STEVENS, Circuit Judge.
There are 60 separate claims in plaintiff’s twice reissued Brainard patent.1 Plaintiff originally contended that 29 of the claims were infringed by defendant’s machines.2 After discovery, plaintiff [581]*581dramatically changed its position. Its charges of infringement were limited to the contention that defendant’s vertical spindle machine (Model 15V) infringed claim 15 and that its horizontal spindle machine (Model 25H) infringed claims 19 and 20. Moreover, before the trial commenced, plaintiff advised its licensees, and presumably the trade generally, that it would not thereafter enforce claims 36 through 60 of its Brainard patent.3
This dramatic narrowing of the traditional issues of validity and infringement was the consequence of revelations during pretrial discovery. That discovery revealed that immediately after signing the notice of allowance of the original Brainard patent, Primary Examiner Beall retired from the Patent Office and became a paid consultant for plaintiff; his services related to the two reissues of Brainard, as well as the acquisition and reissue of the important Morgan patent.4
The district court, 306 F.Supp. 189, held that Brainard’s reissue claims 15, 19 and 20 were valid and infringed. These claims were substantially identical with claims in the original Brainard patent.5 The principal issue raised by this appeal is whether plaintiff’s employment of Beall in connection with the reissues of Brainard renders the claims in suit unenforceable. Defendant also challenges the district court’s findings of validity and infringement and its holding that the asserted invalidity of certain broad reissue claims was irrelevant; and contends that it is entitled to affirmative relief under the antitrust laws and an award of attorneys’ fees.
I.
Several facets of the record have a bearing on our appraisal of the legal issues. Accordingly, it is appropriate to describe the industry, plaintiff’s machines, the Morgan patent, the accused machines, the prosecution of the original Brainard patent and Examiner Beall’s connection therewith, the nature of Beall’s services to plaintiff after his retirement from the Patent Office, and, finally, the corrective steps taken by plaintiff to cure any possible misuse. We shall then explain why we have decided to reverse.
A. The Industry.
The parties manufacture automatic machine tools which can perform a variety of machining operations, such as milling, boring, drilling, threading and reaming. For each operation, the exact location, diameter and depth of the hole to be machined in the workpiece are predetermined and defined by a numerical code which can be translated into perforations on tape or cards. The operations are thus controlled “numerically.” The concept of numerical control of machine tool operations became commercially important in the early 1950s.
The parties use numerical control, not merely to direct specific operations automatically, but also to control a sequence of operations and the selection of the specific tool needed for each operation. Their machines employ automatic tool changers. For the purpose of this case, they have stipulated that the relevant market may be described as “numerically controlled multi-function machine tools with automatic tool changers.”
Since a sequence of operations normally requires the use of a variety of tools, [582]*582the time required to effect a tool change has a direct and significant impact on costs; the efficiency of the tool changer is, therefore, of vital importance. Plaintiff’s automatic tool changer was not the first to be developed, but apparently it is the best.
Plaintiff’s “Milwaukee-Matic” machines were introduced in 1958 and first sold in 1959; sales increased each year through 1967 when volume exceeded $36,-000,000. During those years, plaintiff’s sales of Milwaukee-Matic exceeded $150,-000,000 compared to competitors’ sales of between $75,000,000 and $100,000,000. Moreover, in 1968 plaintiff received approximately $573,000 of royalty income from its Brainard patent. It was a commercial success.
Plaintiff’s machines were sold in different models and sizes at prices ranging from about $90,000 to over $350,000 each. Total sales of over 925 units through March, 1968, indicate an average per unit price of about $180,000. Plaintiff did not sell its automatic tool changer except as a part of the MilwaukeeMatic machine, and prior to 1969 was unwilling to discuss a separate license for the tool changer.
Plaintiff’s customers include practically all types of manufacturers in the metal cutting industry, including the United States and foreign governments. The precision multiple operating capabilities of plaintiff’s machines have contributed to the design and production of new products, especially in the aerospace industry.
Plaintiff’s principal competitor in the numerically controlled automatic tool changing field is one of its licensees; at least two competitors are engaged in patent litigation with plaintiff; the total number of manufacturers is apparently about half a dozen. In 1967 plaintiff’s sales represented over one-third of the total market.
B. Plaintiff’s Tool Changer.
Plaintiff’s machine uses a horizontal spindle. The tool changer may be simply described. It uses a rotating arm with two hands, one of which approaches the spindle to withdraw the tool which has just completed an operation, while the other is approaching the storage magazine to grasp the next tool to be used. The arm then rotates, simultaneously delivering the new tool to the spindle and returning the old tool to storage. As the arm returns to a parked position, the storage magazine rotates to place a third tool in position for selection after the second has done its job. The selection is controlled by coding on the tools themselves; this “tool coding” (as contrasted with “position coding”) makes it possible to place an old tool in the storage position from which the new tool has just been withdrawn without impairing the machine’s ability to find that tool (or any other tool) when it is next needed.
This oversimplified description ignores a number of important features of the tool change mechanism,6 but it serves to identify two concepts which are of principal importance in this litigation: (1) the use of tool coding to identify the tool to be selected from the storage magazine; and (2) the use of a two-handed arm to effect a simultaneous interchange between the magazine and the operating station.
Neither of these concepts is expressly identified in the three Brainard claims which the district court held valid and infringed. Defendant contends, however, that these claims are too broad to be patentable unless limited to machines [583]*583using tool coding, particularly since plaintiff relied on this concept to distinguish the prior art — principally the Morgan patent — during the Patent Office prosecution. These two concepts clearly differentiate plaintiff’s machines from the Morgan machine.
C. Morgan.
Mark Morgan was a project engineer for International Business Machines Corporation. Apparently he was primarily interested in applying the concept of numerical control to an automatic tool changer. However, it is the mechanical features of the Morgan machine which are relevant here.
