This patent appeal is from the October 27, 1982, judgment of the United States District Court for the Southern District of Texas. Sitting without a jury, it held claims 1-3, 6, and 8-12 of U.S. Patent No. 3,816,092 (RH patent or Wilson patent), issued June 11, 1974, to Rohm and Haas (RH or R & H) as assignee of the inventors, for "Herbicidal 3,1-dichloroanilides," valid and infringed by Crystal Chemical Co. and Joe C. Eller (collectively Crystal).1 Because R & H intentionally made material misrepresentations to the U.S. Patent and Trademark Office (PTO) the effects of which were not eliminated prior to issuance, we reverse the district court's holding that there was no fraud and remand the case for reconsideration in light of our decision that the Wilson patent is invalid.

The extensive opinion of the district court is reported at 557 F. Supp. 739-85, 217 U.S.P.Q. (BNA) 515-603 and familiarity with the procedural history and substantive background of this dispute as therein set forth will be presumed. We find it necessary at this juncture to discuss only the issue of fraud as it affects patent validity.

Of 220 Findings of Fact, the district court made the following 19 findings with respect to Crystal's claim that the RH patent was procured with affidavits furnished to the PTO containing misleading, falsified, and misrepresented data. Each of these findings appears under the district court's subheading "6. Renewed Prosecution of the 1961 Application." 557 F. Supp. at 774, 217 USPQ at 541. We summarize them as follows, adding some pertinent comments and giving their USPQ locations and finding of fact numbers:

1. On February 16, 1973, RH held an "extensive interview" with patent examiner Thomas, after intensive preparation "to identify areas of potential interest to the patent examiner and to collect and organize documents in logical sequence." Representatives of both the Department of Justice, which was investigating Monsanto in connection with Monsanto's fraud in the PTO, and the PTO Solicitor's Office, were invited to attend, as was Mr. Irving Marcus who was chief of the examining group. Only Thomas was present throughout the interview. Marcus was present at the beginning and remained available for consultation if serious questions arose. Finding 108 at 557 F. Supp. at 775, 217 U.S.P.Q. (BNA) 541-42. In its brief, RH stated that, on calling Mr. Marcus to arrange an interview, "Mr. Marcus suggested that the Solicitor's Office might wish to attend the interview, and invitations were extended to the Solicitor's Office, the Justice Department, and Mr. Marcus."

2. Thomas had been responsible for declaring Interference 93,751 among Monsanto, RH, and Bayer, and rendered the decisions on motions during the motion period. Finding 109 at 557 F. Supp. 775, 217 U.S.P.Q. (BNA) 542.

3. At the time of renewed prosecution, Thomas was a primary examiner. Since 1956 he had examined applications dealing with herbicides, he was the recognized PTO herbicide expert, in 1973 he had 8 or 10 other pending applications dealing with the same subject matter as RH's 1961 application, he was well aware of the herbicidal activity of propanil and related compounds, he was familiar with the review and interpretation of herbicidal data sheets, and he was familiar with the prior art. Finding 110 at 557 F. Supp. at 775, 217 U.S.P.Q. (BNA) 542.

4. During the interview, documents were spread out for ease of reference on conference room tables. Thomas brought the RH prosecution files with him and indicated that he had reviewed them prior to the interview. Finding 111 at 557 F. Supp. 775-76, 217 U.S.P.Q. (BNA) 542.

5. During the interview, RH summarized and explained key events of prior proceedings including, but not limited to, the interference and the Monsanto litigation. RH left Thomas copies of trial court and appellate opinions, RH's appellate brief, and all four volumes of the appendix before the Third Circuit in the Monsanto case (the appendix being marked for items that might be of particular interest to Thomas). Thomas stated that he intended to review these items after the interview. Finding 112 at 557 F. Supp. 776, 217 U.S.P.Q. (BNA) 542.

6. On returning these documents several weeks later, Thomas stated that he had reviewed the pertinent parts. Finding 113 at 557 F. Supp. 776, 217 U.S.P.Q. (BNA) 542.

7. RH's discussion of the Monsanto litigation and its disclosure of Monsanto's contention that RH had withheld herbicidal data from the PTO, together with its counsel's notation to Thomas of the trial court's comment that "Monsanto and Rohm and Haas were equally disingenuous with respect to the comparison of the herbicidal effects of DCAA and DCPA in their applications to the Patent Office," alerted Thomas to the importance of the comparative data. Only the testimony of McRae, one of the inventors, and of Hutz, RH's lead counsel, is cited in support of this finding. Finding 114 at 557 F. Supp. 776, 217 U.S.P.Q. (BNA) 543.

8. Hutz and McRae also brought a "substantial number of documents" called the "500,000 series" documents. The court found that although Hutz repeatedly offered to file the documents with the PTO, Thomas elected to have them remain with RH. Finding 115 at 557 F. Supp. 776, 217 U.S.P.Q. (BNA) 543.

9. RH also discussed with Thomas the prior art that had been raised during prosecution of the RH applications, the prior art relied on by RH in the Monsanto litigation, and the prior art that had been cited against Monsanto during prosecution of its application. Finding 116 at 557 F. Supp. 776-77, 217 U.S.P.Q. (BNA) 543.

10. In preparation for the interview, RH collected all records available to it reflecting herbicidal test data on propanil and related compounds regardless of whether the related compounds were prior art. These were the "500,000 series" documents. The identity of the compounds for which data existed was brought to Thomas' attention. Finding 117 at 557 F. Supp. 777, 217 U.S.P.Q. (BNA) 544.

