alumina and Zirconia, and No. 1,240,491, Aluminous Abrasive and Method of Preparing the Same, on September 18, 1917. Both of these patents dealt with an alumina/zirconia fused abrasive. The ’490 patent started with a pre-fusion, or raw, mix composed of alumina and zirconia, hereinafter an alumina mix, while ’491 started with a bauxite—the source of alumina—and zirconia raw mix, hereinafter a bauxite mix. ’490, which did not specify the proportions, allegedly produced a powdery (friable) abrasive, denominated as suitable for polishing. ’491, too, did not limit the amount of zirconia, but it did make particular mention of a range of two to five percent, see n.7, post, and allegedly produced a harder abrasive that, like the ’939 patent, was suitable for snagging. Carbo relies on these patents as showing lack of invention, and on Norton’s alleged concealment, or at least failure to cite the ’491 patent to the Patent Office, as part of fraudulent conduct with respect to the patent. The next event of historical importance was that in October, 1951 a Norton employee, Thibault, wrote a letter to a fellow employee proposing an abrasive material to be made by fusing and reducing a mixture of about half bauxite and half zircon and casting the product. Although Thibault remarked that this might result in a snagging abrasive, the proposal was for a tumbling abrasive. A number of fusions, possessing various amounts of alumina and zirconia, were made thereafter and tested for tumbling. Marshall participated in this activity, and on his own included experiments using alumina in place of bauxite. The tests were unrewarding and the undertaking was abandoned. Roschuk came to Norton in 1957. He had no contact with Thibault. He and Marshall undertook some experiments looking towards a snagging abrasive using a pre-fusion mix of alumina with a high zirconia content. This ultimately led to application for the patent in suit. Before reaching questions of invention and infringement I consider first the scope of the patent, and the question whether any fraud, in the broad equitable sense, hereinafter misconduct, was engaged in by applicants. Since the two are interrelated, a single review of events in and off the file wrapper will cover both. The claims of the patent are cast in terms of the product—“A fused abrasive material consisting essentially of a mixture of alpha-alumina and zirconia.” The first question is whether the patent, as drawn, covers a bauxite mix. Carbo says that it does not because the product of such a mix contains other materials which cause it to fall without the description “consisting essentially” of alumina and zirconia. I find against it on this point. “Essentially,” or “consisting essentially,” is inexact. If the patent elsewhere made clear that it covered a bauxite mix, I would find no necessary inconsistency in the use of this phrase even though such a raw mix left in the final product a small amount of other elements of some conceded consequence that would not have been introduced by an alumina raw mix. On the other hand, the use of this phrase does not affirmatively suggest a bauxite mix, see, e. g., Saunders ’490, lines 54-55, and this presents a problem. The fact that the patent description is inexact makes significant what the patent contemplates, or permits, as going into the mix. Norton, itself, asks me to find that “[o]ver the years it has not proven possible to predict what will and what will not be an abrasive of given performance or utility.” Bearing this out, Norton’s position is that the whole invention of the ’491 patent relates to the presence of but two to five percent zirconia. This approximates the range of the additional impurities introduced by a bauxite as against an alumina mix. It is not to be characterized as within *643 the doctrine of equivalents, or dismissed , as insubstantial. Hence it is oversimplistic for Norton to say that Carbo’s accused products which result from a bauxite mix read on its claim. Rather, the question is what is the meaning of the claim when read in the light of the specifications and the file wrapper. I find that in fact it does not ■ cover a product of a bauxite mix. The factors that lead me to that conclusion require detailed consideration. While the product received the customary introductory applicant encomiums, the initial remarks dealt also with the raw mix. “More particularly, it has been found that high purity alpha-alumina may be mixed with zirconia, also of high purity, in the proportions indicated hereinafter. The mixture is fused to a molten state in, for example, an arc tapping furnace and the fused product is quickly cooled.” (Ital. suppl.) Later it was added, “Alpha-alumina which may be successfully utilized in the process of this invention is preferably of high purity, usually at least 99.8% by weight A1203 and containing less than 0.1% sodium oxide. Another form of alumina which may be■ used is of a slightly lesser degree of purity and may contain up to as much as 0.4% sodium oxide. In this instance, about 0.5% carbon is included in the mix for the purpose of reducing the sodium oxide to less than 0.1% during the fusing procedure. The