In this case we are asked, in part, to decide the quantum of proof necessary to recover lost profits in an action for "reasonable and entire" compensation for infringement of a patent by the government under 28 U.S.C. § 1498(a) (1994), but decline to do so because it is not necessary to the resolution of the case. Gargoyles, Inc. and Pro-Tec, Inc. (collectively, "Gargoyles") appeal the February 29, 1996 judgment of the United States Court of Federal Claims1 in 342-88C which awarded Gargoyles a reasonable royalty but denied lost profits because Gargoyles had not demonstrated entitlement, by "strictest proof," under the tests in Panduit Corp. v. Stahlin Bros. Fibre Works, 575 F.2d 1152, 1156, 197 U.S.P.Q. (BNA) 726, 730 (6th Cir.1978) (Markey, C.J., sitting by designation), and Tektronix, Inc. v. United States, 213 Ct. Cl. 257, 552 F.2d 343, 348-49, 193 U.S.P.Q. (BNA) 385, 391, modified, 213 Ct. Cl. 307, 557 F.2d 265 (1977). Gargoyles argues the trial court erred by not applying the lower, "preponderance of the evidence," standard. The government cross-appeals, arguing: (i) the award of a 10 percent royalty rate was excessive in light of asserted errors in the trial court's fact-finding, and what the government perceived as Gargoyles' low expected profits; and (ii) Gargoyles' patent is invalid under 35 U.S.C. § 112 p 1 for failure to disclose the inventor's "best mode." The appeal was submitted for our decision following oral argument on January 29, 1997.

Because we hold that on this record the trial court correctly denied lost profits under either possible standard of proof, we affirm the damages award and do not address the issue of what quantum of proof is necessary to show lost profits in section 1498 actions. On the cross appeal, we do not discern clear error in any of the relevant fact findings or error in any legal conclusions challenged by the government relating to royalty rate or best mode. Accordingly, we affirm the judgment of the Claims Court.

Gargoyles, Inc. received by assignment from Pro-Tec, Inc. in 1988 ownership of U.S. Patent No. 4,741,611 (the " '611 patent") for protective eyewear, and thereafter commenced an action under 28 U.S.C. § 1498 for "reasonable and entire" compensation for direct infringement of the '611 patent by the United States. The suit was based on the Army's use of several thousand pairs of accused ballistic/laser protective spectacles ("B/LPS") procured from American Optical ("AO"). Gargoyles' patent, generally stated, is for "toric wrap" safety eyewear which covers the eye while avoiding the appearance of prior art spherical lenses. Gargoyles sells its eyewear commercially.

While the details of the patent claims are not relevant here, Gargoyles conceded in the trial court that the preferred embodiment section of the patent specification contains several errors. It states that toric lenses "are of uniform thickness throughout and thus have zero power" (col. 2, ll. 67-68), while in fact in order to have zero power, toric lenses cannot be of uniform thickness. The specification also lists a diopter value of -4.7 for the vertical meridian of the lens (col. 3, l. 4), but such a value together with the other values provided cannot achieve a zero power lens as claimed in the patent (the correct value is -4.79). The specification also does not disclose certain preferred lens design information, such as lens power, lens thickness and location of optical centers. Gargoyles argued, however, that the inventor, Dennis Burns, had not knowingly withheld information about the best mode and that one of ordinary skill in the art seeking to practice the invention would recognize the errors and could correct them.

Development of the Army protective eyewear program began in the early 1970's when Colonel Francis LaPiana, an Army optometrist, began discussions with AO regarding the development of attractive but resilient eyewear that soldiers might wear on- as well as off-duty. AO advised Col. LaPiana that the project was uneconomical. However, in 1981, Col. LaPiana learned of Gargoyles' commercial product and spoke with the company about developing "eye armor." Col. LaPiana and Burns, Gargoyles' President and CEO, entered into discussions during the early 1980's about the use of Gargoyles' eyewear in the military. The Army conducted field tests on Gargoyles' commercial product, ultimately concluding it was unfit for Army use.