11. These data were used to prepare several affidavits that had been filed in the PTO during prosecution of the 1958, 1960, and 1961 applications. "To the extent that such base records could be reconstructed, they were identified, keyed to the prior affidavits, compared side by side with each of the affidavits and discussed with" Thomas. RH "did not limit its disclosures to [Thomas] to tests from which affidavit data had been taken. Instead, it supplied all data available to it on related compounds." Finding 118 at 557 F. Supp. 778, 217 U.S.P.Q. (BNA) 544.

12. Thomas was told that RH had been unable to locate all the test data in its records corresponding to data recited in the affidavits filed with the PTO, and that there were data in its records that had not been initially reported to the PTO during the prior prosecution. Finding 119, at 557 F. Supp. 778, 217 U.S.P.Q. (BNA) 544.

13. Test data specifically identified and discussed indicated that propanil was not unique as a pre-emergence or general herbicide, that propanil damages some crops, that related compounds displayed relatively high herbicidal activity, and that related compounds like DCAA and DCMVA possessed some selective, post-emergence herbicidal activity. Finding 120, at 557 F. Supp. 778, 217 U.S.P.Q. (BNA) 544.

14. Even though all data on propanil's herbicidal activity were not discovered by RH, "a search for all of such data was conducted." Finding 121, at 557 F. Supp. at 778, 217 USPQ at 544.

15. The interview lasted about 2 1/2 hours, which is "considerably longer than normal interviews." Finding 123 at 557 F. Supp. 778, 217 U.S.P.Q. (BNA) 545.

16. Following the interview, RH filed an amendment summarizing the interview. It offered again to supply more information and copies of any document discussed. Finding 124, at 557 F. Supp. at 778-79, 217 USPQ at 545.

17. At the conclusion of the interview, RH was of the good faith belief that Thomas appreciated the information supplied and its implications. He was again reminded of these disclosures at a subsequent interview, and he told RH that he was fully satisfied and required no further information. Finding 125, at 557 F. Supp. 779, 217 U.S.P.Q. (BNA) 545.

18. Regarding the affidavits relied on by RH during prosecution: "As defendants argued and as plaintiff has admitted, at the time the six allegedly defective affidavits were submitted on the Patent Office, all of the test data were not disclosed in the affidavits." The employee who prosecuted the application and decided which data to submit died before its issuance, and "the reason behind his actions were not disclosed during the trial * * * and it has been represented to this Court that such motives, if any, are not known." Although not wholly a finding of fact, the court also here stated: "What is important to this Court's inquiry into the validity of Rohm and Haas' patent in light of defendants' fraud allegations is the fact that McRae, the person who executed but did not prepare the affidavits, believed such affidavits and the assertions contained in such affidavits to be accurate."2 RH informed Thomas of the omissions of test data from the affidavits and supplied to the extent possible the omitted data and all other herbicidal data in RH's possession. "Consequently, the Court is of the belief that all of the deficiencies which existed in the prosecution of the three Rohm and Haas patent applications from 1958 to 1963 were corrected by the complete and detailed disclosure of herbicidal data and information to the Patent Office in 1973; such disclosure was made within sufficient time for the Patent Office to take whatever action it deemed necessary." No claims had been allowed prior to RH's "full disclosure in 1973," except claim 9 "which had been allowed for interference purposes only and had been cancelled subsequently." Finding 128 at 557 F. Supp. 780, 217 U.S.P.Q. (BNA) 546.

19. RH's dealings with Thomas during renewed prosecution of the 1961 application "were reasonably based on its belief that patent examiner Thomas was qualified as an expert in the herbicide field, experienced in the prosecution of patent applications, and able to read and comprehend herbicide data sheets of the type that was discussed" during the interview. Finding 129, at 557 F. Supp. 780, 217 U.S.P.Q. (BNA) 546.

The trial court filed 133 Conclusions of Law. Based on the above summarized findings of fact, and in view of Fifth Circuit precedent on "fraudulent procurement" (Conclusion 32), it stated a single conclusion with respect to Crystal's assertions of fraud in the PTO, which we quote verbatim:

33. After carefully reviewing the record in this cause regarding the prosecution of the Rohm and Haas patent application in light of the arguments raised by defendants, the Court concludes that defendants have failed to carry the burden and adduce sufficient probative evidence to establish that Rohm and Haas procured the patent in suit by fraud or inequitable conduct. The record of the prosecution of plaintiff's patent application, particularly that part of the record regarding the renewed prosecution in 1973, unequivocally indicates that the patent examiner was fully informed of all the factors material to the patentability of the methods claimed in plaintiff's patent. Refer to Findings 108-123, 128, 129. Although there is evidence that Rohm and Haas omitted herbicidal test data from affidavits submitted during early prosecution of plaintiff's patent application, these omissions were brought to the patent examiner's attention during the renewed prosecution of the patent in suit. Moreover, Rohm and Haas provided the patent examiner with the omitted data, where possible, and all additional herbicidal data in its possession. Refer to findings 117-120. Nor can the Court find fault sufficient to warrant a finding of fraud or inequitable conduct with the oaths executed by Wilson and McRae and contained in the 1960 and 1961 applications. Refer to Findings 70, 80. Consequently, the Court is unable to conclude that the patent in suit was procured by fraud. Refer to Findings 28, 29. [Case citations omitted; 557 F. Supp. at 810-11, 217 USPQ at 571.]

Crystal admonishes that the trial court's opinion "totally overlooked R & H's most egregious misrepresentations." We certainly agree that the facts revealed by the arguments on appeal are nearly wholly absent from the opinion below.