The Morgan machine uses a vertical spindle which may be moved down to grasp a tool positioned beneath it. Tools are stored in a vertical position in notches around the periphery of a disc-shaped magazine which has two horizontal paths of movement. First, it rotates until the position in which the desired tool is located is next to the spindle; and second, the entire magazine then moves laterally to place that tool directly beneath the spindle, where it is grasped by a wrenching mechanism which holds it until the spindle has descended and secured the tool. The magazine then withdraws by again moving laterally. After an operation with a tool has been completed, the magazines moves in to retrieve that tool, withdraws, rotates to a new position, moves back in to deliver the second tool to the wrenching mechanism, and then withdraws again to await the completion of the second operation.
The Morgan tool change sequence is obviously more time-consuming than plaintiff’s simultaneous two-handed interchange which, because the tools themselves are codéd, permits the old tool to be stored in the position from which the next tool has been withdrawn.
The claims of the Morgan patent are much broader than the particular machine described in its specifications. The claims may be read to cover plaintiff’s machine as well as defendant’s, and one of them might encompass any numerically controlled multi-purpose machine tool with an automatic tool changer.7 Defendant was a licensee under the Morgan patent. Morgan was the principal obstacle to the allowance of plaintiff’s Brainard patent. Plaintiff never accepted a license under Morgan, but seriously considered not only a possible license, but also a challenge to the validity of Morgan and ultimately acquired it.
D. Defendant’s Machines.
Defendant participated in the development of numerical control for machine tools in the early 1950s. Its two accused machines contain improvements over Morgan, but are apparently inferior to the Milwaukee-Matic. One of them is itself patented.8 The vertical spindle machine (15V) has been loosely described as “two Morgans with one spindle”; the horizontal spindle machine (25H) more nearly resembles plaintiff’s because it uses a rotating two-handed arm to withdraw the old tool from the spindle and to replace it with the next tool. Only a brief description of each machine is necessary.
1. Model 15V.
Instead of the one disc-shaped magazine used by Morgan, defendant’s model 15V uses two. Each rotates in a horizontal plane to place the next tool in line with the spindle. Instead of moving an entire magazine under the spindle as in Morgan, however, the next tool is [584]*584pushed into the operating position by a one-handed arm, and then grasped by the descending spindle. If only one magazine were used, the operation would be much like Morgan.
By using two magazines alternately, however, a simultaneous exchange of tools may be effected. While the first tool is being withdrawn into its original position in the first magazine, the second tool can be pushed from the second magazine into the operating position. The first magazine can then be rotated to place the third tool in position for insertion while the second magazine remains stationary to await the withdrawal of the second tool after its use.
Plaintiff contends that the simultaneous interchange by means of two one-handed arms, which is effected when the two magazines are used alternately, infringes Brainard’s claim 15.9 The less efficient Morgan-like use of successive tools from the same magazine is conceded to be noninfringing.
2. Model 25H.
As in the Milwaukee-Matic, the tool change in defendant’s model 25H is effected by rotation of a two-handed arm. The old tool, however, is not replaced directly in the storage magazine, but instead is delivered to a shuttle which then carries it laterally to the magazine; the magazine is then rotated to the position from which the third tool is selected by the shuttle for redelivery to the transfer arm and subsequent exchange with the second tool.
Model 25H does not use tool coding, and the use of the shuttle would appear to be more cumbersome and time-consuming than the simple pivot which positions Milwaukee-Matic’s next tool for transfer. Moreover, as in Morgan (and unlike the Milwaukee-Matic), the spindle in model 25H moves in the direction of its own axis to grasp and release tools. Nevertheless, the resemblance between the machines supports the district court’s finding of infringement of claims 19 and 20.10
[585]*585E. Prosecution of the Original Brainard Patent No. 3,052,011.
The application for the original Brainard patent was filed in the Patent Office on June 27,1958, and assigned to Division 13, for which T. Emmert Beall was Primary Examiner and R. H. Eanes was Assistant Examiner. On January 27, 1959, the Patent Office rejected 40 of the 49 claims originally asserted.
On September 1,1959, the Morgan patent issued. About a month later, plaintiff filed the first amendment to its application, modifying many of the original claims and requesting withdrawal of the rejections.11 Before any office action suance of the Morgan patent and ex-ment, plaintiff called attention to the is-was taken on the first amendment, plaintiff filed a second amendment cancelling 14 claims and adding 12 new claims.12 In the remarks supporting this amendplained that the new claims had been drawn to distinguish Morgan.13
On December 16, 17 and 18, 1959, pursuant to an invitation extended by plaintiff in November, and with the approval of the Patent Office, Examiners Beall and Eanes visited plaintiff’s plant in Milwaukee to observe a demonstration of the Milwaukee-Matic machine. On one evening during this visit they attended a birthday party for plaintiff’s president at his home, and the next evening visited in the home of one of plaintiff’s patent attorneys.14
The first action in the Brainard file wrapper after December, 1959, is an office action dated May 23, 1960, which was signed by Beall but apparently dictated by Eanes. It allowed 12 claims and rejected the remainder, many as being unpatentable in view of Morgan. One of the claims in suit, subsequently allowed as No. 15, was simply “rejected as met by Morgan.” Plaintiff’s response ber 21, 1960, and explained at length why to this office action was filed on NovemMorgan was distinguishable.15
There is then another interval from November, 1960, until June, 1961, in which no activity is indicated by the file wrapper. However, the possibility of initiating an interference with the Morgan patent was discussed with Examiner Eanes in an interview in Washington during the week of January 25, 1961. The interview is summarized in a memorandum in plaintiff’s Patent Department file which indicates that several of the Brainard claims had been rejected on the basis of Morgan.16 The memorandum makes no reference to Examiner Beall, and Beall did not recall any discussion of a possible interference proceeding. He did, however, recall discussing the possibility of “swearing back of the Morgan patent” with plaintiff’s counsel and Eanes.17 In general, Beall testified that [586]*586he gave almost no personal attention to the Brainard application.18 Eanes, however, testified that Beall personally participated in “several interviews” because plaintiff’s counsel, when they did not agree with Eanes, “ * * * continually took me to Mr. Beall for further interview.” 19
On June 1, 1961, plaintiff filed a supplement making various technical changes which apparently resulted from office interviews with Examiner Eanes. The next development of significance occurred on September 27, 28 and 29, 1961, when plaintiff’s patent counsel conducted a series of interviews with both Beall and Eanes. At these interviews they reviewed drafts of a new set of claims including those subsequently allowed as Nos. 19 and 20. Both Eanes and Beall suggested certain changes to make the claims allowable; both also expressed the opinion that claim 10 of the Morgan patent read on the Milwaukee-Matic machine. A second supplement was filed on October 16, 1961, and in an office action dictated by Eanes and signed by Beall on October 25, 1961, 24 claims, including claims 19 and 20 (at that time numbered 62 and 63), were allowed, and others, including claim 15 (at that time numbered 58), were cancelled or rejected. Thereafter another amendment was filed and considered, and additional claims, including No. 15, were allowed, and ultimately, on June 4, 1962, the formal Notice of Allowance was issued. It was signed by Beall.