In the meantime, the Army was developing its own ballistic eyewear in both prescription--the Integrated Field Eyewear System ("IFES") at the Medical Research and Development Command ("MRDC")--and non-prescription--the Ballistic Eye Protection, Emmotrope ("BEPE") at the U.S. Army Natick Research Development and Engineering Center--forms. Col. LaPiana prepared a draft Request for Proposal ("RFP") for the BEPE project which included a requirement that the lenses be "toric wrap" so as to avoid the "bug eye look" of spherical lenses of wrap depth. The final RFP, over Col. LaPiana's objection, omitted the toric wrap requirement.2 MRDC also issued an RFP for the IFES, which did include a requirement of toric lenses with attractive appearance.

With respect to the IFES RFP, the Army awarded two contracts, one to Gentex, Inc. and one to AO (Contract No. DAMD17-85-C-5073 or the " '5073 contract"). Fifty pairs of eyewear were delivered after the issuance of the '611 patent on the '5073 contract.

Thereafter, MRDC issued an RFP for procurement of the B/LPS. The Army awarded Contract No. DAMD17-86-6079 (the " '6079 contract") to AO on January 21, 1986. AO delivered roughly 450,000 units under the '6079 contract and several smaller contracts.

The Army later awarded a contract to produce protective eyewear with cylindrical lenses, called SPECS, to Uvex Winter Optical, Inc. in 1987. By 1989, the Army, including Col. LaPiana, concluded that cylindrical lenses were acceptable, and by 1992 large scale testing evidenced that SPECS were preferred to the B/LPS by emmotrope soldiers (those who did not need vision correction). SPECS are designed only for emmotropes and meet the Army's technical requirements; the Army has, since 1990, purchased 500-600 units.

The liability phase of Gargoyles' action in the Claims Court ultimately resulted in a finding that the government, through procurement of B/LPS, infringed the '611 patent. Gargoyles, Inc. v. United States, 32 Fed. Cl. 157, 169, 33 USPQ2d 1595, 1605 (1994).3 The trial court also found no "best mode" violation had been proven, the only issue from the liability phase now on appeal. See 35 U.S.C. § 112 (1994) ("the specification shall ... set forth the best mode contemplated by the inventor of carrying out his invention"). The Claims Court then conducted a three day trial on damages, resulting in an award to Gargoyles of a 10 percent royalty on a base representing the bulk of the B/LPS units acquired by the Army, and a 50 percent royalty on a small portion of the '6079 contract representing the development phase.DISCUSSION

Gargoyles argues that instead of receiving a reasonable royalty, it should have received lost profits under Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1545, 35 USPQ2d 1065, 1069 (Fed.Cir.) (in banc), cert. denied, --- U.S. ----, 116 S. Ct. 184, 133 L. Ed. 2d 122 (1995), and Panduit, 575 F.2d at 1156, 197 USPQ at 730 (four-part test as one methodology for determining lost profits). Thus, Gargoyles maintains the trial court committed legal error when, relying on Tektronix, it required that Gargoyles prove its lost profits by "strictest proof" and prove by "clear and convincing evidence" each of the four factors of the Panduit test. Instead, Gargoyles urges, under the proper standard for recovery of lost profits in the case of infringement by the government, a patent owner must show only a "reasonable probability" that, "but for" the infringement, it would have made the sales made by the infringer. Rite-Hite, 56 F.3d at 1545, 35 USPQ2d at 1069. Here, since Gargoyles maintains AO was the only competitor, Gargoyles argues this standard is readily met. See generally Pall Corp. v. Micron Separations, Inc., 66 F.3d 1211, 1222, 36 USPQ2d 1225, 1233 (Fed.Cir.1995).

The government responds by arguing Rite-Hite is inapplicable because it involved a private defendant (under 35 U.S.C § 284 (1994)) and that Tektronix dictates a stricter standard for recovery when the government is the infringer. The government argues the preferred method of compensation in a section 1498 action is a royalty, reasoning that government infringement is in the nature of a taking, rather than a tortious act. See Leesona Corp. v. United States, 220 Ct. Cl. 234, 599 F.2d 958, 966, 202 U.S.P.Q. (BNA) 424, 432-33 (1979) ("The fundamental error of the trial judge is that he has taken [section 1498] ... and ... converted it to a consent to suit on a tort theory, and the treatment of the United States as a tort-feasor."). The government concludes that any award must be fair to both the government and the patentee, essentially resulting in a compulsory royalty-bearing license. Thus, the government argues that in order to recover lost profits Tektronix requires "strictest proof the patentee would actually have earned and retained," 552 F.2d at 349, 193 USPQ at 391, the claimed profits, and that here the trial court correctly applied the clear and convincing evidence standard.

