Digital Equipment Corporation v. Sidney A. Diamond, Etc.

LEVIN H. CAMPBELL, Circuit Judge. The district court proceeding from which this appeal is taken was instituted by Digital Equipment Corporation (DEC) with the object of overturning a decision of the U.S. Patent and Trademark Office (PTO) striking from the files, on grounds of “fraud,” DEC’s application for reissue of a patent. Computer Operations, Inc. (COI), DEC’s adversary in ongoing patent infringement litigation, intervened as a defendant. On cross-motions for summary judgment, the district court found for the PTO and COI, and DEC now appeals. Central to the appeal is a question of first impression: whether the PTO is authorized to strike reissue applications on the basis of the “patent law fraud” criteria reflected in the Commissioner’s Rule 56, or whether Rule 56 is supplanted as regards reissue applications by a narrower “deceptive intention” standard found in 35 U.S.C. § 251. I. The dispute revolves around U.S. Patent No. 3,387,293, which teaches a control format relating to a magnetic-tape data storage and retrieval system for small computers. Issued in 1968 to Thomas C. Stoekebrand, a DEC employee, the patent was assigned by him to DEC. Stoekebrand joined DEC in the fall of 1962 and shortly thereafter became involved with a DEC project to develop and commercialize an existing system known as “LINC tape.” Stoekebrand had participated in the development of the LINC tape system as an employee of MIT Lincoln Laboratories, Bedford, Massachusetts. During the course of the DEC development program, Stoekebrand improved upon the LINC tape format, eventually producing a system that could read and write on tape moving in either direction; it is this “bidirectional capability” which, according to DEC, provided the basis for Stockebrand’s patentable claims. 1 DEC has marketed the *704 Stockebrand system under the names “DEC tape” and “Micro tape.” In March 1963, DEC prepared an advertisement for the system modelled on LINC tape which it planned to market — the “Micro Tape Control Type 550,” which could be used in conjunction with up to eight “Micro Tape Transports Type 555.” In May 1963, after DEC’s development program had progressed further, DEC prepared a new advertising brochure to promote a “Type 550 Control” with bidirectional capability, to be used with up to four “Type 555 Dual Tape Transports.” This brochure described in some detail elements of the system disclosed in the Stockebrand patent. On May 15, 1963, DEC entered into an agreement to lease computer equipment, including two Micro Tape Controls and six Micro Tape Transports (to be used together as a system), to Kie Data Corp. However, when this lease commenced in July 1963, these units were still in process of being developed. As a result, DEC loaned Kie Data several conventional “Potter units” to be used in lieu of the DEC tape system. 2 Between June 1963 and August 1963, DEC also accepted three other orders for DEC tape devices; 3 two of the orders called for delivery on August 15, 1963 (DEC maintains this was the date set for “earliest possible” delivery) and the third set a delivery date of November 1,1963. The delivery dates were not met, however. All three orders were listed as “overdue” in DEC memoranda of January 25 and 28,1964, and the first of the three deliveries appears to have occurred on February 14, 1964. *705 During the summer and fall of 1963, Stockebrand worked to perfect the DEC tape system. Sometime in the fall of 1963, a DEC tape system was delivered to Kie Data; the exact date of delivery is not known, but the PTO found on the basis of the evidence before it (including entries in Stockebrand’s notebook) that the system was delivered prior to November 1963. According to Stockebrand, the system delivered was his “breadboard model,” the “prototype” on which he had been running tests. The system suffered various breakdowns and disappointments in performance, and Stockebrand spent much of his time from November 1963 through January 1964 working on the system at Kie Data’s plant. As early as February 1964, Stockebrand conferred by telephone with Hugo Liepmann, DEC’s patent counsel, concerning the DEC tape invention. He did not, however, at this time or later inform Liepmann of the delivery of the system to Kie Data or of the other orders for DEC tape systems accepted in the summer of 1963. Liepmann and Stockebrand first met on November 5, 1964, four days before the application for patent was filed, at which time they reviewed the application and discussed, inter alia, the preexisting LINC tape system. The application was then dated November 6, 1964 and was filed on November 9, 1964. 4 The application did not refer to LINC tape by name (although it contained what DEC characterizes as a description of prior art, which discloses certain features of the LINC tape system), nor did it mention the Kie Data delivery or other early commercial activity relating to DEC tape. (Although at least two other DEC employees who assisted in preparation of the patent application were aware that a system had been delivered to Kie Data, this fact was never disclosed to Liepmann or to his successor patent counsel, O’Donnell). As part of the application, Stockebrand signed an oath stating, inter alia: I do not know and do not believe that this invention was ever known or used before my invention or discovery thereof, or ... in public use or on sale in the United States more than one year prior to this application.... Stockebrand restated the above in a “supplemental declaration” filed prior to issuance of the patent. On June 4, 1968, the PTO issued the patent to Stockebrand, and the patent became the property of the assignee, DEC. In 1973, DEC learned that COI was marketing a tape system with an option that could read and write DEC tape. Following an exchange of correspondence between DEC and COI on the subject of possible infringement, on July 2,1974 COI instituted an action against DEC, asserting antitrust claims and seeking a declaratory judgment of the invalidity and non-infringement of the Stockebrand patent (hereinafter, “infringement action”). COI raised issues relating to possible prior art and “on sale” bars to the patentability of DEC tape. See 35 U.S.C. §§ 102 & 103. To bolster its position in the infringement action, on February 3, 1975 DEC filed an application to reissue the patent pursuant to 35 U.S.C. §251; it sought both to amend certain claims that “might be subject to a construction which covers more than the applicant is claiming as his invention," and to disclose to the PTO for the first time information re *706 lating to the Kie Data delivery and the LINC tape system, said to have been “discovered” in the course of investigating COI’s charges in the infringement action. Having been invited by DEC to participate in the reissue proceedings, COI petitioned the PTO to “strike” the reissue application on grounds of “fraud,” pursuant to PTO Rule 56 (37 C.F.R. § 1.56). After further correspondence with DEC and COI, and following receipt of DEC’S responses to two “Requirements for Information,” on November 3,1976 the PTO issued an “Order to Show Cause” why COI’s petition should not be granted. The Order set forth 13 “grounds” establishing a “prima facie case of fraud”: seven grounds related to the undisclosed commercial activity and to Stockebrand’s declaration that the invention had not been “on sale” in the United States more than one year prior to the date of application (j. e., prior to November 9, 1963, the “critical date”); three grounds related to the undisclosed advertising materials prepared in the spring of 1963, which were said to be “ ‘prima facie’ available as prior art against the application”; two grounds related to the inadequate disclosure of LINC tape, the “closest prior art of which [Stockebrand and DEC] were aware”; and one ground concerned the “mischaracterization” of the circumstances leading to the “discovery” and subsequent disclosure of the above information. Following COI’s submission of comments on the Order to Show Cause, on March 2, 1977 DEC filed its response to the Order, which was accompanied by five affidavits, including Stockebrand’s. On April 4, 1978, the PTO issued a decision striking the reissue application under Rule 56, finding that nine of the thirteen grounds of fraud had been established by “clear and convincing evidence.” The decision was signed by the Assistant Commissioner for Patents. DEC then brought this suit in the district court, and, as already noted, the district court refused to set aside the PTO action. On appeal, DEC makes essentially two arguments: (1) The PTO proceeded in excess of its statutory authority in striking the reissue application on grounds of fraud; and (2) The finding that “clear and convincing” evidence of fraud had been presented was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,” 5 U.S.C. § 706(2)(A). We examine each argument in turn. II. The PTO’s Authority to Strike the Reissue Application Under Rule 56 Reissue of patents is expressly authorized in 35 U.S.C. § 251. This statute, which establishes a corrective mechanism for defective patents, creates one of the few avenues for PTO reexamination of an already issued patent. Section 251 provides, in relevant 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 . . . reissue the patent for the invention disclosed in the original patent . .. for the unexpired part of the term of the original patent. The provisions of this title relating to applications for patent shall be applicable to applications for reissue of a patent. ... The PTO’s striking of DEC’s reissue application was not based on any purported failure to meet the express requirements of section 251, but rather was based on findings of “fraud” made pursuant to the PTO’s Rule 56 (37 C.F.R. § 1.56). Rule 56 had been revised as of March 1, 1977, and thus the version in effect when the PTO handed down its decision was different from the version in effect both when the Stockebrand patent was issued in 1968 and when DEC applied for reissue in 1975. Nevertheless, the PTO stated in its decision, *707 It is considered clear that the striking of this reissue application is necessary, whether considered under 37 C.F.R. § 1.56 as it existed before March 1, 1977, or as it exists at present. The pre-1977 version of Rule 56, in effect since 1949, 5 provided, in pertinent part, simply [A]ny application fraudulently filed or in connection with which any fraud is practiced on the Patent and Trademark Office, may be stricken from the files. As amended in 1977, the current version of the Rule begins by reciting that A duty of candor and good faith toward the Patent and Trademark Office rests on the inventor, on each attorney or agent who prepares or prosecutes the application and on every other individual who is substantively involved in the preparation or prosecution of the application and who is associated with the inventor, with the assignee or with anyone to whom there is an obligation to assign the application. All such individuals have a duty to disclose to the Office information they are aware of which is material to the examination of the application. Such information is material when there is a substantial likelihood that a reasonable examiner would consider it important in deciding whether to allow the application to issue as a patent.... Section (d) of the Rule goes on to state, An application shall be stricken from the files if it is established by clear and convincing evidence that any fraud was practiced or attempted on the Office in connection with it or that there was any violation of the duty of disclosure through bad faith or gross negligence. Notwithstanding the difference in language between the earlier and later rules, there appears to be no dispute that the words “fraudulently” and “fraud” in the pre-1977 version of Rule 56 should be read in light of the gloss placed on those words in patent infringement actions: viz., they encompass “grossly negligent” as well as intentional misconduct, see infra. In this respect, the later version merely makes explicit a standard already implicit in the earlier version. DEC challenges the PTO’s use of Rule 56 to strike its reissue application on two grounds. First, DEC argues that the reissue statute, 35 U.S.C. § 251, overrides Rule 56, and requires the PTO to reissue a patent unless there is a showing of “deceptive intention.” “Deceptive intention,” DEC insists, is a far more restrictive term than “gross negligence” or even “fraud” as understood in patent law. Second, DEC maintains that for the PTO to “strike” a reissue application, as opposed to “rejecting” the reissue claims as would be required were it to find “deceptive intention” within the meaning of section 251, frustrates the statutorily provided mechanism for review of PTO decisions (including the applicant’s right to institute a civil action in the district court following an adverse decision by the Board of Appeals), which is triggered only by PTO rejections. 