The district court found no impropriety in the conduct of either Beall or Eanes prior to the issuance of the original Brainard patent.20
F. Beall’s Post-Retirement Service.
Beall retired from the Patent Office on June 22, 1962.21 The head of plaintiff’s Patent Department made a trip to Washington to attend his retirement party two or three days later. On that visit he explored the possibility of employing Beall as a consultant, but Beall then expressed no interest in working for plaintiff.
On July 12, 1962, plaintiff’s president met with Beall at his home and came to an understanding concerning his employment. Towards the end of July, Beall visited plaintiff in Milwaukee; on August 1 a written contract for the performance of part-time consulting services for $10,000 a year was executed. The contract was terminable by either party annually, but contemplated a term of five years. Beall actually continued to work for plaintiff until August of 1968 when the nature of his services came to the attention of plaintiff’s trial counsel.
Shortly after he was employed by plaintiff, Beall was given two assignments involving an analysis of the relationship between the Morgan and Brainard patents. Ultimately, with the assistance of Beall, both patents were reissued.
1. The Brainard Reissues.
Beall’s study of Brainard in relation to the prior art led him to the conclusion [587]*587that the invention involved a much broader concept than was described in the allowed claims. In general, the claims related to (1) the simultaneous transposition of the spindle tool and a magazine tool by means of a two-handed arm, and (2) the use of tool coding to avoid either the delay associated with the rotation of the magazine after a used tool has been returned to its original position or else the sophisticated programming that would be necessary if each position had to be indexed first for tool selection and then for tool reception. Beall determined, however, that the Milwaukee-Matic machine had actually pioneered the concept of using a transfer means for bodily moving the selected tool from the magazine to the spindle (since in Morgan the magazine itself moved to the spindle, .which then descended to grasp the tool). Thereafter Beall participated in the drafting of the broad claims included in plaintiff’s applications for reissuance of the Brainard patent.
The inventor, Brainard, described the broad concept in an interview with one of plaintiff’s patent attorneys on July 27, 1962: 22
“Hajewski: Who conceived the general concept of a single spindle machine tool with a mechanism for transferring a tool from a tool storage medium to the single spindle?
“Brainard: I conceived this general concept as a desirable objective. The development of a mechanism for achieving this broad concept was more or less a process of evolution in which a number of engineers contributed various ideas.” (DTX 418 p. 2.)
Generally speaking, the claims in the original patent covered the “mechanism” for achieving this broad concept, whereas the reissue claims were intended to cover the broad concept itself.
Beall worked on the reissues primarily in an advisory capacity, but he also discussed the applications with Eanes, his former subordinate, on more than one occasion. After the first Patent Office action on the reissue application, which was unfavorable, Beall forwarded a proposed amendment to his client. In his forwarding letter he stated:
“Incidentally, it would be helpful to me if my reports were temporarily removed from the file if and when the file is used at the Patent Office. Our friend might note the reports with great glee and attempt to inform a possible evasion of the two year rule. You probably removed them anyway. Further I do not see that any real evasion has been done and do not contemplate filing for register as a patent attorney anyway.” (DTX 106, p. 5; A. 123)
Beall did not apply for registration to practice before the Patent Office until February of 1965. On one occasion in December of 1964 Eanes objected to his presentation of an amendment to the second reissue application on the ground that he was not registered. Beall and Eanes lunched together on two or three occasions and discussed the reissue application. The first Brainard reissue was dated May 26, 1964; the second reissue was on March 2, 1965.
In addition to his work on the claims, Beall advised plaintiff on tactics and procedure, including the form and content of the oaths required in connection with the reissue applications. Each oath recited that the patent was partially inoperative because the applicants had failed to claim all that they were entitled to as a result of “inadvertence.” The asserted “inadvertence” was the basic point which Beall made in his opinion analyzing the original claims. It was explained in an oath executed by the applicants to overcome certain Patent Office objections in part, as follows:
“That the complete scope of the invention was not fully appreciated before the patent issued and it is now realized that the Morgan patent No. 2,901,927 is the closest tool change art [588]*588available insofar as it discloses the first structure in which a tool is completely removed from a tool storage medium and transferred to a spindle for the performance of work operations, but * * *
“That the present invention represents an advance in the art and includes among its features the broad concept of a tool transfer member operating to extract a tool from the tool storage magazine and transferring the extracted tool to the spindle as distinguished from said Morgan structure in which movement of the tool storage magazine for effecting the tool transfer is effected; * * (DTX 2 p. 121)
A memorandum prepared by the vice president of a competitor after his counsel had reviewed the claims added by the first Brainard reissue appraised their significance as follows:
“It was the consensus that the new claims are so broad that they encompass not only the SAC Automatic Tool Changing System, and not only every tool changing system now in use, but also every tool changing system which might be conceived in the future.” (DTX 549)
2. The Morgan Reissue.
The original application for the Morgan patent was filed on December 27, 1957, and the patent issued on September 1, 1959. A few months later, plaintiff’s patent counsel made a preliminary inquiry of I.B.M. about the possibility of taking a license under Morgan.23 Although the question whether the Milwaukee-Matic machine infringed the Morgan patent was the subject of study by both I.B.M. and plaintiff, no concrete action had been taken by either party before the original Brainard patent issued in September of 1962. They had a negotiating session in April of that year, however, in which the possibility of a license was discussed. In that meeting, plaintiff’s patent counsel took the position that only claim 10 of the Morgan patent was arguably infringed by the Milwaukee-Matic machine, and that the broad construction required to support a charge of infringement would render the claim invalid. I.B.M.’s position was that claim 10 was valid and infringed.