In Tektronix, the plaintiff, a manufacturer of oscilloscopes suing under section 1498, was denied lost profits damages because, in the words of the Court of Claims:

It ha[d] not been sufficiently shown by clear and convincing evidence that plaintiff would actually have supplied all the 'commercial' scopes the Government bought from the third-party respondents, or that if plaintiff had been the source of those items it would have made and kept the profits it now demands.

552 F.2d at 348, 193 U.S.P.Q. (BNA) at 391. Tektronix had granted a paid-up cross-license to another manufacturer which was capable of bidding on the contract (but did not in fact) and therefore if bidding had been limited to authorized sellers it was conceivable the licensee would have bid, won and received the profits Tektronix claimed or Tektronix would have had lower profits due to the other bidder. See id. The Tektronix court concluded, "If lost profits are ever to be awarded under § 1498, it should be only after the strictest proof that the patentee would actually have earned and retained those sums in its sales to the Government. That kind of demonstration does not exist in this case." Id. at 349, 213 Ct. Cl. 257, 193 USPQ at 391. Tektronix prayed for compensation that would have been, in the Court of Claims' view, excessive rather than "the just compensation payable under the Fifth Amendment." Id.

Tektronix, however, antedates both Panduit and Rite-Hite, both of which clarified the law pertaining to lost profits in private damages cases. See 35 U.S.C. § 284 ("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...."). While Rite-Hite clearly only discusses section 284, which contains language different from section 1498, Gargoyles is reasonable in speculating on the continued viability of the vague language of "strict proof" in Tektronix and the "clear and convincing" proof standard discussed therein, 552 F.2d at 348-49, 213 Ct. Cl. 257, 193 USPQ at 391, with respect to the less-litigated section 1498 actions. Certainly, we have held that in private (section 284) actions a patent owner "need not meet the impossible burden of negating every possibility that a purchaser might not have bought another product or might not have bought any comparable product at all," King Instrument Corp. v. Otari Corp., 767 F.2d 853, 864, 226 U.S.P.Q. (BNA) 402, 410 (Fed.Cir.1985), appeal after remand, 814 F.2d 1560, 2 USPQ2d 1201 (Fed.Cir.1987), and need only "provide proof to a reasonable probability that the sale would have been made but for the infringement." Paper Converting Mach. Co. v. Magna-Graphics Corp., 745 F.2d 11, 21, 223 U.S.P.Q. (BNA) 591, 598 (Fed.Cir.1984). Thus, the standard in private actions is that "the patentee must establish, by a preponderance of evidence, that but for the infringement he would have earned the profits he asserts were lost." Hebert v. Lisle Corp., 99 F.3d 1109, 1119, 40 USPQ2d 1611, 1618 (Fed.Cir.1996).

Further, nothing in Tektronix stands for the proposition that each and every element of Panduit must separately be met by clear and convincing evidence; indeed, on even a generous reading, Tektronix stands only for the proposition that the patentee in a government infringement suit must provide clear and convincing evidence that it would have made the sales "but for" the infringement, for in Tektronix the issue was whether the manufacturer would have received the contract had the supplier been prohibited from bidding. 552 F.2d at 346, 193 USPQ at 389. We also note that the announcement of the "strictest proof" standard in Tektronix is supported only by general references to the Fifth Amendment and the court's reticence to award lost profits against the government; it is not clear from the case that the issue of the standard of proof was properly joined by the parties or that Tektronix urged a lower standard on the court. Since both section 284 and section 1498 speak of "compensation," albeit "adequate" compensation in the former and "reasonable and entire" in the latter, lost profits should be recoverable in at least some infringement actions against the government, even though the Fifth Amendment is implicated. See Hughes Aircraft Co. v. United States, 86 F.3d 1566, 1571-72, 39 USPQ2d 1065, 1069 (Fed.Cir.1996), vacated on other grounds, --- U.S. ----, 117 S. Ct. 1466, 137 L. Ed. 2d 680 (1997) ("Because recovery is based on eminent domain, the proper measure is 'what the owner has lost, not what the taker has gained' ") (quoting Leesona, 599 F.2d at 969, 202 USPQ at 435); see also Welin Davit & Boat Corp. v. United States, 78 Ct. Cl. 772, 1934 WL 2150 (1934) (lost profits awarded). But see Hughes, 86 F.3d at 1571, 39 USPQ2d at 1069 (Although parties agreed royalty was appropriate, "[g]enerally, the preferred manner [of compensation] is to require the government to pay a reasonable royalty for its license."); but see also Motorola, Inc. v. United States, 729 F.2d 765, 768, 221 U.S.P.Q. (BNA) 297, 299 (Fed.Cir.1984) (concepts and phrases in Title 35 may not always "connote or denote the same meaning under section 1498"). Indeed, in Leesona, the Court of Claims noted that section 1498 is "not completely analogous to [the remedies] of Title 35," 599 F.2d at 968, 202 USPQ at 434 (for example an injunction is not available), but also stated "our suggestion in Tektronix [was] lost profits might be used in some circumstances to measure just compensation so long as a royalty was not awarded as well [i.e., 'double-counting']." Id. at 971, 202 USPQ at 436.