purity of zirconia is preferably at least about 99% Zr02” (Ital. suppl.) These paragraphs, hereinafter the raw mix specifications, stop here; there is no mention of bauxite as “another form of alumina which may be used.” Although the patent might well cover an even less pure pre-fusion alumina, in view of the substantial difference between alumina and bauxite, which the ’490 patent itself makes note of, the omission of any suggestion of bauxite seems eloquent. It would be, at best, misleading, for Norton to commence by speaking of high purity alumina, and then explain that by certain procedures one can use alumina of somewhat less purity, and stop there, if it proposed the inclusion of substantially more impurities by not starting with alumina at all. A patent monopoly is given not simply to reward the inventor; a prime consideration is the ultimate benefit of the public. It is basic that to accomplish this the patent must teach, not hide. Confirmation of what the proposed patent taught is to be found in the process claims included in the original application. Process claims 8 and 9 decline a “raw batch” precisely in accordance with the raw mix specifications, and claim 10 recites a raw mix of alumina and 10-60% zirconia. There is no mention of bauxite. Nor, incidentally, was this a fault of draftsmanship. Marshall testified that this was his own understanding. It is true that the raw mix claims were withdrawn. This, however, was because applicants, faced with an almost meticulously negatively-minded examiner even as to the product claims, wisely abandoned their process claims to concentrate on the product. The cancellation was not accompanied by any change in the specifications to suggest an enlargement in scope. The single item that Norton can point to is a remark in the April 2, 1962 response to the examiner’s rejection of October 6,1961. The response stated, “The abrasive product produced by this invention is, chemically, AL203 and Zr02 with minor amounts of impurities introduced with the source (such as bauxite) of alumina and the source of zirconia.” Very carefully examined, this may be read as affirmatively encompassing a bauxite mix. Whether that was its purpose I will come to later. I find, however, that it was sufficiently ambiguous and inconspicuous to give no notice of such an intent. As the patent itself points out, even an alpha alumina raw *644 mix may be expected to contain impurities originating from the bauxite. If this remark had been accompanied by the withdrawal of the process claims, or a change in the specifications, some attention might have been invited to it, but the withdrawal did not occur until October 11, 1963. It is to be noted that even on that date applicants reaffirmed the implication I find in the raw mix specifications, stating, “The present situation in which the applicant [sic] premixes zirconia and alumina, and heats them to the fusion point . . .” (Ital. suppl.) . On the same date, see post, Norton cited, and distinguished, ’490, but made no mention of ’491. It now says that ’490 itself refers to the ’491 application, and its use of bauxite. That is true, but the reference is to distinguish. From the examiner’s standpoint, if he knew of ’491 before the patent issued, Norton’s failure to cite it would confirm an impression of Norton’s concern with an alumina and not a bauxite mix. As against this Norton notes that after the application was granted, but before the patent actually issued, Carbo’s employee Cutts, whose application for an abrasive patent had been disallowed, filed interference proceedings in which he ultimately repeated claims 1, 2 and 3 of the ’939 patent, and claimed prior invention. In the course of these proceedings Cutts asserted that the ’939 patent, as well as being a valid invention, encompassed a bauxite mix. The interference was ultimately rejected upon a finding that Carbo’s date of invention was too late. The grapes, whose sweetness Carbo had extolled, proving to be out of reach, Carbo now professes them to be sour. No principle of estoppel bars this reversal. Haughey v. Lee, 1894, 151 U.S. 282, 14 S.Ct. 331, 38 L.Ed. 162. Non constat, Norton could properly introduce Carbo’s prior interpretation as an admission. LaSalle Street Press, Inc. v. McCormick & Henderson, Inc., 7 Cir., 1971, 445 F.2d 84; Stedman Mfg. Co. v. Redman, 4 Cir., 1958, 257 F.2d 867, cert. denied 358 U.S. 928, 79 S.Ct. 314, 3 L.Ed.2d 302. I must agree with Norton that Carbo’s present, and strenuous, explanation that it “had to” copy the claims, is pure nonsense in the way that Carbo now puts it. It “had to” copy if it wished to prosecute an interference, but if it did not want the patent, or felt it did not apply to a bauxite mix, it did not have to claim the interference. 