In an interview with plaintiff’s patent counsel during the prosecution of the Brainard application, Beall had commented on a possible conflict with claim 10 of Morgan. Shortly after Brainard issued and Beall had agreed to serve as a consultant, plaintiff requested him to make a thorough analysis of the possible infringement of Morgan by the Milwaukee-Matic machine. His conclusions confirmed the opinion already formed by plaintiff’s patent counsel that only claim 10 was infringed and that claim 10 was invalid.24 Promptly thereafter plaintiff reopened negotiations with I.B.M.
[589]*589In March of 1963, plaintiff consulted Beall about the advantages of purchasing the Morgan patent from I.B.M. Among the advantages which they considered was the fact that Morgan taught “the broad concept of tape control of a tool change mechanism wherein a tool is bodily transferred from a storage member to a rotary spindle.”25 Beall recognized that the concept had been disclosed but not claimed in Morgan and immediately suggested the possibility of obtaining a reissue of Morgan after its acquisition.
Plaintiff also recognized that the acquisition “of Morgan would put K&T in a dominating position relative to the tool change art.” To this comment Beall responded: “Especially true [if] Morgan was reissued.” They expressly noted that in a reissue, “Morgan claim 10 could be altered to make it valid.”
In an agreement effective as of April 1, 1963, plaintiff purchased the Morgan patent for $350,000, plus a share of future royalties. Before recording the assignment, plaintiff requested Beall to draft a set of reissue claims for Morgan, to prepare an opinion on the justification for reissue, and to outline the proper approach to be followed in filing the reissue application and processing it in the Patent Office.
With Beall’s guidance and assistance, the application was filed on January 17, 1964, and the patent was reissued on June 29, 1965.26 In the Morgan reissue, claim 7 was narrowed by adding qualifying language and eleven new claims were added. Except for insignificant changes in punctuation, no amendment to claim 10 was made.
Before the Patent Office, unlike the position taken in its own appraisal during the negotiations with I.B.M., plaintiff contended that claim 10 was valid.27 It hedged the possibility of a finding of invalidity, however, by the addition of six new claims, each of which incorporated the language of claim 10, plus additional qualifying language. The reissue was allowed on the theory that it narrowed the claims of the original patent to avoid prior art which had been overlooked inadvertently by I.B.M. The inadvertence was reasonable since I.B.M., the original owner of the patent, was a company that “manufactures electrical controls with very little, if any, experience in the machine tool field.” 28
G. Plaintiffs Corrective Action.
Without conceding any impropriety in the employment of Beall or any misuse of its patent, plaintiff took extensive remedial action summarized as follows in its brief:
1. In August, of 1968 it terminated Beall’s consulting contract when the nature of his services with respect to the reissue patents came to the attention of plaintiff’s trial counsel.
2. The services of Mr. Wutschel, who had headed plaintiff’s patent department for many years, were terminated shortly before trial, he having been made available for extensive deposition by defendant prior thereto.
[590]*5903. Effective in April 1969 plaintiff limited the allegations of infringement to claims 15, 19, and 20.
4. By directors’ action taken May 19, 1969, plaintiff relinquished all claims obtained through the reissue and the re-reissue of the Brainard patent (notwithstanding opinions of counsel that no advantage was obtained through the use of Beall’s services relating thereto) and licensees were so notified.
5. By letters plaintiff invited its existing licensees to discuss a separate license of a tool changer if they so desired.
6. A revised form of Brainard license agreement was adopted.
7. Plaintiff’s patent department was reorganized and thereafter reported to a newly employed attorney-at-law of substantial experience.
II.
Plaintiff denies that Beall’s conduct violated the conflict of interest provisions of the Criminal Code, 18 U.S.C. § 207, or the Rules of Practice of the Patent Office. Even if the policy underlying those provisions was offended by Beall, plaintiff contends that the claims acquired by reissue are not invalid; moreover, since it has virtually abandoned claims 36 through 60, claims 15, 19 and 20 should survive because the original patent was untainted.
For purposes of decision we accept the district court’s conclusion that the original Brainard patent was valid when it issued. Moreover, we accept the born, fides of plaintiff’s significant corrective action. Those remedial steps may not, even impliedly, be taken as an admission of wrongdoing; otherwise we might encourage the obdurate defense of dubious conduct which experienced patent counsel could cause to be corrected. Plaintiff’s curative measures, therefore, are prescinded from our consideration of the merits.
A.
The district court found that there had been no discussion between plaintiff and Beall about his possible employment until after he had already retired, and that his personal participation in oral interviews and his visit to plaintiff’s plant involved no violation of Patent Office policy. We accept those findings. Nevertheless, this background is not irrelevant to our appraisal of Beall’s post-retirement conduct.
First, it establishes beyond dispute that Beall participated personally and directly in the deliberations within the" Patent Office which preceded the issuance of the original Brainard patent. His signature on the Notice of Allowance was not a mere ministerial act. Second, his employment by plaintiff promptly after the patent issued created an appearance of impropriety that imposed a special obligation on both Beall and plaintiff to avoid conduct which might raise questions about the integrity of the Patent Office itself. For this sequence of events, even though wholly innocent, fits precisely the pattern that a corrupt draftsman would have designed.