Decisions of the Court of Claims (but not the Claims Court) that are on point with the facts and legal issue of an appeal bind us unless they are overruled by an in banc decision of this court. Newell Cos., Inc. v. Kenney Mfg. Co., 864 F.2d 757, 765, 9 USPQ2d 1417, 1423 (Fed.Cir.1988) (in case of conflict, earlier decided case governs); South Corp. v. United States, 690 F.2d 1368, 1370 & n. 2, 215 U.S.P.Q. (BNA) 657, 657-58 & n. 2 (Fed.Cir.1982) (adopting precedent). Thus, Tektronix binds this court until overruled. We need not decide here, however, whether the Court of Claims ruling in Tektronix was correctly applied in the instant case or is still good law, because we conclude, as discussed infra, that even under the "reasonable probability" test discussed in Rite-Hite, Gargoyles would not be entitled to recover lost profits. We do not discern clear error in the fact-findings on the Panduit factors made by the Claims Court, and based on those findings we discern no legal error in the denial of lost profits.4 If the trial court had awarded lost profits, even under the lower standard, on this record, such findings would be clearly erroneous. We turn, then, to the errors asserted by Gargoyles in the findings of fact on the Panduit factors.

The government conceded the demand for the claimed invention, leaving in issue only: (i) Gargoyles' capacity to make and sell the eyewear, (ii) the absence of acceptable, non-infringing substitutes, and (iii) Gargoyles' ability to quantify the profits allegedly lost.

The government purchased 442,357 pairs of the infringing B/LPS eyewear over four years. The Claims Court found that Gargoyles could not have met that demand and that this fact was dispositive of the lost profits issue (although the court continued to analyze the other factors). The Claims Court credited Dr. Abram Hoffman, the government's accounting and finance expert, who testified that in order to meet the Army's demand during Operation Desert Storm, Gargoyles would have had to increase its output and coordinate additional suppliers while maintaining quality control. The trial court noted that no testimony had been offered regarding the capacity of these other suppliers or their willingness to meet demand. Dr. Hoffman also examined Gargoyles' finances and questioned its ability to increase production significantly.

Gargoyles claims that the testimony of Mr. Burns is more credible and was given insufficient weight by the trial court. Burns testified that Gargoyles always had an ample inventory of lenses and had excess molding capacity. Because Gargoyles had increased its capacity from 1987 to 1988, Burns concluded that meeting the new capacity would be "merely a matter of hiring a few more people and adding a few more tables in the assembly area" and there was more than one supplier for each part. Further, Gargoyles argues its testimony showed its business plan assumed great growth with financing accounted for (although the plan was not met) and that it could accommodate large orders. Last, Gargoyles argues Dr. Hoffman was not credible because he had not visited its factory and the trial court ignored a pre-litigation visit by Army personnel at which, Gargoyles contends, the Army was satisfied with production capacity.