5 However, at the same time, Norton is faced with the fact that its own representatives, by in-house communications, had on several occasions agreed with Carbo’s present position. I will deal with these more fully in another connection, but for present purposes I find the battle of admissions to be a Mexican standoff. Norton, however, makes another point, based on the fact, which I find to be true, that the same assistant examiner, one Arnold, did all of the work on the ’939 patent until the final allowance and also the work on the Cutt application and subsequent interference. While processing that application, and after the allowance, but before the final issuance, of the ’939 patent, Arnold cited the ’491 patent. I can agree with Norton that although Arnold, as an assistant examiner, could not formally allow its patent, he recommended its allowance to the primary examiners, and if, on hearing of ’491, he had felt he had been diddled by Norton, he might well have taken steps. But if Arnold thought of ’939 as limited to an alumina mix, he would not have thought he had been diddled. There are various suppositions that can be made by Norton, but I find none persuasive. I do not believe it should be rescued by subsequent collateral events and inferences that I find, at best, spec *645 ulative. Norton had the key to its own deliverance, and for reasons already discussed, and those about to be discussed, I do not find it saved. I remain of the view that directly on its face, or read in the light of the file wrapper, the patent does not disclose, or purport to cover, a bauxite mix, and consequently would not be infringed by material so produced. I turn now to certain matters which lead to the conclusion that if, contrary to my interpretation, the patent should be construed as including a bauxite mix, such a claim cannot be enforced because of what, under the circumstances, would amount to Norton’s disabling misconduct. Prior to filing the application applicants, and Norton’s patent counsel, were aware of the Saunders ’490 and ’491 patents. After the filing Norton learned that Carbo was manufacturing an alumina/high-zirconia abrasive produced from a bauxite mix. Norton’s counsel made a note in the file, “In future prosecution of this application we should attempt to draw our claims with sufficient breadth to cover [Carbo’s] composition.” Pursuant to this, counsel wrote to Norton’s Abrasive Division Manager in April, 1962, that their application, “does not suggest the direct use of bauxite as the source of alumina. “We do not know .at this time whether or not we will obtain claims broad enough to cover an abrasive made from a fusion of bauxite and a source of zirconia. We will attempt to obtain such coverage and have a reasonable chance of success.” (Ital. suppl.) Contemporaneously counsel filed with the examiner the “ . . . introduced with the source (such as bauxite)” response previously quoted. Counsel testified before me that he made this “amendment to make it clear that we intended that the claims we hoped would cover materials made with bauxite because we were aware that just prior to that that Carborundum had come out with such a product.” However, Norton made no amendment of its claims. In December, 1962, it even abandoned a projected filing of a separate application for a bauxite mix because, according to its notes, to “emphasize the bauxite/zirconia approach . . . might hurt our chances of obtaining some protection on alumina-zirconia.” Norton did more than not emphasize the bauxite approach. As already mentioned, the examiner was not taken with this application. This was so even though he failed to cite the Saunders patents. On June 14, 1962, Norton made a note in its file, “The Saunders patent [sic] seems to be the one to consider as a reference to prior art. We should distinguish Saunders before the examiner quotes it against us.” 6 In Norton’s next communication to the examiner, on October 11, 1963, it distinguished two of the examiner’s references, and added, “More pertinent to the above references to the present invention is Saunders patent 1,240,490.” Counsel then proceeded to distinguish it. He made no mention, however, of Saunders 491. If the pending application was intended to cover a bauxite mix I find that the ’491 patent was far more pertinent than the ’490 patent. By volunteering ’490, however, Norton would appear to be commendably candid. Justice Holmes is said to have characterized candor as the most effective form of deception. See E. J. Bander, Justice Holmes Ex Cathedra, [402] (1966). In the present case, as already noted, if one assumes good faith, to cite ’490 and not mention ’491 is to lead away from any consideration of a bauxite mix. Although some cases suggest, see In re Multidistrict Litigation Involving Frost Patent D.Del.1975, 398 F.Supp. 1353, an affirmative duty to volunteer prior art in some-' circumstances, I find Norton’s conduct, in cit *646 ing one patent and not the other, not mere nondisclosure but materially different than if it had cited neither. Even if it could be found that Norton did not, at the time, intend to claim a bauxite mix, in view of the fact that it sued Carbo for infringement by products originating from a bauxite mix it must, in legal effect, be charged with such purpose nunc pro tunc. Whatever may be the limits of file wrapper estoppel, a patent proceeding is a single transaction. Norton could not take advantage of conduct on its part, although initially innocent, by making a claim in the light of which that conduct would be misconduct. I do not find any pertinent patent case, but under general principles, at least wlien the relationship of the parties, as here, calls for a high degree of fair dealing, the law is clear. Cf. Stipcich v. Metropolitan Life Ins. Co., 1928, 277 U.S. 311, 48 S.Ct. 512, 72 L.Ed. 895. I find, moreover, consistent with patent counsel’s admitted purpose in his April 2, 1962 oblique response regarding bauxite, that Norton did have that intent. Norton confirms that today. Referring to this response, its reply brief states, “The record was thus set straight with the Examiner, and all who later would read the patent and its file, that the claims with the ‘ . consisting essentially . . .’ clause in them, were broad enough to cover [be infringed by] products made by use of bauxite in the fusion mix.” But it must be clear that if Norton truly wished the world to be aware that it was claiming a bauxite mix, the place to say so was in the claims themselves, or at least in the specifications which, without it, pointed in the opposite direction, and not by a half-hidden implication, itself buried in an extended prosecution file. Rather, I find that this was a felt-slippered remark, an “unemphasized,” carefully measured maneuver made when it saw its invention suffering a serious blow by Carbo’s employing bauxite. Although cognizant of the considerable burden with respect to fraud that is upon Carbo, I find that on and after April 2, 1962 Norton was planning to use the patent, if it received it and nothing of a contradicting nature had occurred, to claim infringements by products originating from a bauxite mix, to the extent that its failure to include a reference to the ’491 patent was affirmatively deceptive, preventing its now pressing such a claim. Precision Instrument Mfg. Co. v. Automotive Maintenance Mach. Co., 1945, 324 U.S. 806, 815, 65 S.Ct. 993, 89 L.Ed. 1381; Carter-Wallace, Inc. v. Davis-Edwards Pharmacol Corp., 2 Cir., 1971, 443 F.2d 867, 881-882; Beckman Instruments, Inc. v. Chemtronics, Inc., 5 Cir., 1970, 428 F.2d 555. Where it consciously avoided “emphasiz[ing] the bauxite/zirconia approach [lest it] hurt our chances on obtaining some protection on alumina-zirconia,” it should achieve no larger objective. This brings me to a further aspect of the fraud defense: does the not citing of ’491 affect the enforcibility of the patent if narrowly construed to cover only products resulting from alumina raw mixes? I am in full accord with the court in Beckman Instruments, Inc. that if misconduct prevents enforcing the claim as made, it is no answer that it would not have operated against a narrower one. That, however, is not this case, because I have interpreted the claim itself as being narrow. Hence the question is the relevancy of the non-disclosure of ’491 vis-a-vis an alumina raw mix. This question must be faced because one of Carbo’s products, the BR sandpaper, originated from an alumina mix, and if an alumina mix produced a valid invention, BR infringed. Although hardly the achievement of the century, I find there was considerable utility in Norton’s product. There would have been more had a bauxite mix been included—I am unfavorably im *647 pressed by Carbo’s denigration of the high zireonia high impact snagging accomplishments, which I find could have been used much earlier, had it been discovered. Alumina is a good abrasive, and I find that zireonia costs more than alumina, and the product sells for more, but, nonetheless, much is sold. It is no answer that Norton’s product is partly met by Carbo’s non-infringing product. I find there -is a difference in accomplishments, but I also consider that the fact one can thereafter achieve substantially the same result in a non-infringing manner does not render a patent invalid. The question is close. I give weight to the presumption of validity, and some to commercial success. Carbo’s complaint as to the former I will return to later. On the subject of obviousness, I would not agree with Carbo that this is disproved by Carbo’s almost contemporary discovery even if Norton had embraced bauxite. Norton, Carbo, and others expend much effort in seeking to devise new products, and what I find significant is the fact that this one was not discovered earlier. Nor, at least for Norton’s narrow claim, was it anticipated, or taught to one skilled in the art, by the prior art. I accept the testimony of Norton’s expert that just what would accomplish what in this field was often highly problematical, and that the reasonable expectation had been that adding more zireonia to high purity alumina would not produce a snagging abrasive. It is true that the ’491 patent showed good results when zireonia was added to bauxite, but this was another matter. Furthermore, I agree with Norton’s witness that ’491’s reference to two to five percent zireonia is more reasonably read as being the optimum teaching. 