Crystal states that propanil was compared to DCAA (3,4-dichloroacetanilide, its adjacent lower homolog) in Affidavit D which was submitted during prosecution of the 1960 application on July 11, 1961, and during prosecution of the 1961 application on July 20, 1961. Apparently, a first draft of the affidavit was destroyed and another one containing different comparative data was substituted. Crystal asserts that the first draft contained data from propanil versus DCAA tests run in December, while the second affidavit compared propanil kill data from the December tests and DCAA kill data from tests run in May, a time of year when plants are well known to be more resistant to herbicides. A June 21, 1961, letter from the R & H patent department to R & H's Washington counsel explains:

You should discard the copy of the Affidavit which I left with you, since this new Affidavit is expanded from the old one. In the old one, we were perturbed by the comparison of [propanil] with [DCAA], since it seemed to be the one weak spot in the Affidavit and might have given the Examiner and/or Board the opportunity to say that the two differ only in degree and not in kind. These new data obtained relatively recently remove any doubt on this score. In case you are wondering how it is possible to get two such different sets of data on the same compound I should point out that the time of year at which the plants are grown has a very real effect on their resistance to kill.

Crystal states that both McRae and Wilson acknowledged that one comparing such data would need to know that the tests were run at different times in order to properly evaluate the data, and emphasizes that these new data showed that at one pound per acre DCAA had zero effect on 13 of 15 plants tested, whereas the original December data showed the DCAA had kill rates ranging from 20% to 100% on 8 different plants. Crystal provides the following chart comparing the data presented with the data withheld:

                      DCAA
  % Post-Emergence Phytotoxicity at 1 lb./acre
                     DECEMBER
                       DATA        AFFIDAVIT D
                  --------------  --------------
Crabgrass ............  50              0
Foxtail ..............   0              0
Millet ...............  20              0
Wheat ................   0              0
Ryegrass .............   0              0
Wild Oats ............  70              0
Sudan grass ..........  20              0
Mustard ..............  20              0
Lamb's Quarter ....... 100              0
Curly Dock ...........  60              0
Pigweed .............. 100              0
Bean .................   0              0
Flax .................   0              0

Crystal also states that R & H deleted the May data ("AFFIDAVIT D") that showed two plants to have been affected by DCAA (wild carrot at 60 and indian mallow at 70). It is emphasized by Crystal that the May data shown above were submitted in the affidavit which contained the December propanil data. R & H's communication to the PTO which accompanied this affidavit stated that the "Affidavit shows comparisons of the herbicidal activity of [propanil] with [nine] compounds which are all homologs, analogs or position isomers of [propanil]." R & H continued:

It is clearly evident from the data in the Affidavit that [propanil] is truly unique in its herbicidal activity when compared to this group of closely related compounds.

Particular attention is drawn to the data in Table III, in which [DCAA], the compound relied upon most by the Examiner in the parent case, is compared with [propanil]. The 3,4-dichloroacetanilide [DCAA] is valueless as a post-emergence herbicide whereas the [propanil] is substantially 100% effective. [Emphasis in the original by R & H patent department.]

Turning to what it calls the "affidavit E misrepresentations," Crystal states that despite affidavit D, the examiner refused to withdraw his prior art rejection. He questioned whether DCIBA (3,4-dichloroisobutyranilide, another homolog) and DCMVA (3,4-dichloroalphamethyvaleranilide, another homolog) would not also be selective herbicides if applied at low rates, and requested R & H to submit appropriate data. He also questioned whether DCAA would be effective when applied at higher rates and requested that R & H submit data on DCAA at 10 lbs./acre.

Crystal submits that tests of DCAA on eight different crops and fourteen different weeds at 10 lbs./acre showed that DCAA had injury rates of 80% to 100% on six of the weeds, and that rice, wheat, ryegrass, tomatoes and barley were tolerant at this rate. Crystal asserts, however, that in preparing affidavit E R & H selected data on eight of the weeds taken three days before the end of the test period (showing 0% to 25% activity) and omitted all data showing high phytotoxicity3 of DCAA. It is also submitted that R & H did not mention data which showed selectivity on ryegrass (25%), tomatoes (0%), barley (30%), and two varieties of rice known as Bluebonnet (0%), and Gulf Rose (0%).

With respect to DCIBA and DCMVA, Crystal states that the test results at 1/8 and 1/4 lb./acre showed DCIBA had no effect on rice, and little effect at higher rates, although it was highly phytotoxic to a number of weeds at those rates. The results also showed, it is stated, that DCMVA had a very low effect on rice at these low rates. Crystal suggests that R & H could not merely omit unfavorable data because all of the DCIBA and DCMVA data were unfavorable. Crystal states that R & H changed it:

They simply took the data which showed no or very little effect of DCIBA and DCMVA on rice, and raised the numbers. Thus, 0% became 30% to 40%, 10% became 45% or 65%, 15% became 70% or 90%, and 30% and 40% became 100%. For the convenience of the Court the data submitted to the PTO compared to the actual data is presented below:

------------- -------------

lbs/acre Rose Bonnet Rose Bonnet

---------- ----- ------ ----- ------

* * * Furthermore, on crops other than rice they selected some data from preliminary readings taken January 12, and some from final readings taken January 15. In each case only the higher number was reported to the PTO, thereby portraying, an edited, worst possible case for DCIBA and DCMVA activity.

Additionally, Crystal notes that R & H omitted all data showing that carrots were tolerant to DCIBA and DCMVA.

In concluding its discussion of what R & H did, Crystal states that the examiner accepted the affidavit representations concerning DCAA, DCMVA, and DCIBA as sufficient to overcome all rejections: "Thereafter the rejections over the Fischer, the CBCC article and the Bielstein reference were withdrawn, the broadest, primary claim--claim 9--was held allowable, and an interference was declared * * *. These prior art references were never again asserted by any PTO Examiner as a basis for rejecting or limiting any claims."