35 U.S.C. §§ 134, 141 & 145. This “short-circuiting” of the normal review procedures is said to be especially egregious in cases such as this one where a dispute involving the “state of mind” of various individuals has been resolved on the basis of a paper record without benefit of demeanor evidence, and is now subject only to limited judicial review under the Administrative Procedure Act. 5 U.S.C. §§ 702 & 706. A. Validity of Rule 56 as an Exercise of the Commissioner’s Rulemaking Authority Rule 56 was promulgated by the Commissioner pursuant to the authority *708 granted by 35 U.S.C. § 6, “to establish regulations, not inconsistent with law, for the conduct of proceedings” in the PTO. In 1970, the Court of Customs and Patent Appeals held, in reviewing a denial of a motion to strike an original patent application, that Rule 56 was a proper exercise of the Commissioner’s rulemaking authority, and that the Commissioner was empowered thereunder to strike patent applications for “fraud,” as broadly defined in the patent field. Norton v. Curtiss, 433 F.2d 779, 793, 57 CCPA 1384 (1970). We think there can be no serious doubt, certainly not at this late date, that Rule 56 constitutes in general a valid exercise of the Commissioner’s rulemaking powers under 35 U.S.C. § 6, and therefore has the force and effect of law except as it may conflict with other statutory or constitutional provisions. The chief question here is whether, as DEC insists, Rule 56 is “inconsistent with law” insofar as the Commissioner now seeks to apply it not only to original applications but to reissue applications. DEC argues that section 251 establishes a statutory standard which conflicts with and therefore necessarily displaces the broader standard of “fraud.” Before turning to this question, it is useful to consider first the genesis and purpose of the concept of “fraud on the Patent Office,” which Rule 56 embodies. Only through understanding the nature of this rather unique concept is it possible to form a complete judgment as to the relationship between Rule 56 and section 251. The term “fraud” does not appear in Title 35. 6 It is a term that has frequently been invoked by courts and litigants in patent cases, however, often with different shades of meaning in each context in which it is employed. Compare Walker Process Equipment, Inc. v. Food Machinery & Chemical Corp., 382 U.S. 172, 86 S.Ct. 347, 15 L.Ed.2d 247 (1965) (patents “procured by intentional fraud” may give rise to antitrust claims) with Monolith Portland Midwest Co. v. Kaiser Aluminum & Chemical Corp., 407 F.2d 288 (9th Cir. 1969) (award of attorneys’ fees may be entered against party who obtained patent by fraud, pursuant to 35 U.S.C. § 285, authorizing such awards in “exceptional cases”) with Beckman Instruments, Inc. v. Chemtronics, Inc., 428 F.2d 555 (5th Cir.), cert. denied, 400 U.S. 956, 91 S.Ct. 353, 27 L.Ed.2d 264 (1970) (patent invalid because procured by fraud). See generally Transitron Electronics Corp. v. Hughes Aircraft Co., 649 F.2d 871 (1st Cir. 1981). The concept of “fraud on the Patent Office,” as something distinct from common law fraud or deceit, derives primarily from judicial opinions in infringement actions, holding patents “unenforceable” on equitable grounds. A seminal case in this area was Precision Instrument Manufacturing Co. v. Automotive Maintenance Machinery Co., 324 U.S. 806, 65 S.Ct. 993, 89 L.Ed. 1381 (1945). The Court there refused to enforce a patent on behalf of a patentee which had failed to disclose its knowledge that another party to an interference proceeding involving the patent had committed perjury. Finding that, in withholding this information from the Patent Office, the patentee “ha[d] not displayed that standard of conduct requisite to the maintenance of [a] suit in equity,” id., at 819, 65 S.Ct. at 999, the Court disposed of the infringement action under the equitable doctrine of “unclean hands.” In reaching this decision, the Court emphasized that, A patent by its very nature is affected with a public interest. . . . [A] patent is an exception to the general rule against monopolies and to the right to access to a free and open market. 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 *709 and that such monopolies are kept within their legitimate scope. Id., at 816, 65 S.Ct. at 998. Thus, the patentee’s willful concealment of fraudulent conduct, committed in connection with a patent application, justified dismissal of the action for “want of equity.” Drawing on Precision Instrument, courts have expanded the concept of disabling “patent fraud” beyond the classical definition of fraud to include “a wide range of ‘inequitable’ conduct found to justify holding a patent unenforceable.” Norton v. Curtiss, 433 F.2d 779, 793, 57 CCPA 1384 (1970). The less stringent standard necessary to show “fraud” as a defense to an infringement action has been characterized primarily by a relaxation of the traditional requirement of “scienter.” It is not necessary to a demonstration of “fraud on the Patent Office,” as it would have been to a demonstration of common law, fraud, to show that a misrepresentation or omission was intentionally made with a particular deceptive purpose in mind. Compare W. Prosser, Handbook of the Law of Torts 700-04 (4th ed. 1971). Culpability, however, remains an element of “fraud,” even in the patent law context. See Carpet Seaming Tape Licensing Corp. v. Best Seam Inc., 616 F.2d 1133, 1138-40 (9th Cir. 1980). While this court “has not defined the exact standard of wrongful conduct that would disable an otherwise valid patent from enforcement,” we have noted that “courts considering allegations of fraud in procurement of a patent have held that some element of wrongful conduct — intentional misrepresentation or inequitable, reckless, or grossly negligent conduct — is a necessary predicate to rendering a patent unenforceable in the courts.” International Telephone & Telegraph Corp. v. Raychem Corp., 538 F.2d 453, 461 (1st Cir.), cert. denied, 429 U.S. 886, 97 S.Ct. 238, 50 L.Ed.2d 167 (1976). See also Carter-Wallace, Inc. v. Davis-Edwards Pharmacal Corp., 443 F.2d 867, 882 (2d Cir. 1971) (“In order for nondisclosure to constitute inequitable misconduct there must be something more than negligence— . . . there must be ‘some element of wrongfulness, wilfulness or bad faith that transgresses the basic concept of doing equity.’ ”). Perhaps the most often quoted description of the level of culpability necessary to support a finding of “fraud on the Patent Office” — a passage quoted with approval in the PTO decision in this case — is taken from the decision of the Court of Customs and Patent Appeals in Norton v. Curtiss, 433 F.2d 779, 795-96, 57 CCPA 1384 (1970): Good faith and subjective intent, while they are to be considered, should not necessarily be made controlling. Under ordinary circumstances, the fact of misrepresentation coupled with proof that the party making it had knowledge of its falsity is enough to warrant drawing the inference that there was a fraudulent intent. Where public policy demands a complete and accurate disclosure it may suffice to show nothing more than that the misrepresentations were made in an atmosphere of gross negligence as to their truth. See also Tegtmeyer, Fraud on the Patent and Trademark Office Under Rule 56, 58 J. Patent Off. Soc. 550, 561 (“there must be an element of bad faith or total recklessness to find fraud or inequitable conduct”). The expansion of the concept of disabling patent fraud has proceeded largely from the courts’ perception of the “relationship between applicants for patent and the Patent Office and their scrutiny of the conduct of applicants in light of that relationship.” Norton v. Curtiss, 433 F.2d at 793. In Precision Instrument, 324 U.S. at 818, 65 S.Ct. at 999, the Court had stated that: Those who have applications pending with the Patent Office or who are parties to Patent Office proceedings have an uncompromising duty to report to it all facts concerning possible fraud or inequitableness underlying the applications at issue. . . . Only in this way can that agency act to safeguard the public in the first instance against fraudulent patent monopolies. The court in Norton v. Curtiss, 433 F.2d at 793-94, elaborated on this idea: *710 [Applicants before the Patent Office are being held to a relationship of confidence and trust to that agency. The indicated expansion of the concept of “fraud” manifests an attempt by the courts to make this relationship meaningful.... The highest standards of honesty and candor on the part of applicants in presenting . . . facts to the office are . . . necessary elements in a working patent system. It is thus well established that a patent is rendered unenforceable by inequitable conduct — falling somewhat short of common law fraud — on the part of a patentee in his dealings before the Patent and Trademark Office. A logical corollary of this proposition, and the rationale on which the Rule 56 power to strike applications is based, is that “conduct which will prevent the enforcement of a patent after the patent issues should, if discovered earlier, prevent the issuance of the patent.” Norton v. Curtiss, 433 F.2d at 792; see also Ex Parte Mallard, 71 U.S.P.Q. 294 (Comm’r Pat. 1946). At least as applied to original applications for patent, Rule 56 clearly serves this purpose. It is the use of Rule 56 in reissue proceedings, however, that DEC attacks. The PTO’s reliance in such proceedings upon findings of “fraud” is said to run counter to the express terms of section 251, which provides that patents shall be reissued to correct certain errors made without “any deceptive intention.” This language, in DEC’s view, reflects a congressional definition of the quantum of misconduct necessary to bar reissuance of a patent (“deceptive intention”); and as it is greater than the quantum of inequitable conduct sufficient to render a patent unenforceable (“fraud”), the latter standard cannot be imposed by the Commissioner under his section 6 rulemaking power. But assuming, arguendo, that “deceptive intention” within the meaning of section 251 encompasses something significantly different from, and requires a greater showing than, “fraud” as used in Rule 56, 7 it does not follow that the PTO may not rely on the “lesser” showing of fraud as a further basis for refusal to reissue a patent. Section 251 sets forth statutory prerequisites to reissue, in addition to the general conditions relating to issuance of patents found elsewhere in title 35: 8 there must be an issued patent deemed “inoperative or invalid”; it must be so because of “error”; and the patentee and applicant for reissue must not be guilty of any “deceptive inten tion” in connection with the error. See In re Oda, 443 F.2d 1200, 1206, 58 CCPA 1353 (1971). If these requisites are not met, section 251 stands as a statutory bar to reissuance of claims, just as sections 102 and 103 might stand as a bar to the issuance of claims. As we have already noted, however, Rule 56 operates independently of the statutorily prescribed conditions for patentability and issuance of claims. The principle that the privileges of a patent monopoly should be denied to applicants guilty of fraudulent behavior before the Patent Office is an equitable one, developed by the Supreme Court and the lesser federal courts. It is scarcely unreasonable that the Commissioner, in exercise of his rule-making authority, should make provision for recognition of this equitable principle in the Patent and Trademark Office. Otherwise the PTO would be obliged to issue claims knowing that they were unenforceable as a matter of law. The question, then, is not simply whether the prohibition of fraud embodied in Rule 56 is a reasonable interpretation of section 251’s “deceptive intention” terminology, but rather whether, taken as an independent duty imposed on *711 applicants, it is “inconsistent with law.” We cannot conclude that this is the case where the courts have determined “fraud” to be a basis of unenforceability of a patent. DEC