7 In this Norton finds some support in the Cutts interference proceedings, for there it was represented to the examiner that the ’491 statement that “larger proportions of zireonia may be used, is so vague as to be no disclosure.” I find that ’491’s statement that larger proportions could be used was simply for protection against a party who sought to avoid infringement by departing from the exact 2-5% recitation. In that sense Saunders’ claim would be a representation that a higher zireonia content would not destroy the claimed advantages of the invention, but not a disclosure, such as Carbo now argues, that by increasing the zireonia content the benefits would be further enhanced—and particularly not the substantial enhancement that the ’939 invention disclosed. Returning to the matter of non-disclosure, the trouble here is that there is room for arguing invalidity, which is why I find Norton chargeable for not citing '491, if its application included a bauxite mix. An applicant who seeks the advantages of a presumption of validity should not conceal prior art that the examiner might think relevant. The presumption attaches because of the Patent Office’s expertise, which cannot operate without the facts. It has been said that the test is whether the knowingly non-disclosed art might reasonably be thought relevant, cf. In re Multidistrict Litigation Involving Frost Patent, D.Del., ante, but even if the test is could-not-reasonably-be-thought-irrelevant on a bauxite mix claim ’491 would qualify. Even accepting my interpretation of ’491’s two to five percent provision, a reluctant examiner ■ might think it not sufficiently inventive to discover that while that amount was thought best, ten percent or more was in fact better. The fact that Arnold in this case may not have so reacted is not dispositive so far as a bauxite mix is concerned. But even applying these principles I find that Norton’s failure to cite ’491 was not' disabling vis-a-vis an alumina mix. In addition, I reject Car *648 bo’s claim that Norton was guilty of other acts of disabling non-disclosure, particularly of failure to give the examiner certain other material that was in its files. Norton v. Curtiss, 1970, 433 F.2d 779, 57 CCPA 1384; United States v. Standard Elec. Time Co., D.Mass., 1957, 155 F.Supp. 949, app. dism. 1 Cir., 254 F.2d 598. Carbo contends that the patent is invalid for failure to name the proper inventor. 35 U.S.C. § 116. Claiming that Norton is between two horns of a dilemma, Carbo asserts that either Thibault was the sole inventor, or Norton’s present motion to add Thibault as a co-inventor should be denied because the initial failure to name him was done in bad faith. I reject both of these contentions. I find that Thibault’s 1951 suggestion was insufficiently reduced to constitute an invention at that time, or to be subject to the rule, Ansonia Brass & Copper Co. v. Electrical Supply Co., 1892, 144 U.S. 11, 12 S.Ct. 601, 36 L.Ed. 327, that additional advantages to a known product do not constitute inventiveness. I draw the same distinction between Carbo’s attempted assessment of the substance of Thibault’s “product” and accomplishment, and the degree it fell short of it, as I do between Carbo’s exaggerated characterization of Thibault’s suggestion during oral argument (a “recognition by him that it could be used for snagging purposes”), and Thibault’s actual language (“the material might show some interesting results in resinoid snagging.”) At the same time, I find that Thibault’s contribution was sufficient to warrant his inclusion as a co-inventor, and correspondingly forbids Carbo’s reliance on work done in consequence thereof as disabling prior art. Since I find that the failure so to include him prior to the institution of suit was an excusble inadvertence, and not attributable to bad faith, I allow Norton’s motion pursuant to 35 U.S.C. § 256 to add Thibault as a coinventor. 8 This brings me to the counterclaim, whether Norton’s conduct in seeking to enforce the patent against a bauxite mix product, or its misconduct in the Patent Office proceedings was so serious, that it can be charged with seeking to monopolize in restraint of trade. 15 U.S.C. §§ 2, 15. Compare Walker Process Equipment, Inc. v. Food Mach. & Chem. Corp., 1964, 382 U.S. 172, 86 S.Ct. 347, 15 L.Ed.2d 247, with Precision Instrument Mfg. Co., ante. As to Norton’s accusing Carbo’s bauxite mix products, Carbo’s indignation must be read in the light of its own contentions in the interference. The ’939 file wrapper was fully known to it. When it thought it to its own advantage it sought to construe the patent both as valid and as applying to a bauxite mix. Indeed, it may be this very activity on its part that encouraged Norton to bring this action. I am not satisfied that Carbo has sustained its burden of proving Norton guilty of legally actionable fraud, as distinguished from mere disabling misconduct. Cases, ante. But even if it had, I would regard it as grossly offensive for Carbo, in light of its own conduct, to turn around and sue Norton for triple damages. I recognize that private antitrust actions are in part for the public benefit, but I think the public has been adequately protected as it is, and although I find no cases in point, one way or the other, I will dismiss the counterclaim. Cf. Allbright-Nell Co. v. Autosteam Process Co., 7 Cir., 1934, 70 *649 F.2d 959, 962, where the court observed that an