Finally, the background reminds us to identify the interests which the Patent Examiner is entrusted to protect. He is not an agent of the inventor or the inventor’s assignee employed to assist in the procurement of a patent monopoly.29 Nor is he a dispenser of monopoly grants, merely arbitrating disputes between rival claimants to a share of the royal prerogative. See Graham v. John Deere Co., 383 U.S. 1, 5-6, 9, 86 S.Ct. 684, 15 L.Ed.2d 545; cf., Darcy v. Allein, 77 Eng.Rep. 1260 (K.B. 1602). For that reason, the significance at[591]*591tached by the district court to the fact that Beall and Eanes were entertained by defendant as well as plaintiff on their trip to Wisconsin was wholly misplaced. The Examiner does not merely decide who shall receive the monopoly privilege; he must also decide whether the claim shall be allowed at all. For in exchange for the inventor’s contribution to the public by the disclosure of the product of his genius, it is the public which pays his justly earned reward.30 It is the public interest which the Patent Office was created to protect,
B.
A few months after Beall’s retirement from Government service, Congress enacted a new conflict of interest statute. In material part it provides:
“(a) Whoever, having been an officer or employee of the executive branch of the United States Government, of any independent agency of the United States, * * * after his employment has ceased, knowingly acts as agent or attorney for anyone other than the United States in connection with any judicial or other proceeding, application, request for a ruling or other determination, * * * in which the United States is a party or has a direct and substantial interest and in which he participated personally and substantially as an officer or employee, through decision, approval, disapproval, recommendation, the rendering of advice, investigation, or otherwise, while so employed, * * *
“Shall be fined not more than $10,-000 or imprisoned for not more than two years, or both: * * * ” 31
Plaintiff does not dispute that, by reason of his former status as a Primary Examiner for the Patent Office, Beall was a person to whom the statute applied. It is also clear that the potential impact on the public of the reissue applications placed them in the category of matters in which the United States has a direct and substantial interest.32 Plaintiff argues, however, (1) that Beall did not act as an “agent or attorney” within the meaning of the statute, and (2) that the reissue applications were not matters in which Beall had participated while employed by the Government.
The first argument, in substance, is that Beall was not counsel of record for the plaintiff, but merely provided it with off-the-record aid and assistance.33 Un[592]*592deniably, however, as a full review of the facts demonstrates, in rendering legal advice and performing drafting and other services for plaintiff, Beall was acting in his professional capacity as an attorney even if we ignore his personal contacts with Eanes. The matters on which he worked were not mere technicalities; they were of great importance. The fact that his participation was not fully and frankly disclosed does not avoid the statutory proscription.
Plaintiff’s second argument is also asserted as a basis for avoiding the operation of Patent Office Rule 341(g). That rule disqualifies a former examiner from the registration which is required as a precondition to practice in the Patent Office “unless he undertakes (1) not to prosecute or aid in any manner in the prosecution of any application pending in any examining division in which he served during his .period of service therein; and (2) not to prepare or prosecute nor to assist in any manner in the preparation or prosecution of any application of another filed within two years after the date, he left such division, and assigned to such division, without the specific authorization of the Commissioner.” 34
Since Beall did not apply for registration until more than two years after his retirement, he apparently considered the language of the rule inapplicable. Quite clearly, however, the rule reflects an underlying policy that former examiners should not perform the kinds of service that would make them ineligible for registration. Moreover, it is equally clear that the second clause made it inappropriate for Beall “to assist in any manner” in the preparation or prosecution of the first Brainard reissue application assigned to Division 13 without the specific authorization of the Commissioner of Patents.35
The language of the first clause of the rule raises the same question as the conflict of interest statute. Should the applications for reissue be regarded as new matter or as something that had been pending before Beall retired ? A fair appraisal of plaintiff’s argument requires consideration of the nature of a reissue.
C.
To place the subject of reissues in proper perspective, it should be noted that less than 1% of the patents which are granted are subsequently reissued.36 The patent privilege is “an exception to the general rule against monopolies and to the right to access to a free and open market.” Precision Instrument Manufacturing Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806, 816, [593]*59365 S.Ct. 993, 998, 89 L.Ed. 1381. Reissues, then, are a narrow category within an exception to the general policy of free competition.
The purpose of the patent grant is to reward, and thereby to encourage, invention by exchanging a limited exclusive right for disclosure of the inventor’s novel concept. Early patent cases identified the contractual character of the “quid pro quo” exchanged between the inventor and the public. See, e. g., Pennock v. Dialogue, 27 U.S. (2 Pet.) 1, 14, 7 L.Ed. 327. At the time the patent issues, the inventor’s disclosure to the public is complete. Having dedicated his idea to the public domain, he has no further consideration to deliver in exchange for additional monopoly protection to be obtained by a reissue of his patent. Nevertheless, the contract analogy is the source of the recognition of reissues.
Before the Patent Office examining procedures had been initiated, the issuance of a patent was a ministerial act, subject to judicial review should infringement litigation arise. Since a technical defect in the specifications invalidated the patent, the inventor was permitted to retain the benefit of his bargain with the public by cancelling his patent and obtaining a reissue “ ‘when the defect in the specification arose from inadvertence or mistake, and without any fraud or misconduct on the part of the patentee.’ ” Grant v. Raymond, 31 U.S. (6 Pet.) 218, 240, 8 L.Ed. 376. In such a case, the reissue was necessary to the “faithful execution of the solemn promise made by the United States.” 37
The reissue privilege which was originally recognized to correct defects in the nature of a scrivener’s error remains in the law today. The statutory privilege has been broadened but still contains the historic limitation that no new matter may be introduced in a reissue application. Maxant Button & Supply Co. v. Sears Roebuck & Co., 388 F.2d 912, 916-917 (7th Cir. 1968). Any reissue, since it is predicated solely on the invention disclosed in the original patent, necessarily relates back to the original application for its justification. The statute provides, in part:
“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 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. * * * ” 35 U.S.C. § 251.