relies exclusively on the seemingly mandatory terminology of section 251 —that the Commissioner “shall reissue” patents to correct errors made without deceptive intention — to argue that, absent a showing of deceptive intention, the Commissioner is absolutely required to act favorably on a reissue application. However, it is clear that an applicant would not be entitled to the reissuance of claims barred under sections 102 or 103, for example, regardless of whether there existed within the claims “error” made “without any deceptive intention.” See In re Wadlinger, 496 F.2d 1200, 1202-03 (Cust. & Pat.App. 1974); In re Doyle, 482 F.2d 1385 (Cust. & Pat.App.1973); cf. Ex Parte Varga, 189 U.S.P.Q. 204 (P.T.O.Bd.App.1973). Likewise, section 251 does not mandate that the Commissioner reissue claims procured by “fraud,” even if those claims contain errors that might otherwise be correctable through the reissue process. See 37 C.F.R. § 1.176: “An original claim, if re-presented in the reissue application, is subject to reexamination, and the entire application will be examined in the same manner as original applications. . . . ” See also 3 Chisum, Patents § 15.03[5], at 15-56 (1980). Furthermore, the supposed conflict between section 251 and Rule 56 is, in this instance at least, illusory for another reason. Section 251 deals with “errors” which render a patent “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. ...” (Emphasis added.) When such errors are found to have been made without deceptive intention, the Commissioner “shall reissue the Patent.... ” However, the sorts of errors to which the PTO’s findings of fraud were addressed — DEC’s failure to disclose relevant prior art and allegedly disabling prior sales — are not the sorts of errors to which section 251 refers. Even had the PTO determined that these errors had been made without deceptive intention, DEC would not have been entitled, at least without proposing amendments to its claims, to have its patent reissued. On the contrary, the PTO would have been compelled to reject the reissue claims; even if the prior art and sales did not bar patentability of the invention, there still would have been no “errors” rendering the original claims “wholly or partly invalid” due to the applicant “claiming more or less than he had a right to claim in the patent.” The requirements of section 251 would simply not have been met. 9 See In re Clark, 522 F.2d 623, 625-26 (Cust. & Pat.App.1975). DEC’s contention that the PTO erred by not examining its inadequate disclosure under a “deceptive intention” standard, rather than a “fraud” standard, is thus somewhat of a red-herring. The more appropriate question is whether the PTO was entitled to consider the inadequate disclosure as a dispositive factor at all — whether it was entitled to *712 determine that DEC’S misconduct in this regard disabled DEC from obtaining reissue on other grounds. 10 As we have already indicated, this question must be answered affirmatively. If one assumes, for purposes of argument, but see Part III, infra, that the PTO’s findings of fraud were well founded, it follows that the Stockebrand patent is unenforceable or invalid. If this determination had been made by a court prior to the PTO’s consideration of DEC’S reissue application, there is little doubt that reissue of the patent would not have been allowed. See In re Clark, 522 F.2d at 627 (“Reissue is not available to rescue a patentee who had presented claims limited to avoid particular prior art and then had failed to disclose that prior art (the examiner not having acted upon it) after that failure to disclose has resulted in the invalidating of the claims.”). Indeed, even if the PTO had reissued the claims in such circumstances, it is likely that a court would have found them nonetheless unenforceable; this conclusion would surely follow from the prevailing view that fraud practiced in connection with particular claims forever destroys the right of the applicant to have those claims mature into an enforceable patent. See Norton v. Curtiss, 433 F.2d at 783; Kearney & Treaker Corp. v. Giddings & Lewis, Inc., 452 F.2d 579, 595-96 (7th Cir. 1971), cert. denied, 405 U.S. 1066, 92 S.Ct. 1500, 31 L.Ed.2d 796 (1972). This being the case, there seems to be no good reason to prohibit the PTO from conducting its own examination of alleged fraudulent conduct, as part of its reissue proceedings, when, fortuitously or otherwise, it must dispose of the reissue application prior to a judicial determination on the issue of fraud. 11 Indeed, it would be an exercise in futility to require the PTO to reissue claims “tainted” by fraud, only to have those claims held unenforceable by the courts. We conclude, therefore, that the PTO does not exceed its statutory authority when it refuses to reis *713 sue a patent on the ground that it was procured by fraud. 12 B. Rule 56 Procedure DEC contends further, however, that the PTO may not take such action by means of a “summary” procedure — the striking of an application — that cuts off the channels of review established in 35 U.S.C. §§ 134, 141 & 145. This alleged denial of DEC’s right to de novo judicial review is criticized in particular as having eliminated DEC’s opportunity to introduce demeanor evidence on the issue of “fraud.” As an initial matter, we are not persuaded that decisions of the Assistant Commissioner of Patents striking applications on grounds of fraud would not be appealable under the above provisions. It is true that the jurisdiction of the Board of Appeals, pursuant to section 134, is limited to PTO decisions “which relate, at least indirectly, to matters involving the rejection of claims.” In re Hengehold, 440 F.2d 1395, 1404, 58 CCPA 1099 (1971). 13 Here, DEC’s “claims” have not been “rejected” in a technical sense. (Furthermore, decisions made by the Board as an “agent” of the Commissioner are not “decisions” within the meaning of section 141. In re James, 432 F.2d 473, 57 CCPA 1371 (1970).) Nevertheless, in determining whether a particular PTO action is appealable to the Board (and hence reviewable either by the Court of Customs and Patent Appeals or by the district court in a civil action instituted pursuant to section 145), one must “look both to the language employed [by the agency in taking the action] and the effect thereof,” and consider “both form and substance.” In re Haas, 486 F.2d 1053,1055 (Oust. & Pat.App. 1973). In Haas, the court held that the Board had jurisdiction to review an examiner’s “withdrawal” of claims, which “amount[ed] to a rejection” because it precluded further consideration of the claims “not only in this application but prospectively in any subsequent application because of their content.” Id., at 1056. At least one commentator has pointed to Haas as support for the proposition that PTO decisions striking applications under Rule 56 are appealable. Gholz, Review of Decisions Striking Patent Applications for Fraud, 61 J. Patent Off. Soc. 52 (1979). (The PTO’s position is otherwise, see id., at 52). Indeed, the PTO’s failure to label its present action in terms of “rejection” may have little practical significance other than to purportedly insulate from the review mechanism mandated by Congress a decision that is highly factual and substantive in nature, and thus particularly appropriate for such review. One might well conclude, based on a reading of the statutory provisions cited above, that Congress intended actions such as this one — constituting a final adverse disposi *714 tion of an applicant’s asserted right to a patent — to be appealable in the manner provided. See also 35 U.S.C. § 7. Even assuming, however, that such decisions are not reviewable under section 134, the PTO’s administrative application of the fraud standard to reissue applications is not thereby rendered unsupportable. We have already noted that the Commissioner is empowered to promulgate regulations “not inconsistent with law” to govern PTO proceedings. 35 U.S.C. § 6. Regulations establishing PTO inquiries addressed to possible “fraudulent” conduct are not inconsistent with law simply because decisions made pursuant to them are not reviewable in the same manner as are PTO decisions relating to other matters. Whatever the desirability of more extensive review of the PTO’s “fraud” findings, the possible failure of Congress to have provided for such review cannot be taken to preclude the PTO from acting in the first instance. Moreover, as our disposition of the instant case demonstrates, the courts retain power, under general provisions relating to review of agency action, e. g., 5 U.S.C. §§ 702 & 706, to correct any serious errors committed by the PTO. As for the “summary” nature of the administrative proceedings presently afforded by the PTO — the absence of provision for live testimony or oral argument — we think that, while such procedures may be imperfectly tailored, our review would be limited to whether they comported with the requirements of due process. The district court engaged in a thorough treatment of this issue and concluded that DEC was not denied due process under the circumstances of this case. DEC has not disputed this finding on appeal, and we thus consider the matter foreclosed. 14 III. We next turn to whether the PTO’s findings that “fraud” was committed in connection with the application for the Stockebrand patent, and in connection with the reissue application, can be sustained. The PTO determined that nine grounds of fraud were established by clear and convincing evidence. The district court found that four of these grounds for decision were “particularly clear” and that each supported the PTO’s determination that fraud had been committed. 15 We are unable to find, however, that the PTO has articulated any basis sufficient to support its action. We fully recognize the limited scope of judicial review of agency action under 5 U.S.C. § 706(2)(A), which provides for setting aside agency actions, findings and conclusions only if they are “arbitrary, capricious, an abuse of discretion or otherwise not in accordance with law.” 16 Nevertheless, though our review is constricted, we must still determine whether the PTO’s decision “was based on a consideration of the relevant factors and whether there has been a clear error of judgment.” Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 824, 28 L.Ed.2d 136 (1971); see also FCC v. National Citizens Committee for Broadcasting, 436 U.S. 775, 802-03, 98 S.Ct. 2096, 2115-16, 56 L.Ed.2d 697 (1978). We are convinced that the PTO’s decision was not based on an examination of the relevant factors and that, particularly in light of the PTO’s burden, articulated in its decision, to establish fraud by “clear and convincing” evidence *715 (see also Norton v. Curtiss, 433 F.2d at 796 — 97), the finding of fraud on the grounds advanced by the PTO amounts to “clear error” and lacks a “rational basis.” We are thus compelled to set aside the PTO’s action. A. The deficiencies of the PTO’s analysis are similar with respect to the first eight grounds on which it relied. In general, the PTO concluded that fraud had been committed because DEC and Stockebrand had in their possession information that was not disclosed in connection with the patent application. In each instance, however, the PTO failed to find all the elements necessary to a determination of patent office fraud; its findings did not reflect significant (let alone “clear and convincing”) evidence of “culpable” behavior on the part of the applicant or assignee, nor did they establish that the withheld information was of requisite materiality. As we point out in Part II A, an intentional scheme of deception is not necessary to a showing of “fraud on the Patent Office.” Nevertheless, a not insubstantial degree of “wrongfulness” is required to be shown. Applicants for patents do not engage in fraudulent conduct when they make “honest mistakes in judgment” concerning the information material to consideration of their applications, nor does some slight deviation from the requisite standard of care suffice as evidence of fraud. In the PTO’s own discussion of the law governing its decision, it acknowledged that a finding of fraud would be warranted only when an applicant has made misrepresentations with “knowledge of [their] falsity” or “in an atmosphere of gross negligence as to their truth.” Clearly the PTO did not mean to apply in this case some new administratively fashioned standard penalizing innocent or merely negligent nondisclosure. (It could hardly have done so, since all the relevant events occurred during a period when the only relevant codified administrative standard governing the conduct of applicants before the PTO was former Rule 56’s cryptic proscription of “fraud.”) Yet in the discussion of grounds one through eight, the PTO points to little that would indicate the requisite culpability. A fraud finding also requires that the withheld information meet a rather high standard of materiality. As the Court of Customs and Patent Appeals said in Norton v. Curtiss, In patent cases, “materiality” has generally been interpreted to mean that if the Patent Office had been aware of the complete or true facts, the challenged claims would not have been allowed.... Findings of materiality should not be limited only to those situations where there can be no dispute that the true facts, or the complete facts, if they had been known, would most likely have prevented the allowance of the particular claims at issue or alternatively, would provide a basis for holding those claims invalid. ... It is our view that a proper interpretation of the “materiality” element of fraud in this context must include therein consideration of factors apart from the objective patentability of the claims at issue, particularly (where possible) the subjective consideration of the examiner and the applicant. Indications in the record that the claims at issue would not have been allowed but for the challenged misrepresentations must not be overlooked due to any certainty on the part of the reviewing tribunal that the claimed invention, viewed objectively, should have been patented. If it can be determined that the claims would not have been allowed but for the misrepresentation, then the facts were material regardless of their effect on the objective question of patentability. 