admission of infringement by a particular defendant was not affected with a public interest, and could not be withdrawn. Finally, Carbo asks for attorneys fees because this is an “exceptional case.” 35 U.S.C. § 285. Whatever sympathy I might otherwise have evaporates in the light of Carbo’s extensive interference proceedings. Carbo cannot hunt with the hounds and run with the hare. SUPPLEMENTAL OPINION Norton’s motion for amplification of the June 18, 1975 opinion contained two misconceptions. Although the exact proportions do not appear, Carborundum’s products (other than BR sandpaper) are the result of a raw mix not only of bauxite and zirconia, but of scrap residue of a previous fused mix. Norton would equate this residue with alumina within its patent. This is error. Where Carborundum’s bauxite mix is not within the patent, it does not become so by being used over again, absent express teaching, but even if the scrap were to be regarded as alumina, there is a further error. Norton suggests, on the assumption that Carborundum uses alumina in addition to bauxite, that the alumina infringes its patent pro tanto and Carborundum should be accountable in that proportion. If Norton had a patent on which combinations of A and B and Carborundum added a small amount of C, not sufficient to vary the basic teaching of the patent, fairness might dictate, if damages were to be measured by weight or volume, that the amount of C be subtracted. Even if, because -of the large “Loss” that was testified to, there is considerable reusable scrap, Carborundum clearly must use a substantial proportion of bauxite, or there would soon be no more production. This is a significant departure from the patent. Norton’s discovery was that one can add a substantial proportion of zirconia to alumina, not that one can add zirconia to bauxite, or zirconia and alumina to bauxite. SUBSTITUTE INTERLOCUTORY JUDGMENT AND FINAL JUDGMENT 1. Patent No. 3,181,939, Fused Alumina-Zirconia Abrasives, Marshall and Roschuk, now ascribed to Marshall, Roschuk and Thibault, May 4, 1965, is to be read as limited to products originating from a raw mix consisting essetially of alumina and zirconia, and not to include a raw mix containing bauxite. 2. As so read the patent is valid and enforcible. 3. Carborundum’s BR sandpaper, so-called, but none other of its accused products infringes the patent. 4. The counterclaim asserting violations of 15 U.S.C. §§ 2 and 15 is dismissed. 5. Damages for infringement, and the allocation of costs, are reserved. The court determines that there is no just reason for delay and expressly orders and directs that the judgment contained in paragraph 4 hereof is a final judgment. Footnotes [1] Certain matters of clearly ephemeral interest are placed in fractional footnotes appearing in an appendix to the original opinion, but omitted from publication. [*] By designation. [2] Norton’s abbreviation omits two clauses denominated therein (a) and (b), on which Carbo makes a construction argument which I reject. [3] Eutectic, in terms of two components, is an alloy or solution having the components in such proportions that the melting point is the lowest possible with these components (and lower than either individually.) A eutectic mix is composed of very fine crystals (fines), each containing both primary components. If the proper proportion of components is fused, only the eutectic will re-suit. If the proportion of one component is increased there will be a eutectic mix plus separate, larger, crystals of the excess component. The patent specifications correctly recite, “These [“primary”] crystals are generally separated from each other (or cemented together) by areas of the naturally fine eutectic.” [4] As did the parties, I use the term high as meaning 10% and over, as distinguished from low, particularly the 2 to 5% zirconia specifically mentioned in the Saunders ’491 patent, post. [5] Carbo’s counsel’s statement during oral argument, that he was “perfectly willing to assume” that if Carbo had won the interference it would not have made these broad claims, is a singular “assumption.” I draw the opposite inference. [6] Coneededly, this was the ’490 patent. Not to be hypocritical, the expressed motive seems one of calculation, not of good faith disclosure. [7] “Zireonia additions in amount as small as two to five per cent are sufficient to produce strongly marked effects of the character indicated above. Either smaller or larger proportions of zireonia may be used.” [8] However, I do not, by allowing this motion, enlarge the concept of the patent to include bauxite, and it may be that on an alumina mix basis an argument could be made that I should not grant it at all. No one has made that point. Carbo makes the opposite claim, that if Thibault is the inventor, he is the only one. This I reject. It is often difficult to determine who is, and who is not to be included as an inventor. Cf. Mueller Brass Co. v. Reading Industries, Inc., E.D.Pa., 1972, 352 F.Supp. 1357. The important finding is that there was no deceptive intent in not naming Thibault in the first place. On this record I see no purpose in pursuing the matter further.