Essentially, Gargoyles maintains that had the standard been a preponderance of the evidence, rather than "clear and convincing" evidence, the testimony of Mr. Burns might have proved sufficient over that of Dr. Hoffman. Certainly, if Tektronix dictates that the standard is "clear and convincing" evidence, then the matter merely becomes a choice between the credibility of two experts, and the Claims Court's conclusion that Gargoyles lacked capacity to produce the eyewear is supported by Dr. Hoffman's testimony. Anderson v. City of Bessemer City, 470 U.S. 564, 575, 105 S. Ct. 1504, 1512, 84 L. Ed. 2d 518 (1985) (decision to credit one witness when there is conflicting testimony can "virtually never be clear error"); First Interstate Bank of Billings v. United States, 61 F.3d 876, 881 (Fed.Cir.1995). The question then becomes whether the lower standard would avail Gargoyles and thus make a remand necessary. Even assuming this court had the authority to consider the rule in Tektronix distinguishable or otherwise not controlling because the question of the proper standard of proof was not in issue, we hold that on these facts Gargoyles has not proven capacity to produce the eyewear sufficient to receive lost profits even under the lower standard discussed in Rite-Hite. A holding by the trial court otherwise would be reversible error. The government notes that the commercial eyewear in Gargoyles' inventory would not have been acceptable for military use. Further, Dr. Hoffman testified that Gargoyles would expect lower profit margins on government sales. Gargoyles missed projections in its business plan to a significant extent, so the probative value of the projections to demonstrate capacity for further financing or production is low. Thus, we hold there is insufficient evidence of record to entitle Gargoyles to lost profits even under the "reasonable probability" standard.

Although the capacity issue standing alone was sufficient for the trial court to conclude lost profits should not be awarded, the trial court found SPECS, the competitor product with cylindrical lenses made by Uvex, to be an acceptable, non-infringing alternative. Gargoyles argued in the trial court that SPECS was not an acceptable alternative because the RFP from MRDC contained a requirement of "toric lenses with substantial wrap depth and zero power" and the Army's technical requirements during Desert Storm included a specification of toric wrap so as to meet its "Comfort and Attractiveness" requirement. Although the RFP from the Natick Center did not contain such a requirement, Gargoyles argues Col. LaPiana criticized the removal of the specification from the original draft.

Moreover, Gargoyles argues that SPECS were not available or "on the market" during the relevant accounting period because they were not "type classified" (as standard Army issue). See Zygo Corp. v. Wyko Corp., 79 F.3d 1563, 1571, 38 USPQ2d 1281, 1287 (Fed.Cir.1996) (product which is not available for purchase cannot be acceptable alternative). Finally, Gargoyles claims that because the Army ultimately purchased eyewear with toric lenses from AO, it is inferable that the Army had a clear preference for the patented item. See Kalman v. Berlyn, 914 F.2d 1473, 1484, 16 USPQ2d 1093, 1102 (Fed.Cir.1990) ("A product lacking the advantages of [the] patented [device] can hardly be termed a substitute 'acceptable' to the customer who wants those advantages.") (quoting Panduit, 575 F.2d at 1162, 197 USPQ at 734). Gargoyles concludes that, applying its proposed standard, there was a reasonable probability that some sales would have been directed to it if B/LPS was not available, and this fact in turn gives rise to a rebuttable presumption that all sales would have gone to Gargoyles.

The government provided evidence that the Army began developing SPECS in 1987, and by 1992 or 1993, Col. LaPiana was convinced that cylindrical lens eyewear could meet all military requirements. The government claims that SPECS had been recommended for type classification during Desert Storm, and Gargoyles' own commercial product was not type classified either, due to problems encountered during field testing and a lack of laser protection. Further, the Army sometimes allows for purchase of non-type classified items when special need is shown, as it did with Gargoyles' product for certain Army units.

We agree with the government's contention that causation of the loss of profits cannot be inferred here through the lack of acceptable substitutes; although Gargoyles characterizes the market as a two-competitor one, evidence established that: (i) Gargoyles' commercial product was not acceptable to the Army, (ii) SPECS had been in development for the Army from 1987 and the Army had accepted 500-600 units by the time of trial,5 and (iii) Col. LaPiana, who was only one of several analysts in the Army on eyewear, eventually also considered toricity to be unnecessary. Thus, it was not clear error for the trial court to infer that, had B/LPS not been available, the Army would have considered SPECS "acceptable to the Army despite their lack of the patented feature in question" and that clear and convincing evidence had not been presented that sales would have gone to Gargoyles. We think the same result would also obtain under the asserted lower "preponderance of the evidence" standard.