In light of the above facts, Crystal argues that the "subjective but for" test of materiality has been satisfied, citing Norton v. Curtiss, 433 F.2d 779, 167 U.S.P.Q. (BNA) 532, 57 CCPA 1384 (1970). It also asserts that "Virtually identical procedural circumstances" that led the Third Circuit to affirm the trial court's determination that Monsanto's claims to propanil per se were invalid for fraud in the PTO are presented here. Monsanto Co. v. Rohm and Haas Co., 312 F. Supp. 778, 164 U.S.P.Q. (BNA) 556, and 165 U.S.P.Q. (BNA) 683 (E.D.Pa.1970), aff'd, 456 F.2d 592, 172 U.S.P.Q. (BNA) 323 (3d Cir.), cert. denied, 407 U.S. 934, 92 S. Ct. 2463, 32 L. Ed. 2d 817 (1972).

Crystal also states, and we agree, that the court was in error in saying that "what is important to" the court's inquiry into fraud is "the fact that McRae, the person who executed but did not prepare the affidavits, believed that such affidavits and the assertions contained in such affidavits to be accurate." See footnote 2 supra. Crystal emphasizes that intent is not often demonstrated by direct proof, but only by circumstantial evidence, and that

It is undisputed that the purpose of the affidavits was to convince the Examiner that propanil was unique in being a selective, post-emergent herbicide * * * [; that] McRae admits that he knew the purpose of affidavit E * * * [;] that he knew that the data he had actually observed during the rice tests confirmed that DCIBA and DCMVA were selective to rice * * * [; that he] knew the tests from which the affidavit E data were taken were run specifically for the purpose of responding to the Examiner's request for additional data * * * [; and that he] knew that R & H field test data showed that DCIBA was selective on a number of crops, and in fact was more useful on some crops * * *. Dr. Myers, R & H's registered patent agent, in the "remarks" section of an amendment accompanying affidavit E, replied to the Examiner's question regarding the herbicidal activity of DCIBA and DCMVA at lower application rates representing to the PTO that "Dr. McRae has undertaken tests in the greenhouse along this line * * * [t]he results are summarized in the TABLE on page 46 of Dr. McRae's Affidavit, and should clearly be a complete answer to the Examiner's question." Myers went on to represent to the PTO that based on this affidavit DCIBA and DCMVA were "completely unsuitable for weed control * * * there is complete lack of the remarkable selectivity shown by [propanil] * * *. [Emphasis added by Crystal.]

Crystal also states that both McRae and Myers knew that the data compared in Affidavit D were taken from different times of the year and that, thus, while accurate they were not fair. It is further argued that both reliance on the misrepresentations and injury to the public are manifest.

Turning to the effect of the renewed prosecution in 1973, after the interference, Crystal asserts that the court "erred in concluding that the February 16, 1973 representations to Examiner Thomas were effective retractions or corrections of the acts of fraud completed during 1958-63." First, Crystal argues that this conclusion is based on inadmissible evidence (memoranda of R & H's lead counsel concerning the interview) and testimony (that of R & H's lead counsel) that was heavily relied upon. Prior to trial, Crystal asserts, it sought production of the documents relied on and R & H refused to produce them on the grounds of privilege and work product. Thus,

* * * Appellants objected to their admission on the ground of (a) the previous refusal to produce, (b) the lack of opportunity to discover other attorney documents and depose witnesses concerning the subject matter of these exhibits, (c) irrelevance, since they were offered solely to support Hutz' testimony, and (d) hearsay, since they were offered to prove the truth of the statements made therein * * *. Appellants also moved to reopen discovery to inquire into any other documents which might exist relating to this subject matter * * *. However, the Trial Court overruled both the objection and the Motion * * *.

Second, Crystal argues that the information supplied to the PTO in 1973 was legally ineffective because the alleged attempted retraction of material misrepresentations occurred only after prosecution on the merits had been closed, citing Penn Yan Boats, Inc. v. Sea Lark Boats, Inc., 359 F. Supp. 948, 964-65, 175 U.S.P.Q. (BNA) 260, 271-72 (S.D.Fla.1972), aff'd, 479 F.2d 1328, 178 U.S.P.Q. (BNA) 577 (5th Cir.1973); Tipper Tie v. Hercules Fasteners, 130 F. Supp. 3, 9, 105 U.S.P.Q. (BNA) 182, 187 (D.N.J.1955), aff'd, 232 F.2d 635, 109 U.S.P.Q. (BNA) 373 (3d Cir.1956); Louis A. Grant, Inc. v. Keibler Industries, Inc., 377 F. Supp. 1069, 1082, 181 U.S.P.Q. (BNA) 1, 10-11 (N.D.Ind.1973), dismissed, 541 F.2d 284 (7th Cir.1976), for the proposition that such retractions must be made before the receiver acts in reliance on the misrepresentation. Crystal argues that prosecution on the merits was effectively closed because Examiner Thomas' decision would have been based on the written record after the interview. Thomas stated on deposition (1) that the interviews were just considered "friendly discussion," and that no review of the file was made prior to the interview; (2) that he would have had "no interest * * * at all" in the interference record; (3) that at the time of the interview the subject matter of the 1961 application had been allowed and the question of patentability was not before him; (4) that he had not been aware of Judge Masterson's "disingenuous" comment in the Monsanto case; (5) that he was not aware of DCAA data omission in the Monsanto application (which he did not examine) and was not aware of any DCAA data omissions in R & H's application, or for that matter any other data omission; and (6) that he would not have considered the raw data sheets unless they had been submitted to him for inclusion in the written record.4 Because "the only difference between [allowed] claim 9 and claim 1 of the [RH] patent is the substitution of the word "established" for "agronomic," Crystal says that