433 F.2d at 794-95 (emphasis in original). The PTO decision cites a similar passage from Union Carbide Corp. v. Filtrol Corp., 170 U.S.P.Q. 482, 521 (C.D.Cal.1971), aff’d, 179 U.S.P.Q. 209 (9th Cir. 1973): [a] patent applicant’s duty of disclosure to the Patent Office extends to prior art or other facts known to him which. . ., but for the nondisclosure, would have prevented the patent from issuing or *716 would have restricted the scope of the claims. (Emphasis added.) See also Parker v. Motorola, Inc., 524 F.2d 518, 535 (5th Cir. 1975), cert. denied, 425 U.S. 975, 96 S.Ct. 2175, 48 L.Ed.2d 799 (1976); Schnadig Corp. v. Gaines Manufacturing Co., 494 F.2d 383, 392-93 (6th Cir. 1974); Langer v. Kaufman, 465 F.2d 915, 921-22, 59 CCPA 1261 (1972). This standard of materiality is not even mentioned, however, let alone found to have been met, in the PTO’s discussion of the various grounds. 17 This is not to say that a strict “but for” test of materiality must be applied in every instance, or that it would necessarily be inappropriate to test the materiality of misrepresentations or nondisclosures against a standard somewhat different from the formulation used in Norton. See True Temper Corp. v. CF&I Steel Corp., 601 F.2d 495, 504 (10th Cir. 1979); In re Multidistrict Litigation Involving Frost Patent, 540 F.2d 601, 604 n.9 (3d Cir. 1976); Timely Products Corp. v. Arron, 523 F.2d 288, 297-98 (2d Cir. 1975). Questions of “materiality” and “culpability” are often interrelated and intertwined, so that a lesser showing of the materiality of withheld information may suffice when an intentional scheme to defraud is established, whereas a greater showing of the materiality of withheld information would necessarily create an inference that its nondisclosure was “wrongful.” See United States v. Standard Electric Time Co., 155 F.Supp. 949, 952-53 (D.Mass.1957). To establish fraud, however, the nondisclosed information must be such as to have a likely effect on the scope of allowable claims or the issuance of the patent. It is not enough that the information be simply “relevant” in some general sense to the subject matter of the claimed invention, or even to the invention’s patent-ability. Nevertheless, the PTO, in its discussion of the first eight grounds of fraud in this case, consistently shied away from making any concrete findings concerning the significance the withheld information would have had to an examiner’s consideration and allowance of the Stockebrand claims. B. The PTO’s discussion in support of “Ground I” is illustrative of the manner in which it proceeded. Ground I states: The applicant, Thomas C. Stockebrand, signed the oath for application Serial No. 409,913 on November 6, 1964, when he knew, or reasonably should have known, that the invention was “prima facie” “on sale in the United States more than one year prior” to the date of the application. The Assistant Commissioner’s decision and earlier Order to Show Cause set forth with specificity the facts which the PTO found to establish this ground by clear and convincing evidence. In brief these facts are as follows: DEC leased a DEC tape system to Kie Data in May 1963, to be delivered in July 1963; DEC accepted orders for DEC tape systems from three other companies between May 1963 and August 1963, with deliveries scheduled prior to November 1963; DEC delivered a system to Kie Data before November 1963; prior to the Kie Data delivery Stockebrand reported that the system was “now running satisfactorily (reading while writing)”, and Stockebrand testified that prior to delivery the system could “read and write” but “broke” after *717 delivery; 18 Stockebrand was aware of, and participated in, the above orders and deliveries, was told in August 1963 that the prototype system had been “sold” and he was “late,” and was “very upset at the sales department” which had made these arrangements without informing him beforehand. We note at the outset our confusion over the PTO’s “prima facie on sale” terminology. The applicant, Stockebrand, is charged with “fraud” because he signed the application oath while knowing that the invention was “prima facie on sale in the United States more than one year before the application date.” The meaning of “prima facie on sale” is not, however, explained, nor can we readily understand the utility of such a concept in the context of determining whether fraud has been committed. The oath that Stockebrand signed recites that: I do not know and do not believe that this invention was ever ... on sale in the United States more than one year prior to this application.. .. Stockebrand did not promise the invention was not “prima facie on sale.” The crux of Ground I must be that Stockebrand withheld information — the facts recounted above — that was relevant to an examiner’s determination of whether or not the invention was “on sale.” But since the PTO stopped short of finding that the invention was in fact “on sale” more than one year prior to the application date, 19 an essential question becomes whether the undisclosed facts were nonetheless so “material” that failure to disclose them was either in itself fraudulent or rendered the application oath fraudulently misleading. That the undisclosed information would amount to a “prima facie showing” of an “on sale” bar avoids the question whether it was significant enough to meet the Norton v. Curtiss “but for” standard or some other adequate standard of materiality. It is true that at one point the Assistant Commissioner states that “[e]ven if it could be construed that the sales involved an inoperable invention that had not been tested or that the sales were only for experimental purposes, the *718 question would be so close as to impose an obligation to disclose the sales to the examiner for a determination of their effect on patentability.” To say that the matter is “so close” as to impose an obligation of disclosure, however, is little more than conclusory. What should have been undertaken was an explanation of how close a question it was, in terms of the likely effect that the undisclosed information would have had on Stockebrand’s application for a patent. 20 Even were we to hold the findings on materiality marginally sufficient, the PTO’s discussion of Ground I would still be deficient in that it fails to address adequately the question of Stockebrand’s culpability for the lack of disclosure. After reciting the facts relating to the various pre-November 1963 orders for DEC tape and the Kie Data delivery, the PTO goes on to stress Stockebrand’s involvement in and awareness of these activities. It concludes that “it is incredible that Stockebrand would not remember the sales under the circumstances set forth above.” The implication seems to be that this demonstrates that Stockebrand signed the application oath with “knowledge of its falsity.” However, the oath that Stockebrand signed states a legal, or at least complex factual, conclusion: that the invention he sought to patent was not “on sale” prior to November 9, 1963. An invention cannot be “on sale” until it is operable and “reduced to practice,” nor is there an “on sale” bar if the invention is “sold” and delivered primarily for “experimental purposes.” DeLong Corp. v. Raymond International, Inc., 622 F.2d 1135 (3d Cir. 1980); In re Theis, 610 F.2d 786 (Cust. & Pat.App.1979); Timely Products Corp. v. Arron, 523 F.2d 288 (2d Cir. 1975); In re Yarn Processing Patent Validity Litigation, 498 F.2d 271 (5th Cir.), cert. denied, 419 U.S. 1057, 95 S.Ct. 640, 42 L.Ed.2d 654 (1974); B. F. Sturtevant Co. v. Massachusetts Hair & Felt Co., 124 F.2d 95 (1st Cir. 1941). 21 DEC does not maintain, as the PTO decision seems to suggest, that when Stockebrand signed the oath he did “not remember” the Kie Data delivery had ever occurred, that he developed some sort of selective amnesia. 22 Its contention is *719 that the conclusion stated in the oath was true, regardless of the Kie Data delivery and the other orders recounted by the PTO, or at least that Stockebrand reasonably believed it to be true. The PTO’s decision is entirely unresponsive to this contention; if anything it tends to support DEC’s position by expressly leaving open the possibility that the undisclosed “sales” involved “an inoperable invention that had not been tested or that the sales were only for experimental purposes.” If such were the case, then it could hardly be said that Stockebrand acted wrongfully by stating that there was no “on sale” bar to his invention. We find fatal the PTO’s failure to determine expressly that Stockebrand acted in bad faith by signing the oath, or at least had sufficient reason to believe the oath was false so that he acted “in an atmosphere of gross negligence” by signing it. The deficiency is especially glaring in light of Stockebrand’s sworn affidavit in which he avers that he did not deliberately conceal information from the PTO and had no reason to believe at the time he signed the oath, and does not now believe, that the oath could not properly be made. 23 The PTO points to no evidence that would indicate that Stockebrand actually considered the withheld information to be significant to the patentability of his invention yet made a deliberate decision not to disclose it. Cf. Beckman Instruments, Inc. v. Chemtronics, Inc., 428 F.2d 555 (5th Cir.), cert. denied, 400 U.S. 956, 91 S.Ct. 353, 27 L.Ed.2d 264 (1970); Norton v. Curtiss, 433 F.2d at 796-97; see also note 31, infra. Indeed, there is little, if anything, in the PTO’s present decision necessarily inconsistent with the conclusion that Stockebrand made an “honest mistake” in deciding that, in light of his substantial work on the system between November 1963 and January 1964, the invention he sought to patent was not reduced to practice, and was not “on sale,” prior to the critical date. Cf. DeLong Corp. v. Raymond International, Inc., 622 F.2d 1135, 1143 (3d Cir. 1980) (“A reduction to practice ‘is not achieved until the inventor has sufficiently tested the prototype to prove its utility and to determine that no further refinements are necessary.’ ”). As noted by the court in DeLong Corp., 622 F.2d at 1146, “[t]he mere misstatement in the patent application that there had not been a prior sale of the invention for more than one year prior to the date of the application, particularly when the circumstances pertaining to the transaction required a sifting of the facts, will not support a finding of fraud. . . . ” At most, the PTO has established that Stockebrand was “negligent” in not exploring, with his patent attorney or others, the possibility that the early commercial activity constituted an “on sale” bar. Such negligence is not by itself equivalent to fraud. C. Grounds II, IV, and VII suffer, in large part,- from the same deficiencies as Ground I. Ground II charges that The assignee, Digital Equipment Corporation (DEC), permitted, or caused, the application to be filed on November 9, 1964, when the assignee knew, or reasonably should have known, that the invention was “prima facie” “on sale in the United States more than one year prior” to the date of the application. Again, the PTO does not attempt to assess the probable impact of the withheld information on the ultimate issue of patentability. DEC’s culpability is made to rest primarily on the fact that three of its officers or employees, in addition to Stockebrand, were aware both of the Kie Data delivery and of the contents of the patent application generally. 24 No reasons are advanced *720 why these employees could not reasonably have believed that the application oath was true regardless of the Kie Data delivery. (It is also apparent that the two employees found to have actually “read” the application did so in order to scrutinize that portion of the application describing the invention and setting forth Stockebrand’s claims; they had little reason to be concerned with the contents of Stockebrand’s oath concerning prior sales.) In any event, these employees were at most shown to have been negligent in their supervision of the application process and in inadequately weighing the possibility that the early commercial activity might stand as a bar to patentability. 25 Grounds IV and VII are, in essence, simply repetitive of Grounds I and II. Ground IV charges that The applicant, Thomas C. Stockebrand, signed a supplemental declaration for application Serial No. 409,913 on April 23, 1968, when he knew, or reasonably should have known, that the invention was “prima facie” “on sale in the United States more than one year prior” to the date of the application. Ground VII charges that The applicant and the assignee failed to disclose to the Patent and Trademark Office (PTO) prior to the issuance of U.S. Patent No. 3,387,293 on June 4, 1968, the fact that they knew, or reasonably should have known, that the invention was “prima facie” “on sale in the United States more than one year prior” to the date on the application. No additional evidence is recited in support of these grounds that might enhance the PTO’s treatment of the issues of materiality and culpability, discussed above with regard to the first two grounds. (The only additional fact recounted in support of Ground IV is that on April 11, 1968 DEC’S patent counsel sent Stockebrand the supplemental declaration with a cover letter informing him that “this Declaration brings up to date the statements you made when the application was filed, e. g., that the invention was not published or in public use more than a year prior to the date of filing.” Ground VII simply “incorporate[s] by reference” the discussions of Grounds I— IV.) D. Grounds VIII and IX deal with the failure of Stockebrand and DEC, respectively, to bring to the attention of the PTO two advertising brochures published by DEC in spring of 1963, which are said to have been “ ‘prima facie’ available as prior art” against the Stockebrand application. It is established that Stockebrand and various DEC officers and employees knew of the existence and contents of the brochures. The PTO accepted, arguendo, DEC’s characterization of the first brochure, which *721 had a reference date of March 1963, as describing “the tape system that DEC had planned to build (the Line tape unit) but never did.” On this basis, however, the PTO found “its materiality is clearly and convincingly established,” due to DEC’s stipulation that, at the time the application was filed, “LINC tape was the most pertinent prior art known to Stockebrand, DEC or DEC’s attorneys.” As is developed more fully below, see Part III E, this equation is ill-conceived; to say that a prior invention is the “most pertinent” prior art does not necessarily establish that invention’s materiality absent an evaluation of how pertinent it is. In this instance, however, the issue is further muddled, because in virtually the next breath the PTO concedes that “the disclosure in this brochure was quite skimpy standing alone and its materiality questionable.” The second brochure, dated May 1963, does describe certain aspects of the Stockebrand system. DEC nevertheless contended that the brochure “was still preliminary in nature” and maintained that it lacked “enough [detail] to disclose the invention.” While accepting, arguendo, DEC’s description of the brochure as “preliminary,” the PTO concluded that “the obvious materiality of the brochure .. . cannot be denied.” The PTO did not elaborate on this bare conclusion, and its decision is certainly devoid of any finding that the brochure would likely have been a bar to patentability under sections 102(b) or 103. Our own limited examination of the brochure suggests that there could at least be a reasonable difference of opinion concerning the extent to which the brochure “teaches” the Stockebrand invention. More serious than the truncated discussion of the brochures’ materiality is the absence of any discussion of Stockebrand’s or DEC’s culpability .in withholding the brochures from the PTO. Again, the PTO seems to have simply assumed that since the applicant and the assignee had “knowledge” that these materials existed, their failure to bring the materials to the PTO’s attention amounted to “fraud.” There is no evidence, however, to indicate that Stockebrand or any other DEC employee either 1) believed these brochures significant to the patentability of the invention or 2) were “grossly negligent” in failing to realize that the brochures would be a determinative factor in the examiner’s evaluation of the patent application.' The PTO could hardly make the latter finding, of course, since it made no predicate finding that the advertising materials would have played a dispositive role in the examination of Stockebrand’s claims. The PTO’s discussion of Ground IX, relating to DEC’s failure to disclose the materials, is devoted largely to demonstrating that DEC undertook in the spring of 1963 to generate “strong customer interest in Micro-tape,” and had established tentative prices for the system as early as March 30, 1963. It is difficult to comprehend what relevance these fácts have to the status of the advertising brochures as “prima facie prior art.” While this aspect of DEC’s advertising and sales program might conceivably be pertinent to the question of an “on sale” bar, it has no bearing on whether the brochures taught the invention sufficiently to render it unpatentable. E. Grounds XI and XII concern the failure of Stockebrand, DEC, and patent counsel (Liepmann and his successor, O’Donnell) to adequately disclose “LINC tape,” the “closest prior art of which they were aware.” It is established that Stockebrand was fully aware of the existence of LINC tape, having even played a part in its development, and that he disclosed the central elements of the LINC tape system to patent counsel. The application, however, did not identify LINC tape by name. It did contain a 12-line description which Stockebrand testified described LINC tape “exactly.” The PTO’s only discussion of the materiality of LINC tape to the patentability of the Stockebrand invention is its recital of DEC’s stipulation that “LINC tape was the most pertinent prior art known to Stockebrand, DEC or DEC’s attorneys.” This con *722 cession is not responsive, however, to the issue of “but for” materiality, as defined in Norton v. Curtiss, since it is quite possible (and even probable) that the “closest” prior art will not render an invention unpatentable or restrict the scope of the claims made. Nor did the PTO attempt to explain the effect that full disclosure of LINC tape would likely have had on Stockebrand’s application for patent. We can understand the frustration of the PTO at having discovered that it examined the Stockebrand application without having been apprised of materials it would have found “relevant” to that examination. Yet we can point to no source establishing an ironclad duty on the part of applicants for patent, as of June 4, 1968, to bring to the PTO’s attention “the most pertinent prior art” of which they were aware, even if that prior art was “more relevant than the prior art cited by the Primary Examiner.” See United States v. Standard Electric Time Co., 155 F.Supp. 949 (D.Mass.1957). The dispositive question would have been, at a minimum, whether the prior art was so close as to create a substantial likelihood that it would bar issuance of the patent. 26 Moreover, contrary to the PTO’s suggestions, in the context of an inquiry into fraud, “failures to disclose the most relevant prior art” can “be excused” by proof that those involved with the application acted in “good faith.” See International Telephone & Telegraph Corp. v. Raychem Corp., 538 F.2d 453, 461 (1st Cir.), cert. denied, 429 U.S. 886, 97 S.Ct. 238, 50 L.Ed.2d 167 (1976), and cases cited in Part II A, supra. In this case, the patent application contains, at col. 1, lines 52-64, a brief passage which DEC contends adequately describes LINC tape. It reads: The data stored on a magnetic tape are often organized into blocks encompassing a number of consecutive storage spaces. These spaces are arranged in parallel data tracks extending along the tape. The location of each block is recorded on the tape in a track separate from the data blocks. With this arrangement, as the magnetic tape advances, a mark track read/write transducer produces signals identifying the blocks moving into a read/write position with the data transducers. These signals are used to schedule the retrieval of information recorded in the tape and, alternatively, the recording of information. In his deposition, Stockebrand testified, referring to this passage: A. Well, specifically LINC tape is described exactly. The language is such that you could — it certainly also describes what I call Supertape somewhere else in here, the original TX-2 tape. Though it isn’t specific in the sense that you can’t say, well, it must have been that you were talking about — if you see what I mean. Certainly the only tape in existence that describes what I personally know about is LINC tape. The PTO’s response was that It appears from the above quotation that Stockebrand was using his knowledge of LINC tape to identify the passage as describing LINC tape. Further, Stockebrand’s knowledge of the original TX-2 tape was apparently also sufficient to enable him to identify the passage as applying to that tape also. However, there is no indication that one not having specific knowledge of either LINC tape or the original TX — 2 tape would recognize the same from the description in question. While the PTO’s comment is not entirely clear, we take it simply to be a finding that the application’s disclosure of LINC tape was not adequate, despite Stockebrand’s ability to associate the description with the preexisting system. .Taken as such, the finding is certainly supportable. It is obvious that the brief description of “prior art” at lines 52-64 of the application was not as complete a disclosure of LINC tape as *723 would have been possible. This conclusion is mandated in part by the fact that DEC now seeks to amend certain claims so as to avoid a “construction” that might be taken as reading on LINC tape — a construction that reveals elements of the LINC tape system not ascertainable from the brief description quoted above. Furthermore, as COI pointed out to the PTO, in a comment the PTO found “well taken,” the description at lines 52-64 does not reveal “that LINC searched bidirectionally and read and wrote monodirectionally,” that “LINC employed complement obverse block marks in the data tracks,” or that “as of 1962 ... it was a matter of choice to put information in obverse complement form or put in the obverse and use a complementing amplifier to achieve the result when data was stored in two different locations.” 27 Nevertheless, we find nothing in the PTO’s subsidiary findings indicating that the inadequate disclosure amounted to a misrepresentation “made in an atmosphere of gross negligence.” 28 The PTO does not seem to dispute DEC’s contentions that as of the filing date of November 9, 1964, there was no publicly available description of LINC tape, 29 that Stockebrand had to rely on his recollection of LINC tape, and that Stockebrand was not even aware of certain features of the LINC tape system (such as the system’s “fortuitous” use of complement obverse numbers as control *724 words.) 