The Claims Court also found that Gargoyles did not properly quantify its lost profits. Mark Peterson, an accountant and Gargoyles' financial expert, estimated that the government saved a certain amount per unit by purchasing B/LPS and the government further saved a large sum in medical costs. That calculation was based on the price paid by Gargoyles' largest commercial buyer and the gross margin reflected in Gargoyles' business plan, adjusted upward to compensate for advertising benefits provided by the commercial buyer. Gargoyles argues that, "approached in different ways, and using numbers which are not in dispute ..., the evidence of record allows the calculation of lost profits that are close to, and in some cases higher than, the lost profits Gargoyles requested ... in its proposed findings...."

Dr. Hoffman examined Peterson's analysis and testified for the government that the analysis had several flaws, including (i) not allowing for "commercial endorsement value" of having the military use the glasses, (ii) using the lowest price paid by the Army for B/LPS, not the average, (iii) wrongly including the profits of another entity,6 (iv) failing to account for additional fixed costs of government contracting, and (v) improper methodology in calculating medical benefits. Thus, using Peterson's analysis as a starting point, Dr. Hoffman found a much smaller incremental profit to Gargoyles per unit.

Again, the trial court found Hoffman's analysis to be "credible and well reasoned." It also rejected Gargoyles' argument regarding the cost savings from reduction in injuries as speculative and not correctly performed, as it compared Gargoyles' protection to no protection, rather than the next alternative, SPECS. Again, the trial court selected Hoffman's analysis over Peterson's, and did not clearly err in concluding Gargoyles "failed to carry its burden of quantifying its lost profits by clear and convincing evidence." Given the trial court's decision to credit Hoffman, though, even under the lower standard of reasonable probability, Gargoyles still fails to meet its burden of showing its lost profits on this record. A remand is not necessary since the decision to credit Hoffman precludes a supportable finding that Gargoyles has shown lost profits even under the lower standard.

In sum, under Tektronix, Gargoyles would be required to show by clear and convincing evidence that but for B/LPS, it would have made and kept the profits it can quantify as lost. The trial court, in crediting the government's expert and other witnesses over Gargoyles', did not clearly err in determining Gargoyles failed to make that showing. Even if this panel were to distinguish Tektronix or deem it overruled by Rite-Hite, and accordingly apply a lower standard, Gargoyles' prayer for judgment on the fact findings or a remand to apply the lower standard would fail, as the fact findings made by the trial court would still compel us to affirm judgment in favor of the government. Given the irreversible credibility determinations made by the trial court, there is insufficient evidence of record to entitle Gargoyles to lost profits, even if the private infringement standard applies. Thus, we need not remand even if the lower standard applies, because a holding that Gargoyles would be entitled to lost profits would be clear error. Because Gargoyles cannot succeed under either standard of proof, we affirm the holding that Gargoyles is not entitled to lost profits damages.II. Reasonable Royalty

After correctly concluding that lost profits were not appropriate, the trial court, applying the fifteen factor test set out in Georgia-Pacific Corp. v. United States Plywood Corp., 318 F. Supp. 1116, 1120, 166 U.S.P.Q. (BNA) 235, 238 (S.D.N.Y.), modified and affirmed, 446 F.2d 295, 302, 170 U.S.P.Q. (BNA) 369, 374 (2d Cir.1971), assessed damages as a reasonable royalty. See slip op. at 14-21. The government asserts several factual errors in the Claims Court analysis of the Georgia-Pacific factors. The government contends the royalty rate "was double the highest rate supported by any credible evidence." We disagree.

First, the government asserts the Claims Court neglected what the government deems an offer to it of a license by Gargoyles. Central to this determination is the interpretation of a letter by Mr. Burns to the Army which states "under appropriate circumstances we would be amenable to the negotiation of the sale of a technical data package or a similar contractual vehicle, that would permit competitive procurement of our products." The government argues that Gargoyles' main business was in the commercial, fashion eyewear industry and it was consistent with the company's strategy to license the government business. The trial court, however, credited Burns' testimony that, as the eyewear was the "core and heart" of the company, it never had, nor would it have, been willing to license the patent. Further, as Gargoyles notes, the patent does not restrict its claims to fashion eyewear. In interpreting the letter, the trial court found that it was merely an offer to sell large orders to the government at lower rates. Since the trial court's interpretation of the letter is entitled to deference just as is its interpretation of the testimony of witnesses, see R.C.F.C. 52(a), we find the decision to credit Burns and interpret the letter as an offer to sell, but not license, was adequately supported and not clear error.