* * * the Trial Court's reference to the fact that only claim 9 had been allowed in 1963, that it had been allowed for interference purposes only, that is [sic] had been cancelled and replaced by a claim which became claim 1 of the Wilson patent, and that nothing whatsoever prevented Examiner Thomas from rejecting claims after allowance of claim 9 clearly reveals a lack of understanding of patent examining procedure, and in particular reveals misunderstanding concerning what issues the Examiner actually was considering on his own volition compared to those issues he was relying on R & H to identify. Contrary to the inferences drawn by the Trial Court from the general testimony of Examiner Thomas, the Examiner's specific testimony demonstrates that (a) he was not interested in disturbing the prior decision on patentability; (b) he was interested only in what was "important", which he relied on R & H to bring to his attention; and (c) never in his 27 years of experience with the PTO had anyone brought to his attention their prior fraud.

Third, Crystal argues that "the raw data submissions to the examiner were too complicated and were presented too quickly to have been effective retractions or corrections of the previous misrepresentations." Crystal emphasizes that about 3,771 pages were presented to Thomas at the interview, some 1,326 of these being the "500,000 series documents."5

It is also emphasized that the compounds tested are variously referred to on the raw data sheets "by code names, structural formulae, chemical formulae, chemical names, and tradenames; the test data are taken at various application rates, some of which are expressly noted, some of which are coded; the test data are taken variously as single data points and as multiple data points; and the tests were run on a variety of plants and under a variety of test conditions."

When R & H wanted the PTO to act in reliance on its data submissions, these raw data were converted into a form--Tables--in which meaningful comparison could readily be made by the PTO examiner. * * * However, in 1973 when R & H "supplied to the extent possible" the data to the PTO only the raw data were presented. * * * Furthermore, Hutz admits that he made no effort to point out to Examiner Thomas that affidavits D and E had been presented to Examiner Meros, the reasons for the filing of the affidavits, the representations that had been made concerning the data in the affidavits, the fact that affidavit E had been submitted specifically in response to Examiner Meros' request for additional data, or the effect that the affidavits had. Instead, Hutz testified that he merely showed the Examiner the data alongside the affidavits, and let the Examiner draw his own conclusions * * *. Furthermore, in his remarks filed later, purporting to summarize what took place at the interview, Hutz made no mention of comparing data with any affidavit, or of any "omitted" data or other misrepresentation.

Fourth, Crystal alleges that "the trial court erred in determining that R & H was entitled to rely on Examiner Thomas." Instead, it is argued:

The record cannot be more clear that he was relying on R & H to bring to his attention anything of importance and bring it to his attention in the written record * * *. There is simply no proposition of law, nor basis in the factual circumstances of this case that R & H was entitled to rely on Examiner Thomas to ferret out discrepancies between affidavit table data submissions made in 1961-62 and raw data presented to him once during an interview in 1973. * * * [I]t was the PTO Examiner who was entitled to rely, and who actually did rely, on R & H to fully and specifically inform him of any corrections or retractions relevant to the already-determined patentability of the claims and to do so in writing * * *.

Finally, Crystal contends that, in any event, "once fraud on the PTO has been committed, it cannot be corrected," citing Ex parte Mallard, 71 U.S.P.Q. (BNA) 294 (Comm'r Pat.1946) and Kuzel v. Heany, 1911 C.D. 148 (Comm'r Pat.1910).

In light of the above arguments, Crystal concludes its discussion of fraud by maintaining that during the post-interference interview R & H continued to conceal effectively the previous wrongs by burying the relevant data in a mountain of irrelevant information.

Dozens of issues were considered and resolved by the trial court, many involving Crystal's antitrust counterclaims under sections 1 and 2 of the Sherman Act and section 3 of the Clayton Act. Although comparatively few "findings" and only one "conclusion" of the court are devoted to the fraud allegations, R & H devoted almost forty percent of the "Argument" portion of its brief to this issue, as did Crystal.

Preliminarily, we note that nowhere does R & H deny the facts that underlie Crystal's allegation of fraud with regard to affidavits D and E. R & H variously characterizes those facts simply as "defects which they claim existed," "all deficiencies which may have existed," or in terms like "whatever those deficiencies may have been," and so on.

The R & H argument is directed solely to contentions that no fraud or inequitable conduct was proved because, ultimately, R & H "made a complete disclosure of all material facts."R & H makes the following assertions regarding the 1973 interview:

(1) it was decided that six broad areas should be discussed--new claims; Dr. Takematsu's alleged publication in January 1959; Dr. Takematsu's status as a purported independent inventor; Monsanto's attack on R & H's conception date; prior art; and the herbicidal activity of propanil and chloroanilides having a related chemical structure.6

(2) in view of the implications of Monsanto, R & H undertook to collect all herbicidal data records for related compounds available to R & H from any source.

(3) counsel prepared a list of the McRae affidavits, identifying them as "A" through "E", correlated them to the original data records, and designated the records accordingly.

(4) the knowledge that Thomas was an expert in both patent law and the herbicide art, and that he possessed extensive background information about the previous prosecution, strongly influenced the manner of data presentation at the interview.

(5) the documents were arranged in logical sequence on conference tables and, after a general discussion of the purposes of the interview, the documents were placed before him for explanation and review. Thomas was advised before and after the interview that he should ask any questions which came to mind, make notes, request any document that he wished, and that the interview was for his benefit and would be conducted in accordance with his wishes.

R & H asserts that it discussed the following matters at the interview:

(1) the prior Monsanto litigation and interference history, including the finding of fraud due to, inter alia, Monsanto's withholding of herbicidal data. Thomas was also advised of Monsanto's contention that R & H had not disclosed to the PTO all herbicidal data in its files, and of Judge Masterson's comment regarding "equally disingenuous" conduct.