30 More important, there “is no attempt to directly confront Stockebrand’s assertion that he did not believe that the undisclosed elements of LINC tape would affect the patentability of his claims. Cf. Beckman Instruments, Inc. v. Chemtronics, Inc., 428 F.2d 555 (5th Cir.), cert. denied, 400 U.S. 956, 91 S.Ct. 353, 27 L.Ed.2d 264 (1970). 31 While Stockebrand might have been wrong in this belief (a fact, we point out, that the PTO has yet to determine), this in itself would not necessarily establish his culpability. And while Stockebrand can be faulted, in a general sense, for failing to explicitly refer the examiner to the LINC tape system, thus enabling him to undertake his own examination of LINC tape’s relevance, it is not clear that Stockebrand’s failure to do so was either part of an intentional scheme of deception or so grossly wrongful as to constitute fraud. P. Ground XIII, the only ground relating to the contents of the reissue application, charges as fraudulent conduct Applicant’s signing of the “Petition, Declaration and Power of Attorney” for reissue application Serial No. 546,238 containing a material mischaracterization or misstatement regarding the circumstances surrounding the discovery of the delivery to Kie Corporation. The “material mischaracterization” thus referred to is the following statement in the application for reissue: In investigating Computer Operations, Inc.’s charges as to the present patent, the assignee discovered that it had delivered to Kie Data, Inc., Watertown, Massachusetts, a prototype computer system containing a tape transport in accordance with the invention. This is incorrect, the PTO maintains, because DEC, the company “which actually made, and was responsible for the delivery,” would certainly have “known” of the delivery at the time it was made, and thus could not have “discovered” it over ten years later. We think this ground borders on being frivolous, and find disturbing the PTO’s reliance on it to establish “clear and convincing” evidence of fraud. The PTO’s dissection of the single word “discovered” is an exercise in hypertechnical nitpicking. First, in the context in which the statement was made, it heralds little more than the fact that the Kie Data delivery was “brought to DEC’s attention” in connection with the infringement litigation, regardless of whether DEC had been aware of the *725 delivery at some point in the past, 32 We note in this regard that the “memory” of a corporation is dependent on the memories of its individual officers and employees, not all of whom are kept continuously apprised of the company’s entire range of affairs, past and present. Moreover, while it may be “incorrect” to maintain that a company does not “know” of its own deliveries, and “illogical” to say that information can be “discovered” once it has been “known,” such a position is so “illogical” that it cannot reasonably be inferred that DEC intended the sentence in question to “deceive” the PTO concerning the extent of DEC’s “corporate knowledge” of the Kie Data delivery. 33 Moreover, even if one reads DEC’s perhaps euphemistic reference to the “discovery” of the Kie Data delivery as a gross misstatement of the facts, the statement’s immateriality is beyond question. The reissue application contains a brief description of the Kie Data delivery and promises that “[d]etailed documentation concerning the foregoing is being assembled by the assignee and will be forwarded to the Patent Office shortly.” This “detailed documentation” was subsequently filed on September 15, 1975, prior to initiation of the “fraud” proceedings, and contains a lengthy recital of the circumstances surrounding the Kie Data lease. The PTO was thus well positioned, by virtue of materials filed by DEC, to form its own “characterization” of DEC’s “awareness” of the lease, and had no reason to rely on the ambiguous one-sentence description in the reissue application or to be misled by any “mischaracterization” the sentence conveyed. In such circumstances, it was at best unnecessary for the PTO to parse the sentence in search of “fraud.” By this, I did not mean to suggest that sales by DEC were unknown to DEC since, obviously, its acts were known to it. IV. We wish to emphasize, in closing, the limited nature of our decision. We find only that the PTO has failed to set forth any basis to reasonably support its findings of “fraud,” and therefore lacked justification for its action striking DEC’s reissue application under Rule 56. Our decision cannot be taken as a determination that no “fraud” was committed; we are limited, on this appeal, to scrutiny of the agency record and analysis of the reasons advanced by the agency to support its action. See Vermont Yankee Nuclear Power Corp. v. National Resources Defense Council, Inc., 435 U.S. 519, 549, 98 S.Ct. 1197, 1214, 55 L.Ed.2d 460 (1978); SEC v. Chenery Corp., 318 U.S. 80, 63 S.Ct. 454, 87 L.Ed. 626 (1943). Nor do we mean to suggest, even on our examination of the present record, that DEC has behaved admirably before the PTO. A more complete disclosure of facts relating to LINC tape and to the Kie Data delivery would certainly have assisted the PTO in performance of its functions, and the present controversy concerning DEC’s “good faith” could have been avoided had Stockebrand and DEC chosen to resolve any doubt about the relevance of such facts in favor of full disclosure. It might even be desirable to impose a duty' of “due diligence” on applicants and their counsel to disclose any “relevant” facts to the PTO, thus ensuring full consideration of such facts by the agency. 34 There is a limit, *726 however, to the extent to which the pertinent requirement that applicants not act “fraudulently” can be used to establish such an affirmative “duty of disclosure.” This case provides a good illustration of the pitfalls of a concentration on an applicant’s alleged fraudulent conduct in preference to a thorough examination of the validity of his patent. The PTO had pending before it serious questions concerning the patentability of the Stockebrand invention in light of prior art and potentially disabling sales. While a determination of patentability would have required resolution of some disputes of a technical nature (such as the date of the invention’s reduction to practice, the applicability of the “experimental use” exception to section 102(b), and the “obviousness” of Stockebrand’s advances over LINC tape), the inquiry would have been in many ways more straightforward than the one undertaken here. Moreover, it is with regard to such matters, and not with regard to ferreting out “inequitable” conduct, that the PTO possesses unique expertise. The comments of the court in Pfizer, Inc. v. International Rectifier Corp., 538 F.2d 180, 196 (8th Cir. 1976), cert. denied, 429 U.S. 1040, 97 S.Ct. 738, 50 L.Ed.2d 751 (1977), directed to the proper conduct of infringement litigation, are at least as germane to proceedings before the PTO: “An infringement defendant in complex litigation should not be permitted to sidestep the [ ] main issues [of the validity and enforceability of the patent] by nitpicking the patent file in every minute respect with the effect of trying the patentee personally, rather than the patent.” See also Norton v. Curtiss, 433 F.2d at 795 (In cases where the “misrepresented” facts, in and of themselves, would probably render the patent invalid or unenforceable, “[w]hether the claims would also be unenforceable because a fraud was committed in misrepresenting the facts to the Patent Office would really be of secondary importance.”). We thus vacate the judgment of the district court and remand with instructions that it remand the case to the PTO for further consideration of DEC’S application for reissue. As we have noted, the PTO has not yet made a final determination on the merits of the application, aside from the issue of “fraud.” Nevertheless, as Part II of our opinion suggests, the PTO has authority to renew its Rule 56 proceedings should it think such a course fruitful. We are of the opinion on the basis of our reading of the present record, however, that its efforts would more profitably be directed elsewhere. Vacated and remanded, with instructions to remand to the Patent and Trademark Office for further consideration. Footnotes [1] The patent, headed “Bidirectional Retrieval of Magnetically Recorded Data,” claims an invention that “provides rapid and flexible access to the data storage spaces on [a moving] record medium,” through use of “selected digital control words” on a mark track to designate data storage “blocks.” Because the selected control words “are located in corresponding positions at the beginning and end of each block and have the property that the signal produced by reading one of them when the medium is moving in one direction is identical with the signal produced by the corresponding word when the medium is moving in the opposite direction,” the location of information can be identified with equal facility during motion of the medium in either direction. While the invention is claimed to be “useful with data storage media such as drums, tapes and disks, using magnetic, optical or other storage techniques,” the DEC tape system developed by Stoekebrand uses a magnetic tape medium and incorporates various features of “conventional” and “previously known” tape storage systems. Data is stored on multiple *704 track tape in the form of “magnetic spots.” The tape is moved, by tape drive motors, past an electrical device called a “read/write transducer" which (by transforming the magnetic field into an electrical signal) “reads” each spot as either a “1” or a “0”; the same magnetic spot will be read as a 1 when the tape is moving forward and as a 0 when the tape is moving in reverse. Combinations of l’s and 0’s form binary-coded numbers, interpreted as “data words.” The DEC tape system shares these features with many preexisting systems. According to the invention, the multiple track tape is organized into “blocks” and one of the parallel tracks is designated as a “mark track” for purposes, inter alia, of identifying the location of each block. This is accomplished by recording an identifying “control word” at the beginning of each block. This method of organization is used in various systems predating Stockebrand’s, including “LINC tape.” As described by DEC, Stockebrand’s major contribution over the previously known systems was his discovery that a “reverse block mark” could be placed at the end of each block which would be the “complement obverse” of the binary number constituting the “forward block mark,” placed at the beginning of that block. (For example, 010110 and 100101 constitute a pair of numbers each the complement obverse of the other.) With the tape moving in reverse, this reverse block number would be read the same as would the forward block number when the tape was moving in the forward direction; thus the position of the data would be demarcated by an identical signal regardless of the direction of tape movement. This advance allegedly permits more rapid access to data than would have been possible through use of preexisting systems, such as LINC tape, at least absent additional, complex circuitry. i [2] According to the affidavit of John T. Gilmore, Jr., co-founder and Vice President of Kie Corporation, and since 1974 an employee of DEC, the computer equipment leased from DEC by Kie in July 1963 was used in connection with a large data processing project that Kie was performing. Gilmore avers that Kie “had a close working relationship with DEC,” that it “was not uncommon [for Kie] to cooperate with DEC in developing and testing new product applications,” and that Kie was aware that the Micro tape units it was leasing “would still have to be proven in actual use and that changes might have to be made in them after connection into the system at Kie Corporation in order to ensure satisfactory operation.” Because the Micro tape units “were still being developed and were not available for use” when the lease began in July, DEC supplied Kie with “its then standard Potter units” which were “used until the Microtape units were available and operating.” During December 1963, Kie began transferring the data stored on the Potter tape units to the Micro tape units, a process that continued through January 1964. A DEC document dated January 20, 1964, “Minutes of the Engineering Projects Committee,” notes that “[t]he operation of these Microtapes at Kie will be an excellent