Next, the government challenges the finding that Gargoyles' success and profitability in the fashion market was not indicative of the value of the invention for protective eyewear. However, the Army's own documents indicate that attractiveness was a concern so the soldiers would wear the product off-duty. While perhaps in some cases government contractors can expect less profit when licensing to the government than in the private sector, the success in the private marketplace was a reasonable starting point for the trial court to analyze the hypothetical royalty negotiation.

Third, the government argues that inclusion of another entity's profits in the royalty calculation was clearly erroneous and in conflict with the prior holding that Gargoyles had impermissibly included that entity in the lost profits calculation. However, the trial court only considered that entity as one of several factors in the value that a reasonable licensor would put on the patent. Based on the nature of the relationship between the entity and Gargoyles, it would be reasonable for Gargoyles to put a high value on a license if it realized licensing would force the other entity to lose profits. Even if the government's argument that this is impermissibly compensating a non-party for loss of business opportunity is correct, the conclusion that Gargoyles placed a high value on the patent is still more than adequately supported by the success of the product and its centrality to the company's business.

The government also asserts that the trial court impermissibly compensated Gargoyles for actions outside the accounting period, arguing that before the accounting period AO had conducted research and development and delivered 2,000 pairs of B/LPS to the government, and arguing that the trial court's conclusion that AO would compete with Gargoyles in 2005 when the patent expires was speculative. The government also maintains Gargoyles should not receive a higher royalty because AO received the benefit of the government's challenge to validity. We note that together these asserted errors would not result in a significant change in the royalty rate, but in any event the Claims Court correctly considered the Army's actions in encouraging the creation of AO and that Gargoyles would have demanded a large royalty to compensate for foregoing the business advantage (in the form of vastly increased sales and the accompanying research benefits and production experience) which Gargoyles would have realized if it had licensed the Army to produce the eyewear. While we agree with the government that (i) the right to exclude does not inure until the patent issues, (ii) the government has not waived sovereign immunity for collateral acts like inducement and contributory infringement, see Decca, Ltd. v. United States, 225 Ct. Cl. 326, 640 F.2d 1156, 1167, 209 U.S.P.Q. (BNA) 52, 61 (1980), and (iii) consequential damages are not recoverable against the government under section 1498, see id., here it is reasonable to assume that Gargoyles would not have abandoned the opportunity to produce the eyewear without compensation. Also, allowing Gargoyles to recover, partly, for the unsuccessful challenge to the validity of the patent does not, as the government argues, impair the government's right to challenge a patent. AO clearly benefited here from the government's decision to pursue the invalidity of Gargoyles' patent.

As for the ultimate determination of the 10 percent royalty, we review to determine whether there has been clear error, overturning the royalty rate only if we "are left with the definite and firm conviction that a mistake has been made." Hughes, 86 F.3d at 1571, 39 USPQ2d at 1069, citing SmithKline Diagnostics, 926 F.2d at 1164, 17 USPQ2d at 1925, and quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948). The trial court here selected a rate within the bounds of the two parties' estimates proffered by their experts. Contrary to the government's assertion, the trial court did not rely on rates used in prior cases; the opinion is clear that the rates in other cases were merely a check against the court's reasoning on Gargoyles' unwillingness to license, the value of the patent to Gargoyles, and other factors. While obviously the facts of each case (and royalty determination) differ, there is nothing objectionable about a court examining similar cases to confirm that its decision was a reasonable one. In sum, it was not clear error to award the royalty rate of 10 percent on the compensation base described in the trial court opinion. The district court adequately explained the reasoning for its selection of that royalty figure, and given the hypothetical nature of the determination of what number reasonable negotiators would select, see State Indus. v. Mor-Flo Indus., 883 F.2d 1573, 1580, 12 USPQ2d 1026, 1031 (Fed.Cir.1989), we will not disturb the Claims Court's selection.