(2) the herbicidal activity of propanil and related compounds by reference to (a) data from Monsanto's files, (b) original R & H laboratory notebook records, (c) publications by Monsanto, R & H, and others, (d) internal R & H memoranda, (e) Greenhouse Biological Evaluation Sheets, (f) the Monsanto decisions, and (g) the R & H applications and prosecution histories themselves.

(3) "As to the McRae affidavits, Rohm and Haas specifically compared the data reported therein with the underlying records from which the data had been taken to give the Examiner a complete and side-by-side picture of what had and had not been originally before the PTO."

(4) "The existence of some selective activity for related compounds (e.g., 3.4-DCiBA and 3,4-DCMVA, which were not in the prior art, and 3,4-DCAA which was a known compound but not recognized as a selective, post-emergence herbicide ) was shown not only by the data records discussed at the interview but also by (1) express statements of selectivity in the Wilson applications as filed, which the Examiner had read, (2) the previously filed McRae affidavits, (3) finding 49 in the Monsanto decision [sic, opinion], and (4) publications present at the interview, discussed with the Examiner and left with him for further study." [Emphasis by R & H. Unless otherwise noted, record and exhibit references omitted in all quotes from briefs.]

R & H also states that "On December 12, 1973, a final interview was held with Examiner Thomas [at which R & H] 'summarized the history of the interference and litigation, Monsanto's fraud, our prior prosecution, our prior affidavits, Fred Myers' problems and failure to present all available data (both good and bad) and our subsequent complete disclosure to Thomas'. Against this background, the Examiner was asked what additional information or data, if any, he desired, and Examiner Thomas answered, 'none'." From all of this, R & H asks us to conclude that Crystal's "fraud by burying" assertions are untenable.

With respect to Crystal's fraud arguments, R & H first notes that Crystal's emphasis on the pre-1973 prosecution totally ignores

* * * the well-established law that an applicant's conduct must be evaluated when the patent claims are finally allowed. The law does not examine an applicant's prosecution at intermediate points in a multi-year prosecution, conclude patent-defeating behavior due to temporary withholdings or misstatements, and then shut its eyes to the later presentation of corrective information prior to allowance of the patent. The pertinent inquiry is whether the Examiner received complete and accurate information during prosecution within sufficient time to act upon it before he allowed the claims.

R & H also maintains that Hutz' testimony and related documents were properly admitted as rebuttal evidence. R & H points out that Crystal did not seek production of those documents by motion to compel, nor did Crystal seek to depose any witness other than Hutz and McRae regarding the circumstances of the interviews.

R & H declares that Crystal's assertion that prosecution on the merits was closed by 1973 "is belied by the Examiner's ready rejection of claims during the renewed prosecution." Further, R & H contends that Thomas readily understood its interview disclosures. With respect to the written record:

Even assuming that the amendment following the interview did not provide adequate detail, this does not undercut or render "clearly erroneous" the Trial Court's express findings that detailed disclosure of all material facts was made. Those findings are based upon all of the evidence including a determination upon observation of the credibility of Mr. Hutz and Dr. McRae, who testified at length as to what occurred at the interview. A technical defect in the description of the interview would not change the complete disclosures made to Examiner Thomas. Moreover, the record establishes that there was no defect in the description. Defendants' complaint totally overlooks Rohm and Haas' preservation of the documents and notes thereon which, together with the summary remarks in the amendment which followed the February, 1973 interview, constitute a complete and accurate record of what was discussed. [Emphasis by R & H.]

R & H further states that "Defendants' contention that Mr. Thomas had to be expressly told about each prior rejection and how each affidavit responded to each rejection is fallacious." It contends that R & H "was entitled to rely on the reasonable belief that the Examiner had taken its disclosures into account when allowing the Wilson application," for it "placed evidence before the Examiner that it reasonably believed should have alerted the Examiner to the true state of the facts." [Emphasis by R & H.]

Responding to Crystal's argument that fraud cannot be corrected, R & H says that "The law plainly holds that the question of 'correction' is properly assessed at the end of prosecution where the bases of allowance can be accurately determined," citing a number of cases.

R & H also stresses that the district court properly determined that R & H acted in good faith, although that finding is only with regard to the alleged attempted cure. While Crystal argues that R & H did not apprise the PTO that the test data obtained for DCAA were obtained from tests run at different times and that Dr. McRae knew the tests would guarantee different results, R & H counters that it "presented all available herbicidal data to the Examiner, and this gave the Examiner a basis for reaching his own conclusion as to the alleged similarity of [DCAA] to propanil in its 'selectivity as to rice and tomatoes'." (Emphasis by R & H.)

Defendants' assertion that Rohm and Haas' representatives had "no intention" to discuss alleged deficiencies in the McRae affidavits is untenable. This accusation begins with the legally false premise that Rohm and Haas' complete disclosure to the PTO had to be coupled with some sort of statement that the facts disclosed rendered the application improper; that is not the law. The purported admissions of counsel were never made. * * * Rohm and Haas did precisely what was required, namely it submitted and explained all of the material facts and permitted the Examiner to draw on his own evaluative expertise and reach his own decision with knowledge equal to Rohm and Haas. The records of the information made available to Examiner Thomas do not reflect an intent to withhold anything from the Examiner. They confirm the intent to make a complete and honest disclosure, answer any questions, and submit any documents, all to an official of the PTO (others had been invited) with years of experience and clothed with the PTO authority to take whatever action the PTO deemed appropriate.