Broadening reissues contain risks of abuse not involved in the mere correction of an error in the specification or in a narrowing reissue. As the Supreme Court has pointed out, in cases involving broadening reissues the circumstances of the original application may have special relevance to the deliberations within the Patent Office.38
[594]*594We, therefore, conclude that Beall’s participation as an attorney in the prosecution of the Brainard reissues involved the same conflict of interest as if he had transferred his allegiance while the original application was still pending. His conduct was indefensible.
D.
In our opinion claims 36 through 60 of Brainard Re.-Re. 25,737 are invalid.39 The misconduct evidenced by this record is no less serious than concealment of the true identity of the author of a technical article, cf., Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 64 S.Ct. 997, 88 L.Ed. 1250, or the concealment of a prior use by the brother of the inventor, Keystone Driller Co. v. General Excavator Co., 290 U.S. 240, 54 S.Ct. 146, 78 L.Ed. 293. As those cases hold, therefore, the facts call for nothing less than a complete denial of enforceability of any of the claims acquired by reissue.40
In Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172, 86 S.Ct. 347, 15 L.Ed.2d 247, the Court characterized a patent procured by fraud as not merely unenforceable but invalid. The Court recognized a distinction between so-called “technical fraud,” which might result from “an honest mistake” (id. at 177, 86 S.Ct. 347) or from a “judicial finding of ‘obviousness’ ” (id. at 179, 86 S.Ct. 347, Harlan, J., concurring), and fraud involving the willful misrepresentation of facts to the Patent Office. Walker’s allegations fit the latter category. To support its conclusion that Walker’s action was not barred by the rule that only the United States may sue to cancel or annul a patent, the Court quoted a passage from the Precision Instrument opinion which is relevant here:
“The far-reaching social and economic consequences of a patent, therefore, give the public a paramount interest in seeing that patent monopolies spring from backgrounds free from fraud or other inequitable conduct and that such monopolies are kept within their legitimate scope.” 382 U.S. at 177, 86 S.Ct. at 350 (quoting from 324 U.S. at 816, 65 S.Ct. 993, 89 L.Ed. 1381).
We think the phrase “fraud or other inequitable conduct” encompasses the [595]*595kind of conduct disclosed by this record. A violation of the proscriptions against conflict of interest to obtain a broadening reissue involves the same threat to the public interest as actual fraud practiced on the Patent Office. Cf., United States v. Mississippi Valley Generating Co., 364 U.S. 520, 81 S.Ct. 294, 5 L.Ed.2d 268.
A review of the projects on which the former Primary Examiner provided technical and tactical guidance reveals gross abuse of the statutory privilege of reissue to cure defects resulting from “inadvertence.” It is quite clear that the inventor who conceived the desirable objective of using an independent means to transfer a tool from the storage magazine to the operating spindle shortly after he began work on the MilwaukeeMatic project in 1956 did not consider that broad concept patentable. It was Beall, several years later, who discovered the inadvertence in failing to file claims encompassing that broad concept. Beall also had recognized the overlap between claim 10 of the Morgan patent and the Milwaukee-Matie machine while he was still Primary Examiner of Division 13; that overlap was a factor subsequently considered in connection with the acquisition and reissuance of Morgan to enlarge plaintiff’s “dominating position relative to the tool change art.” Whether the reissue of Morgan could truly be defended as narrowing 41 when the validity of claim 10 was reasserted and reenforced by the addition of related claims after plaintiff and Beall had privately concluded that claim 10 was invalid, is a question we need not decide. It is, however, manifest that the use of Morgan was ancillary to plaintiff’s purpose to obtain a monopoly on the broad concept recognized by Brainard in 1956, but not protected by the patent issued in 1962.
Though the facts, the time intervals, and the relevant statutory provisions differ, we believe the Supreme Court’s comments in Miller v. Brass Co. are strikingly apt:
“But there is another grave objection to the validity of the reissued patent in this case. It is manifest on the face of the patent, when compared with the original, that the suggestion of inadvertence and mistake in the specifications was a mere pretence; or if not a pretence, the mistake was so obvious as to be instantly discernible on opening the letters-patent, and the right to have it corrected was abandoned and lost by unreasonable delay. The only mistake suggested is, that the claim was not as broad as it might have been. This mistake, if it was a mistake, was apparent upon the first inspection of the patent, and if any correction was desired, it should have been applied for immediately.
“These afterthoughts, developed by the subsequent course of improvements, and intended, by an expansion of claims, to sweep into one net all the appliances necessary to monopolize a profitable manufacture, are obnoxious to grave animadversion. The pretence in this case that there was an inadvertence and oversight which had escaped the notice of the patentee for fifteen years is too bald for human credence.” 104 U.S. 350, 351-352, 26 L.Ed. 783.
Having concluded that claims 36 through 60 are invalid, we must consider the effect of that conclusion on claims 15, 19 and 20 in view of the provisions of the 1952 amendments to the Patent Code dealing with reissues and disclaimer.
E.
Unlike an original patent in which all claims have effect from the same date of issue, the claims in a reissued patent may have different effective dates. As 35 U.S.C. § 252 now provides, [596]*596to the extent that the claims of a reissue patent are identical with the original patent, they “constitute a continuation thereof and have effect continuously from the date of the original patent.” Thus, claims 15, 19 and 20 were effective as of September 4, 1962, whereas claims 35 through 60 were effective as of May 26, 1964, or March 2, 1965. This provision supports plaintiff’s argument that the invalidity of the added claims should not impair the enforceability of the original claims.