test of the equipment as Kie is being forced to use them since we are removing the Potter units.” [3] These orders were from the Atomic Energy Commission of Canada, Ltd. (AECL); Fort Meade; and MIT “Project MAC.” [4] Stockebrand informed Liepmann of LINC tape, as well as other data storage systems, in a letter of February 14, 1964; Liepmann was thus aware of LINC tape at the time he wrote the application for patent. In his affidavit, Liepmann avers that “[p]rior to the preparation of the application, it was determined that the subject matter was patentable over what was known as ‘LINC’ and other prior art." At the November 5, 1964 meeting, Stockebrand and Liepmann reviewed the disclosure in the application “to ensure that it contained a correct and accurate description.” Liepmann was concerned that a paper discussing DEC tape written by Leonard Hantman, a colleague of Stockebrand’s at DEC, “might possibly give rise to a [35 U.S.C.] section 102 bar.” Although the paper was presented to a conference on November 17, 1963, it was determined that the paper was first printed in draft form on November 7, 1963. It was thus decided to file the application by November 9, 1964 (a Monday) to satisfy the requirements of section 102. [5] Prior to 1949, PTO Rule 31 provided that: Every application signed or sworn to in blank or without actual inspection by the applicant of the petition and specification, and every application altered or partly filled up after being signed or sworn to, will be stricken from the files. Rule 56 incorporated most of Rule 31, while adding the references to “fraud” and “fraudulently filed” applications. While Rule 31 did not contain the “fraud” terminology, decisions applying the Rule often referred to the conduct it proscribed in terms of “fraud.” E. g., Ex Parte Mallard, 71 U.S.P.Q. 294, 296-97 (Comm’r Pat. 1946). [6] Sections 116, 251 & 256 of Title 35 speak of errors made “without any deceptive intention”; section 288 deals with instances when, “without deceptive intention,” one of multiple patent claims is invalid; and section 255 refers to mistakes which occurred “in good faith.” Section 32 provides' for suspension from practice before the PTO of persons “guilty of gross misconduct” or who “with intent to defraud in any manner, deceive, mislead, or threaten any applicant. ... ” [7] DEC relies primarily on dictionary definitions of “deception” and “deceive” to support the proposition that “deceptive intention,” as used in section 251, is equivalent to the common law concept of “scienter.” DEC seems to ignore the fact that section 251 refers to “error without deceptive intention”; one might conclude that misrepresentations made in an atmosphere of gross negligence do not constitute “error.” See In re Clark, 522 F.2d 623 (Cust. & Pat.App. 1975). [8] Section 251 also incorporates by reference the provisions of Title 35 “relating to applications for patent....” See also PTO Rule 176, 37 C.F.R. § 1.176. [9] PTO Rule 175(a)(4), 37 C.F.R. § 1.175(a)(4), as amended in 1977, entitles an applicant to apply for reissue: When the applicant is aware of prior art or other information relevant to patentability, not previously considered by the Office, which might cause the examiner to deem the original patent wholly or partly inoperative or invalid. . .. Nevertheless, section 1414.02 of the Manual of Patent Examining Procedure (MPEP) provides that “no reissue application will be passed for issue with only a § 1.175(a)(4) type oath or declaration.” If a reissue application is filed only pursuant to section 1.175(a)(4), e. g., to disclose prior art and sales not previously considered by the PTO, the proper disposition is to “reject [the application] as lacking statutory basis for a reissue, if there are no other grounds of rejection. ...” However, “the record of prosecution of the reissue will indicate that the prior art has been considered by the examiner.” Id. In essence, then, the section 1.175(a)(4) “reissue” process is simply a mechanism by which patentees can seek a PTO determination that their claims are patentable over newly cited prior art (or other information relevant to patentability) — a determination that might prove useful to the patentee, for example, in prosecuting a pending infringement action. See, e. g., Dunner & Lipsey, The New Reissue Practice, 61 J. Patent Off. Soc. 68, 80 (1979). [10] The primary “error” which DEC sought to cure through reissue was the statement, in claims 1 and 12, that the Stockebrand invention includes pairs of control words, each the complement obverse of the other, “selectively positioned” on the mark track to achieve bidirectional reading and writing of data. See note 1, supra. Since the LINC tape system utilized control words “one or more of which may, in fact, fortuitously be the complement obverse of the other” (but, according to DEC, not “ ‘selectively’ positioned with respect to each other to produce the desired result”), DEC sought to “elaborate” on claims 1 and 12 to maintain that the marks are “selectively and correspondingly" positioned. Insofar as the original claims are, as DEC contends, “subject to a construction which covers more than the applicant is claiming,” thus rendering the patent “partly inoperative,” the section 251 requisites would seem, on this basis, to be met, provided the original terminology (i. e., “selectively positioned”) was not used with a “deceptive intention” as to the scope of the claims. (Even so, to the extent the proposed change in terminology is an attempt to distinguish prior art not cited to the examiner (i. e., LINC tape), a finding that the prior art was “fraudulently” withheld might preclude a determination that the original, broader claim was “error” within the meaning of section 251. See In re Clark, 522 F.2d at 626-27; but cf. id., at 628-33 (Miller, J., concurring).) [11] As a result of the PTO’s 1977 expansion of its reissue procedures (see note 9, supra), the Office was enabled to reexamine, upon the patentee’s application for reissue, the validity of a patent in light of prior art or other information relevant to patentability not considered prior to the time of the patent’s issuance. It has been widely suggested that courts should take advantage of the new procedure, by staying infringement actions in which the alleged infringer cites such prior art or information as evidence of the patent’s invalidity, in order to allow the patentee to seek “reissue” and thus obtain the PTO’s expert determination of the effect of the prior art or information on the patentability of the claims. See Markey, New Patent Office Rules May Aid Patent Litigation, Vol. 9, No. 9, The Third Branch 7 (Sept. 1977) (recommendation by Chief Judge of the Court of Customs and Patent Appeals that infringement actions be stayed in favor of reissue proceedings); see also Dunner & Lipsey, The New Reissue Practice, 61 J. Patent Off. Soc. 68, 84 n.8 (1979) (cataloguing cases in which requests for such stays have been granted or denied). It may thus become more and more commonplace for the PTO to be required to act on a reissue application prior to a judicial determination of the validity and enforceability of the original patent. While it does not appear that a stay of the related infringement action was sought here, DEC did successfully resist COI’s request that the PTO stay its proceedings pending the outcome of the companion litigation. [12] DEC also argues that “Rule 56 by its terms does not apply in reissue applications to alleged ‘fraud’ in the procurement of the original patent, since the alleged ‘fraud’ was not practiced ‘in connection with it [the reissue application].’ ” We think, however, that fraud in the procurement of claims can reasonably be said to have been practiced “in connection with” an application seeking reissue of the same or similar claims. Furthermore, the PTO’s interpretation of its own rule in this regard is entitled to substantial deference; we have already concluded that its interpretation is not “inconsistent with law.” [13] In Hengehoid, the court held that the Board lacked jurisdiction to review an examiner’s “requirement for restriction.” 35 U.S.C. § 121 provides that “[i]f two or more independent or distinct inventions are claimed in one application, the Commissioner may require the application to be restricted to one of the inventions”; section 121 explicitly refers to the Commissioner’s action as a “requirement for restriction.” The terms “requirement” and “objection” are used in Title 35 to refer to actions distinct from “rejection.” See 35 U.S.C. § 132. It is thus quite reasonable to conclude, as did the court in Hengehoid, that when Congress provided in section 134 for appeal to the Board by applicants whose claims had been “twice rejected,” it intended to exclude from the appeal process “requirements” and “objections” (which, in any event, are normally actions of a procedural, provisional, or discretionary nature). 440 F.2d at 1402-03. The same, however, cannot be said of actions taken pursuant to Rule 56, an “extra-statutory” provision; that the Commissioner has chosen to style such actions in terms of “striking of applications” rather than “rejection of claims” is not evidence of Congress’ “intent” regarding appealability. [14] DEC maintained in its principal brief that “in the constitutional law, . .. the cases uniformly recognize the need for an opportunity to present demeanor evidence prior to resolution of issues of credibility and veracity similar to those here resolved by the PTO on paper.” However, in response to the PTO’s and COI’s arguments that the PTO procedures comported with “due process,” DEC’S reply brief states: In order to simplify the issues before this Court, and to clearly focus the Court’s attention on the inconsistency of the PTO action with the existing statutory framework, DEC does not now, and did not in its principal brief, assert a claim to denial of due process per se. [15] These were grounds I, IV, VII, and VIII (mislabeled “VII”). [16] The parties agree that section 706(2)(A) supplies the proper standard of review of this issue. [17] We note, in this regard, that the current version of Rule 56, promulgated in 1977, seems to adopt a definition of materiality more expansive than that applied in “fraud” cases such as Norton. The rule states that All [inventors, assignees, attorneys, etc.] have a duty to disclose information they are aware of which is material to the examination of the application. Such information is material when there is a substantial likelihood that a reasonable examiner would con sider it important in deciding whether to allow the application to issue as a patent. (Emphasis added.) Such a “duty of disclosure” may well be desirable as a matter of regulatory policy: we do not suggest that it could not be applied prospectively in proper circumstances. It would obviously be improper, however, to judge the conduct of these applicants retroactively in terms of a “duty” created by a regulation promulgated years after the events at issue. [18] A DEC document dated July 15, 1963, “Minutes of Engineering Projects Committee,” observes that “Tom Stockebrand reported that Microtape was now running satisfactorily (reading and writing).” At his deposition, Stockebrand was shown the document and asked whether the statement was true. His response was: A. It was reading while writing in July 1963. It was not working. Q. Why was it not working? A. Because reading while writing, while a very important part of the operation of the system, was only a small part.. .. At another point in his deposition, the following exchange occurred between Stockebrand and COI’s attorney: Q. And before you shipped [the machine delivered to Kie], could it read and write on tapes? A. Sure. Q. And when it was delivered, could it read and write on tapes? A. No. Q. Why not? A. Broke. Q. What broke? A. Wires. It overheated. Had to cut a hole for a fan. We had made a couple of logic errors, so that we thought it worked before we shipped it. While the PTO has quoted selectively from these sources to suggest a concession that the Kie system was operable prior to delivery, DEC’S position is that the part of the system that was running “satisfactorily” prior to delivery “had nothing to do with the invention,” and that the repair of the “logic errors” discovered after delivery was fundamental to completion of the invention. The PTO decision does not attempt to refute this position. [19] The PTO decision notes DEC’S contention that “the Stockebrand invention was neither in the public use