The government also argues the patent is invalid for failure to disclose the best mode of practicing the invention. It takes issue with the trial court's conclusion that although there were "obvious mistakes" in the patent (to which Gargoyles admits) and a failure to disclose certain design information, nonetheless it was not established by clear and convincing evidence that the inventor, Burns, concealed a known better method of carrying out the invention or that one skilled in the art could not practice the invention without undue experimentation. See Chemcast Corp. v. Arco Indus. Corp., 913 F.2d 923, 927-28, 16 USPQ2d 1033, 1036-37 (Fed.Cir.1990).

With respect to the knowledge requirement, the trial court found the evidence was conflicting as to whether Burns knew the details of optical design and appreciated the best mode. The government argues that Burns, as owner of Gargoyles, worked directly with Gentex Corporation to design and manufacture the eyewear product and selected a design with certain refractive power and lens thickness variance. Although, as noted by the government, the finding that "there is conflicting evidence regarding whether Burns knew the details of the optical design ... and appreciated it" is not further explained in the Claims Court opinion, it is clear from the record that Burns testified that Gentex would not give him technical information and that Burns gave the finished eyewear to an optometrist to measure (with imprecise instruments) and obtain the key technical data. Although the record indicates Burns might have had access to the technical information at the time, the trial court's finding that at the time of filing Burns knew of no better mode is adequately supported and we are not convinced that clear error was committed, especially given the high burden on the government to show concealment. Although the government claims Burns knew of the best mode but simply did not know the details of the best mode, citing United States Gypsum Co. v. National Gypsum Co., 74 F.3d 1209, 1214, 37 USPQ2d 1388, 1392-93 (Fed.Cir.1996) (fact the ingredient was trade secret did not excuse failure to disclose), there was no duty for Burns to disclose details of which he was not aware. See Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1050, 34 USPQ2d 1565, 1569 (Fed.Cir.), cert. denied, --- U.S. ----, 116 S. Ct. 516, 133 L. Ed. 2d 424 (1995) ("the best mode inquiry focuses on the inventor's state of mind at the time he filed his application"). Although it is incontrovertible that the commercial embodiment was in fact different than what was in Burns' mind at the time he filed his application for the patent, the specificity of the information known to the inventor and "facts within the possession of the inventor" were the ones incorporated into the specification without change or omission. Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524, 1535, 3 USPQ2d 1737, 1745 (Fed.Cir.1987). Further, the trial court was correct, given the facts as it found them, in not imputing knowledge of Gentex to the inventor, Burns. See Glaxo, 52 F.3d at 1050, 34 USPQ2d at 1569; Texas Instruments Inc. v. U.S. Int'l Trade Comm'n, 871 F.2d 1054, 1061, 10 USPQ2d 1257, 1262 (Fed.Cir.1989). Simply put, Burns did not actively conceal anything that he himself knew. See Chemcast Corp. v. Arco Indus. Corp., 913 F.2d 923, 926, 16 USPQ2d 1033, 1035 (Fed.Cir.1990) (Sole purpose of best mode " 'is to restrain inventors from applying for patents while at the same time concealing from the public preferred embodiments of their inventions which they have in fact conceived.' ") (quoting In re Gay, 50 C.C.P.A. 725, 309 F.2d 769, 772, 135 U.S.P.Q. (BNA) 311, 315 (1962)). Accordingly, the trial court's conclusion that there was no proof by clear and convincing evidence of knowing concealment was not clear error, and we therefore need not reach the issue of whether one skilled in the art would have known to correct admitted defects in the specification. See Spectra-Physics, Inc., 827 F.2d at 1535, 3 USPQ2d at 1745. Since the government did not prove that Burns knew of and withheld the best mode,7 there can be no violation of section 112 p 1.

The trial court's factual findings are all supported by evidence of record and are not clearly erroneous. Regarding lost profits, Gargoyles' argument that Tektronix is distinguishable or should be overruled (which may only be effected by an in banc court) is unavailing; even if Tektronix does not apply, the record indicates an insufficient showing of lost profits under either the clear and convincing or preponderance of the evidence (reasonable probability) standard. On the cross appeal, concerning both royalty and best mode we are not left with a definite and firm conviction that the trial court erred in assessing the relevant facts. Accordingly, the judgment of the Claims Court is in all respects affirmed.

AFFIRMED.

The parties shall bear their own costs.