We believe, however, that the answer to the precise question presented here is found in other provisions of the Code. Section 252 was enacted to deal with the problem of intervening rights of third parties who may have relied on the limited scope of the original patent before the reissue was allowed. See Frederico, “Commentary on the new Patent Act,” preceding 35 U.S.C.A. at pp. 45-47. This provision did not eliminate the requirement that the original patent be surrendered for. cancellation at the time of a reissue. Indeed, except for the provision relating to intervening rights, § 252 itself expressly provides that “every reissued patent shall have the same effect and operation in law, on the trial of actions for causes thereafter arising, as if the same had been originally granted in such amended form, * *
If an original patent contains some claims which are valid, and some which are invalid, it is now settled that the patentee may sue on the valid claims provided that a specific statutory condition is met. Formerly, unless promptly disclaimed, an invalid claim would vitiate the entire patent. Ensten v. Simon, Ascher & Co., 282 U.S. 445, 51 S.Ct. 207, 75 L.Ed. 453. The rule was based on the potential deterrence to competitors who might not realize that an allowed claim was in fact invalid. Id. at 453, 51 S.Ct. 207. This deterrence would be eliminated by filing a formal disclaimer which would then come to the attention of any member of the public investigating the status of the patent. See Marconi Wireless Telegraph Co. v. United States, 320 U.S. 1, 58, 63 S.Ct. 1393, 87 L.Ed. 1731.
Now, even without a formal disclaimer, the invalidity of one claim does not thereby render the remaining claims invalid. Like the right to sue on a valid claim in a patent containing invalid claims, however, the statute expressly identifies a condition which must be met. That condition is that the claim be invalid “without any deceptive intention.”
The disclaimer section now provides, in part:
“Whenever, without any deceptive intention, a claim of a patent is invalid the remaining claims shall not thereby be rendered invalid.” 35 U.S.C. § 253.
And the authorization for an action for infringement of a patent containing an invalid claim provides, in part:
“Whenever, without deceptive intention, a claim of a patent is invalid, an action may be maintained for the infringement of a claim of the patent which may be valid.” 35 U.S.C. § 288.
Plaintiff cannot meet the statutory condition. As we have already noted, in Walker the Supreme Court recognized a distinction between so-called technical fraud, based for example on a finding of obviousness, and actual fraud or other inequitable conduct. We believe the statutory requirement “without deceptive intention” identifies the latter category and that Beall’s conduct falls within it. Claims 36 through 60 are invalid but not “without any deceptive intention.” Fairly read, § 288 prohibits the maintenance of any action on a patent which includes claims which are invalid by reason of deceptive intention. This reading is supported by the inclusion of the same condition in § 253 42
We have no doubt that plaintiff is chargeable with responsibility for [597]*597Beall’s conduct. The trial court so found. His employment was negotiated by the president of the company. His services related to matters of vital importance to the company. He worked closely with the head of the company’s patent department and cautioned him about the desirability of avoiding reference to his participation in the matter. The fact that he accepted relatively nominal compensation for his part-time labors does not exclude plaintiff from full responsibility for the work it employed and directed Beall to perform. Accordingly, plaintiff is chargeable with “deceptive intention” within the meaning of the Patent Code. As we read the statute, the infringement action must therefore fail.
We have considered and reject plaintiff’s contention that its significant curative steps, including virtual abandonment of the claims acquired by reissue, have cleansed its hands sufficiently to protect the viability of claims 15, 19 and 20. Since plaintiff is charged with responsibility for, and knowledge of, Beall’s conduct, its commencement of this litigation, asserting infringement of the reissue claims as well as those still in dispute, was itself an abuse of the patent privilege. Not until after Beall’s conduct was revealed during pretrial discovery did plaintiff renounce its efforts to exploit the broadening reissues. More fundamentally, since the conduct is within the “deceptive intention” category, the curative steps cannot overcome the statutory objection to maintenance of the action.
III.
Two related questions remain. Defendant contends that it should recover its attorneys’ fees pursuant to 35 U.S.C. § 285, and treble damages pursuant to § 4 of the Clayton Act, 15 U.S.C. § 15.
We agree that this is an exceptional case and that the recovery authorized by § 285 should be allowed.43 The factors which we consider of controlling importance are not the protracted and complex character of the litigation but rather the nature of plaintiff’s wrongdoing and its potential impact on the public. See Monolith Portland Midwest Co. v. Kaiser Aluminum & Chemical Corp., 407 F.2d 288, 294 (9th Cir. 1969); cf., Scott Paper Co. v. Fort Howard Paper Co., 432 F.2d 1198, 1204-1205 (7th Cir. 1970), cert. denied, 401 U.S. 913, 91 S.Ct. 882, 27 L.Ed.2d 812. It is of paramount importance that neither the integrity of the Patent Office, nor public confidence in that integrity, be impaired. Moreover, the public has a comparable interest in identifying and forestalling the kind of abuse of the narrow statutory reissue privilege which is disclosed by this record. Instead of the proper correction of an inadvertence, necessary to .ensure that the inventor or his assignee will receive a fair reward for his disclosure, the Brainard and Morgan reissues were strategic maneuvers in a co-ordinated campaign to aggrandize control over an entire basic industry. A patent is not a beachhead from which, once secured, a patentee may scout adjacent territory for possible conquest by means of acquisition and reissue. By casting a revelatory light on the potential impact of broadening reissues, defendant, in the proper pursuit of its own private interests, has performed a significant public service.
The wrongful acquisition of monopoly power is prohibited by § 2 of the Sherman Act. The parties have stipulated that the relevant market is multifunction machine tools with automatic tool changers. Since plaintiff controls [598]*598only about one-third of that market, it has not achieved monopoly status. See United States v. Aluminum Co. of America, 148 F.2d 416, 424 (2nd Cir. 1945). Plaintiff’s actual position in the industry is not, however, the sole test of violation of § 2. By condemning attempts to monopolize, that section directs itself against dangerous probabilities as well as the completed result.44
A reliable or accurate measure of the actual likelihood that plaintiff would have achieved monopoly status if Beall’s wrongful conduct had never been disclosed is, of course, not possible. However, we do not understand the “dangerous probability” test to involve an evaluation of the actual likelihood that an attempt would have succeeded if not frustrated by an intervening event. Rather, it requires an appraisal of the alleged offender’s ability to achieve the forbidden result,45 his intent,46 and the nature of his overt actions.47 In an antitrust context we must consider the firm’s capacity to commit the offense, the scope of its objective, and the character of its conduct. The ultimate concern is the firm’s actual or threatened impact on competition in the relevant market.48
Plaintiff’s size, representing one-third of the industry, was “impressive.” See United States v. Columbia Steel Co., 384 U.S. 495, 533, 68 S.Ct. 1107, 92 L.Ed. 1533. At least one significant competitor was a licensee under the Brainard patent. Plaintiff’s patent position was fortified by its ownership of Morgan. Its sale and license arrangements were highly profitable. Clearly plaintiff was a firm with the capacity to make a serious attempt to acquire monopoly status.