nor on sale in 1963” and responds that “the facts to the contrary are convincing,” because “not only were orders solicited and accepted prior to the critical date, but one of the deliveries actually occurred prior to the critical date.” (No mention is made here of DEC’S position that the “delivery” consisted of an incomplete and inoperable invention and was made primarily for “experimental purposes.”) Later, however, the PTO retreats from any finding that the invention was “on sale” prior to the critical date, concluding that “[e]ven if it could be construed that sales involved an inoperable invention that had not been tested or that the sales were only for experimental purposes,” Stockebrand should nevertheless have disclosed the underlying facts. [20] We observe that, prior to the PTO’s determination of “fraud,” the examiner assigned to DEC’S reissue application rejected all claims as barred by section 102(b), due to the “admission that the device was ‘on sale’ more than one year prior to the filing date of the original patent.” While it is arguable that this fact would have provided support for a finding that the original examiner likewise would have rejected Stockebrand’s claims had he been apprised of the same information, the PTO’s decision does not rely on (and indeed nowhere mentions) the reissue examiner’s action. It is not our function to supply a rationale to support the PTO’s action; the PTO decision must “stand or fall” on the reasoning stated therein. Camp v. Pitts, 411 U.S. 138, 143, 93 S.Ct. 1241, 1244, 36 L.Ed.2d 106 (1973); SEC v. Chenery Corp., 318 U.S. 80, 63 S.Ct. 454, 87 L.Ed. 626 (1943). [21] In its brief, the PTO seems to suggest that since Stockebrand “is not a lawyer and was not familiar with the requirements for an ‘on sale’ bar,” he is precluded from relying on such factors as the invention’s lack of operability and reduction to practice or the sale’s “experimental purpose” to support his averment of a “good faith belief’ that the invention was not on sale before the critical date. The concept that an invention must be operable and reduced to practice before it can be deemed “on sale,” however, is rooted in scientific reality and simple common sense; certainly this notion could have occurred to Stockebrand, a scientist and inventor. (Indeed, DEC argues that Stockebrand’s lack of experience with patent law cuts the other way — that as a scientist he “just did not appreciate that the very activities in 1963 which failed to achieve the intended realization of DEC tape might, nonetheless, be deemed sufficient to preclude his right to a patent. .. . ”) [22] It is true that Stockebrand avers in his affidavit that At the time I signed the application as well as at the time I signed the supplemental declaration, I do not believe that I specifically considered any acts such as leases, sales, offers for sales, brochures, advertisements, or any such acts or information with specific relation to the documents I then signed, nor do I now believe that I had any reason to believe at this time that I could not properly make the oaths I then signed. This does not indicate, however, that Stockebrand had “forgotten” that such leases, sales, and the like ever occurred, only that he did not deem them relevant to the papers he was signing. [23] The PTO, of course, was entitled to assess Stockebrand’s “credibility” (at least to the limited extent that “credibility” can be determined from a paper record of the sort before the PTO), and to reject his testimony on this basis. As already noted, however, Stockebrand’s affidavit was dismissed as “incredible” solely on the basis of a misapprehension of its contents. [24] Richard Best, DEC’s Chief Engineer, was found to have been “aware of both the sales and the application,” and to have “read the patent application before it was filed.” Ken *720 Olsen, President of DEC, “knew of the sales” and was “aware of ... the filing of the application,” although there was “no testimony or evidence to indicate that Olsen actually read the application. . . . ” Donald Vonada, a DEC employee hired in September 1963 to introduce Stockebrand’s invention into production, was “involved in the preparation of the application” (by virtue of his having reviewed a draft and made comments in the margin) and, the PTO found, must have known of the sales, because it would be “incredible that DEC would hire Vonada to introduce Stockebrand’s invention into production and not inform him of the various sales made to Kie, AECL, MIT and Ft. Meade.” [25] We cannot, however, accept DEC’S position before the PTO that it was exonerated from responsibility for the nondisclosure because “the disclosure would normally be made, and can be expected to be made, by the inventor.” Here, the inventor did not make the disclosure, either to patent counsel or to the PTO, yet various DEC employees, aware of the sales and of the fact that an application for patent was about to be filed, were in a position to intervene and fill the gap. We certainly would not approve the proposition that a corporate assignee of a patent cleanses itself of all culpability for misrepresentations inserted in a patent application filed on its behalf by purposely keeping its officers in the dark concerning the application’s contents and placing the entire burden of proper disclosure on an inventor not versed in patent law and an attorney kept ignorant of putatively relevant facts. We hold only that on the evidence presented here, DEC was not shown to have acted recklessly or grossly negligently in this regard. [26] In its brief, the PTO asserts that “the applicant must disclose the closest prior art of which it is aware which would anticipate the invention or in fact would prevent the patent from issuing’ (emphasis added). [27] It was not established, however, that these similarities would bar patentability of the Stockebrand invention. With respect to LINC’s ability to search bidirectionally and read and write monodirectionally, for example, Stockebrand testified: A. Look, what I’m inventing or is in the patent is built on LINC tape. Clear. It goes beyond it with complement obverse, with symmetry, and one or two other things.... As far as I’m concerned, it was what was added to the prior art, is my invention. The PTO made no finding to the contrary. DEC also contends that the prior art that was considered by the PTO prior to issuance of the Stockebrand patent “clearly sets forth that bidirectional searching was known.” As mentioned infra, Stockebrand denied knowledge, prior to issuance of his patent, that LINC made any use of complementary obverse numbers in the data tracks. DEC argues further that such usage is in any event immaterial and would not render the Stockebrand invention “obvious.” COI also maintained before the PTO that the Stockebrand application did “not disclose the LINC mark track format,” did “not disclose the technique which LINC employed to read the mark track,” and did “not disclose the data track format employed by LINC in the prior art system.” Aside from the issue of the materiality of the above to what Stockebrand claimed as his invention, we note that these statements are not entirely accurate. The Stockebrand application did disclose, in general outline at least, 1) the practice of recording the location of each block in a mark track separate from the data tracks, 2) the prior use of a “read/write transducer” to retrieve information recorded on the mark track, and 3) the previous usage of separate data tracks. [28] The PTO makes much of the fact that Vonada, who reviewed a draft of the application for patent prior to filing, pencilled the following comment in the margin adjacent to lines 52-64: What did you want to say. There are other systems in the field doing this or there is only ours? (In turn, Stockebrand apparently placed a question mark next to Vonada’s comment.) Vonada testified at his deposition that I guess the intent in the writing here was to ask whether he should be more clear about whether he was making a distinction that in general magnetic tape is organized into blocks, or that he’s trying to prescribe specifically DEC’S method of doing it. From this the PTO concluded that Stockebrand and Liepmann were alerted that the description was inadequate. Even so, it seems far-fetched to assert that their failure to accede to Vonada’s rather vague editing suggestion (particularly in light of the fact that the passage already began with the words “The data stored on magnetic tape are oñen organized....”) amounts to clear and convincing evidence of fraud. Liepmann avers in his affidavit that it was his “recollection that Vonada was only inquiring whether the passage meant to deal with prior art or with the invention,” and, as such, did not consider it “to call for further description of prior art.” [29] The record does contain a letter dated May 13, 1966 from Spear, Inc. to a project engineer at Johns Hopkins Medical School, sent in response to an inquiry about LINC tape, to which is attached a lengthy description of LINC tape “in extremely rough form.” The letter states that “there has never been a formal write-up prepared to describe the tape system in any detail.” The record also contains a one-page, undated advertisement entitled “Plug a Line tape into your PDP-11 Computer,” which includes a sketchy description. [30] Although it is not entirely clear from the record, it appears from Stockebrand’s deposition testimony that, while he was associated with the LINC tape project at MIT Lincoln Laboratories, he had primarily a “maintenance function” and did not contribute to development of LINC’s control format. [31] In contrast to the lack of such findings in this case, the court in Beckman found that the applicant and the corporate assignee’s patent employees clearly “understood the significance” of the withheld prior art “and knew that it presented a serious threat to the validity of the patent they were trying to secure.” 428 F.2d at 564. See also In re Clark, 522 F.2d 623, 624 (Cust. & Pat.App.1975) (applicant clearly knew prior art “to be of pivotal importance in determining the patentability of [his] claims”). This conclusion was drawn in part from a memorandum written by a patent liaison employee sent to the company’s manager of engineering, stating: A related matter has now come up that creates a rather ticklish situation. As of early 1954, Dr. Richard W. Stow . . . corresponded with our Sales Department ... concerning a glass electrode . .. [which] is similar to the Clark design.... The Stow device is clearly within the scope of claims now in the Clark application .. . Clark and Stow have known about each other’s work.... (Emphasis added.) Furthermore, the applicant and assignee had “not only failed to present the Stow device to the Patent Office, but also made affirmative representations in the papers they filed to the effect that there had been no previous inventions displaying the properties that the Clark device had in common with the Stow device.” 428 F.2d at 566. Finally, the court found that the Stow device in fact rendered the Clark patent invalid, as anticipated by prior art. It was on the basis of these findings that the court rejected, as “utterly incredible,” the appellants’ contention that they did not consider Stow’s work relevant. Id [32] DEC’s amendment to its reissue application, filed September 15, 1975, three months prior to the filing of COI’s petition to strike the application for fraud, states that the information “came to the attention of applicant” during the course of discovery proceedings in the pending infringement action. [33] In his affidavit, Counsel O’Donnell, who prepared the application for reissue and drafted the sentence in question, averred: [34] See note 17, supra. It has for some time been a matter of concern that in a significant number of instances the PTO has issued patents without being aware of prior art later cited as a bar to patentability by defendants in infringement actions. See, e. g., Waltershied & Cage, Jurisdiction of the Patent & Trademark Office to Consider the Validity of Issued Patents, 61 J. Patent Off. Soc. 44 (1979). This concern prompted the PTO’s 1977 expansion of the reissue process. 37 C.F.R. § 1.175(a)(4). It has also led to the recent enactment of legislation enabling the Commissioner, at the request of any person or on his own initiative, to reex *726 amine the validity of a patent in light of newly cited prior art. 1980 Act Affecting Patents, Pub.L.No.96-517, 94 Stat. 3015 (to be codified at 35 U.S.C. §§ 301-307).