Finally, additional to the above, R & H argues that the criticized data were neither material nor relied on by the PTO, because "the alleged deficiencies in the original prosecution did not result in the allowance of any claim in the Wilson patent." (Emphasis by R & H.) R & H stresses that the patent did not issue until about a year after the interview and that, "During that year the application was twice rejected by Examiner Thomas, and the claims at issue were thereafter allowed. Thus, unlike most cases where fraud and inequitable conduct are alleged, the Trial Court was not compelled to speculate on the probable impact of the alleged withholdings on the ultimate issue of patentability. The Patent Office, with all of the criticized herbicidal data before it, found the data did not detract from patentability." Further,

Defendants concede that propanil possesses "truly remarkable and completely unexpected" properties when used as a selective, post-emergence herbicide in rice and that those properties "resulted in the outstanding success of propanil both in the United States and throughout most of the world." At the same time, defendants contend that propanil's herbicidal activity differs in an insignificant manner from the activity of structurally related, prior art compounds. The Trial Court rejected defendants' obviousness arguments and found that "the record demonstrates clearly that propanil's selective, post-emergence herbicidal activity is unique." We have set forth the evidence in support of this determination elsewhere in this brief. Suffice it to say, proof of this nature, i.e. commercial success, widespread publications and patents, industry acclaim, indifference to the related compounds, and the like, and not any earlier affidavit showings, was relied on to distinguish propanil during the renewed prosecution. [Emphasis by R & H.]

With respect to the prior art, R & H concludes as follows:

Defendants' contention that the Fisher [sic, Fischer], CBCC and Bielstein references render the subject matter of the Wilson patent obvious and that these references were not employed to reject claims during the renewed prosecution thereby showing no disclosures were made is equally fallacious. First, Examiner Thomas was well aware of these references and prior rejections based thereon. Second, all such prior art was presented to and considered by the PTO during the renewed prosecution. Third, defendants do not even rely on Bielstein in support of their obviousness arguments and Bielstein was not listed in defendants' notice pursuant to 35 U.S.C. Sec. 282. Fourth, the Trial Court specifically assessed both Fisher [sic, Fischer] and CBCC and found that neither reference rendered obvious the use of propanil as a selective post-emergent herbicide.

Noting that courts have not required a showing on all the elements of common law fraud, although "the evidence clearly shows all the elements * * * exist with respect to the pre-interference prosecution," Crystal responds to R & H's argument on materiality:

R & H does not deny (and there is no contrary Trial Court fact finding) that misrepresentations were made in the pre-interference prosecution, that those misrepresentations were made for the purpose of convincing the Examiner that claims should be allowed, that the Examiner relied upon those misrepresentations to allow claim 9, and that claim 9 is substantially identical to claim 1 of the patent in suit.

With respect to R & H's argument that Examiner Meros could have relied on commercial success, Crystal states that

* * * it is significant that R & H patent agent Myers emphasized in his post-interview memo the effect of the test data, but was silent on the effect of the commercial success. Also, propanil's commercial success resulted from other factors * * *. Moreover this information on commercial success is not available to rebut the prima facie case of obviousness on the broad, generic claims because the information relates solely to commercial success on rice * * *. [Emphasis by Crystal.]

Aside from this, says Crystal, "it is not necessary to prove that the comparative test data misrepresentations were the only evidence the Examiner considered in allowing claim 9; it is sufficient that it 'resulted in substantial measure' from representations made in affidavits."

Crystal, for the most part, summarizes its previous arguments regarding intent and reliance, and states that "No matter what the law on timeliness of a retraction or correction, none occurred here."

Crystal states that for Thomas to have been aware of the significance of the data, it would have been necessary for him to have in mind, during the interview, the specific deficiencies in the various affidavits, the relationship between those deficiencies and the prior art rejections by the then examiner, and the various representations that were made concerning the affidavit data:

The Trial Court did not find either that any of this was told to Examiner Thomas at the interview or that he was otherwise aware of it. R & H does not contend that any of this was told to him. The only affirmative, uncontradicted, and unbiased evidence on this point is Thomas' testimony, which is that he knew nothing of any fraud.

Crystal points out that, while Thomas' memory was dim on much of what took place at the interview, he was quite clear on what did not take place. "In the 27 years of my Patent Office career, those things [affidavit misrepresentations] have not been pointed out to me, never." Crystal further states that "Hutz testified nearly two months after Thomas, and had full opportunity to directly and affirmatively contradict Thomas' testimony. However, he did not. Instead, he testified that he merely let Thomas 'draw his own conclusions' ":

By his handling of the interview and his failure to make of record in the PTO any documents from which it could be determined that misrepresentations had been made, Hutz effectively precluded Thomas from independently making the necessary comparisons in examining the application and the new claims presented in the later amendments. At the interview itself Hutz talked about numerous items, leaving the alleged data comparison until last. By this time Thomas, understandably, "just wanted them to leave." Following the interview, the R & H data sheets were taken back to Hutz' office, and Thomas never saw them again. Then Hutz, while purporting to file a paper "summarizing" what took place at the interview, went into detail concerning all the things discussed at the interview except the alleged comparison of data with the affidavits. This is not mentioned at all. Nor is any mention made about any problem with any affidavit. [Emphasis by Crystal.]

R & H relies heavily on the fact that the Trial Court found that R & H "informed patent examiner Thomas of the omissions of herbicidal test data from the earlier filed affidavits," and that R & H "supplied to the extent possible the omitted data." If these findings mean that R & H exhibited to Thomas the test data from which the affidavits were derived, and told Thomas there were data that were not shown in the affidavits, then they are at least supported by testimony of Hutz and McRae. However, if these findings are construed to mean that Thomas learned that misrepresentations were made, they are clearly erroneous. There is no evidence to support such findings. [Emphasis by Crystal.]