Plaintiff cannot deny that it intended to acquire the power to exclude competition from the portion of the market described in the broad claims of the twice reissued Brainard patent. A competitor described those claims as broad enough to cover “ * * * not only every tool changing system now in use, but also every tool changing system that might be conceived in the future.” We need not accept that description to recognize that exclusive control of the broad concepts encompassed by these claims, when added to the firm’s other attainments (including its acquisition of Morgan and the products of its own ingenuity), would endanger the competitive process. Fulfillment of its objectives would have significantly enhanced its market position.49
[599]*599We have concluded that plaintiff’s employment of Beall was conduct comparable to that condemned by the Supreme Court in Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172, 86 S.Ct. 347, 15 L.Ed.2d 247. Because of its wrongful character and its potential for harm to the public, that conduct is appropriately characterized as “predatory.”
Thus, the uncontradicted evidence establishes plaintiff’s capacity to commit the offense, its intent to acquire the power to exclude competition from a significant share of the market, and its use of predatory means to accomplish its purpose. In Walker the Court held that the enforcement of a fraudulently procured patent would give rise to a § 2 violation if the exclusionary power of the illegal patent claim in the relevant market for the product involved exceeded accepted antitrust criteria.50 Since we find the same objection to the procurement of a patent by violation of conflict of interest principles as by fraud, under our appraisal of the exclusionary power of the Brainard reissue claims, that holding persuades us that plaintiff has violated § 2 of the Sherman Act.
This conclusion is consistent with other evidence in the record upon which defendant places great stress. Thus, the provisions in plaintiff’s licenses prohibiting separate manufacture or sale of the patented automatic tool changer and requiring licensees to pay royalties on an entire machine tool even though the patented unit represented only about 10% of its cost are probative of an intent to exclude competition and to enhance prices. Cf., American Photocopy Equipment Co. v. Rovico, Inc., 359 F.2d 745 (7th Cir. 1956). Similarly, although the acquisition of one of two dominating patents by the owner of the other has not been held to constitute a violation of the antitrust laws, plaintiff’s acquisition of Morgan may be taken as evidence of an intent to restrain competition. Cf., United States v. Singer Manufacturing Co., 374 U.S. 174, 83 S.Ct. 1773, 10 L.Ed.2d 823. We do not, however, rely on that evidence because it is subject to different interpretations which would require the kind of evidentiary analysis that should be made by the district court in the first instance. The basic factors on which we do rely do not involve the resolution of any disputed questions of fact. We have not disturbed any of the findings made by the district court and we have no doubt that if he had considered the reasoning of Walker applicable to plaintiff’s employment of Beall, as we have held, he would also have found a violation of the Sherman Act.
Even though defendant has proved a violation of the Sherman Act, to recover damages he must also demonstrate that plaintiff’s conduct “injured [him] in his business or property.” 15 U.S.C. § 15. We are not at all sure that defendant has met this additional burden.
Defendant contends that plaintiff unlawfully acquired a dominant position in the market and that a sales advantage, once obtained, tends to “pyramid” to the inevitable disadvantage of competitors. There is force to this contention, but certain qualifications must be identified.
We have accepted the district court’s conclusion that the original Brainard patent was valid when it issued on September 4, 1962. By then, presumably by virtue of the superior quality of its machine, plaintiff had already obtained a position of market leadership. No adverse impact on defendant appears to have resulted from the acquisition of Morgan in 1963.51 Not until after the [600]*600date of the first reissue of Brainard on May 26, 1964, can defendant properly suggest that the in terrorem effect of the invalid claims obtained by reissue had any impact on its own research and development. Indeed, it was not until May of 1966 that plaintiff notified defendant of its position that the Brainard patent was being infringed. The sales pyramid may, therefore, have been constructed on a lawful foundation.
Defendant has also argued that plaintiff’s dominant position, coupled with its restrictive licensing practices, has enabled it to enhance the price level. If this be so, defendant, as a competing manufacturer of machines, would not have been injured.52 Defendant asserts, however, that it was also a customer of plaintiff. We are not sure how significant its purchases were, when they occurred, or even whether they were Milwaukee-Matic machines.
Finally, defendant’s president testified that certain potential customers, particularly in Europe and Japan, declined to make purchases from it because of concern about possible infringement of the Brainard patent, including the claims acquired by reissue. We are unable to appraise the significance of this testimony as a matter of first impression.
In the district court the parties agreed to postpone any determination as to the amount of damages until after final decision on the merits. The trial did encompass the question whether defendant suffered any injury by reason of an antitrust violation. Because it found no violation of the antitrust laws, the district court had no occasion to make any findings on the fact of damage issue. On this issue the burden of proof is upon defendant; as indicated, we are not at all sure that it has sustained its burden. On the other hand, we cannot say as a matter of law that it suffered no injury whatsoever. The issue is one which the district court is better able to resolve than we are. See Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 123-125, 89 S.Ct. 1562, 23 L.Ed.2d 129. Whether it is necessary or appropriate to hear additional evidence on this factual issue we leave to the district court’s sound discretion.
The judgment is reversed and the case is remanded to the district court with directions to enter judgment in favor of the defendant on the complaint, to make an appropriate award of attorneys’ fees, and to conduct such additional proceedings on the counterclaim as it deems appropriate in the light of this opinion.