Crystal makes the same argument with respect to the conclusion by the district court that "the record * * * indicates that the patent examiner was fully informed of all factors material to the patentability of the methods claimed * * *."

Crystal again argues that the court abused its discretion in admitting the Hutz testimony and documents relevant to the interviews while denying it an opportunity for discovery, and adds that it was error for the court to place controlling weight on the testimony and memoranda of R & H's trial counsel.

Finally, Crystal continues to maintain that the post-interference allowance of claims by Thomas was not based on a re-determination of the Sec. 103 issue passed on earlier by Meros, and that the 1973 "disclosure" concealed rather than revealed. R & H seeks

* * * to explain the large number of data sheet documents as being necessary to comply with their duty of candor. However, neither the Monsanto court, Mr. Dunner [Crystal's patent law expert witness], nor any other authority has stated that the duty of candor requires the submission of all data on all compounds tested, whether or not they were prior art. [Emphasis by Crystal.]

Crystal says a total of 14 pages would have sufficed to have shown the truth of matters discussed in the affidavits.

The questions presented by Crystal's allegation of fraud in the PTO are:

1. Did R & H intentionally make material misrepresentations to the PTO?

2. If it did, could that conduct have been "cured" during prosecution by subsequent disclosure to the PTO before the patent issued?

3. If R & H prosecution activities were illicit, yet could have been "cured," was R & H's 1973 interview disclosure sufficient to that end?

R & H has not denied Crystal's key factual assertions that data in Affidavit E were falsified and that the differing experimental conditions that lay behind the data comparison in Affidavit D were not revealed to the PTO. Nevertheless, it argues (1) that this activity does not amount to fraud because the affidavits were not material and because (2) the intent of the deceased patent agent who prepared the affidavits is not known.

It cannot be disputed that the purpose of the affidavits was to overcome prior-art rejections that had been entered by the examiner. R & H wished to overcome those rejections by demonstrating the allegedly unique activity of propanil when compared with a number of other closely related compounds. Findings of Fact 81 and 82 at 557 F. Supp. 768, 217 U.S.P.Q. (BNA) 536. Although the examiner did not retract his rejections on consideration of Affidavit D, the rejections were withdrawn and claim 9, the broadest claim in the application, was allowed after presentation of Affidavit E, wherein R & H advised that the data (which were falsified) "should clearly be a complete answer to the Examiner's question" and that the related compared compounds "were completely unsuitable for weed control" and showed a "complete lack of the remarkable selectivity shown by [propanil]." Thus the affidavits were material and the misrepresentations therein were material. In contrast to cases where allegations of fraud are based on the withholding of prior art, there is no room to argue that submission of false affidavits is not material.

It cannot be said that these misrepresentations to the PTO were the result of an honest mistake. Cf. Norton v. Curtiss, 433 F.2d 779, 794, 167 U.S.P.Q. (BNA) 532, 544 (CCPA 1970). While direct proof of intent to mislead is normally absent, such submissions usually will support the conclusion that the affidavit in which they were contained was the chosen instrument of an intentional scheme to deceive the PTO. In any event, proof of the actual state of mind of the applicant or persons associated with or representing an applicant is not required. "The intent element * * * may be proven by a showing of acts the natural consequences of which are presumably intended by the actor." Kansas Jack, Inc. v. Kuhn, 719 F.2d 1144 at 1151 (Fed.Cir.1983). Thus, the trial court's assertion that "what is important" to the fraud inquiry was affiant-inventor McRae's actual state of mind was legal error. Question "1" must be answered in the affirmative.

Crystal argues that "Once fraud on the PTO has been committed, it cannot be corrected," citing cases supra. R & H counters that the cases cited do not necessarily support such a broad proposition, attempting to distinguish them in a very brief discussion which concludes with the statement:

The law plainly holds that the question of "correction" is properly assessed at the end of prosecution where the bases of allowance can be accurately determined * * *. The lower court properly found that all deficiencies which may have existed in the early period of prosecution "were corrected by the complete and detailed disclosure * * * in 1973; such disclosure was made within sufficient time for the Patent Office to take whatever action it deemed necessary" * * *. Law, logic and sound public policy clearly support this ruling.

R & H does not say what the policy considerations are.

Surely, a very important policy consideration is to discourage all manner of dishonest conduct in dealing with the PTO. At the same time, the basic policy underlying the patent system is to encourage the disclosure of inventions through issuance of patents. Another policy of the system is to stimulate the investment of risk capital in the commercialization of useful patentable inventions so that the public gets some benefit from them, which may not occur in the absence of some patent protection. Clearly, we are faced with questions of both socioeconomic policy on the one hand, and morals or ethics on the other. We think we should not so emphasize either category as to forget the other.

Respecting the possibility of curing misconduct, we wish to focus as sharply as possible on the exact issue before us. In this case, we face specifically an assertion of invalidity of an issued patent, not a rejection by the PTO for "fraud" of a still-pending patent application. The assertion of invalidity, nevertheless, is based on what occurred during the prosecution of the application--more accurately, the whole sequence of copending applications. The problem, therefore, is what, if anything, could have been done in the PTO during prosecution to cure or overcome misconduct consisting of intentional misstatements of asserted material facts so as to have saved the later issued patent from the consequences of the misconduct. We are not dealing with the question of what, if anything, can be done after the patent issued to alleviate the effect of misconduct, as did one of our predecessor courts in In re Clark, 522 F.2d 623, 187 U.S.P.Q. (BNA) 209 (CCPA 1975) (unsuccessful attempt to overcome failure to inform the PTO of highly relevant prior art by reissue of patent).

Specifically, the narrow issue we now deal with is whether voluntary efforts during prosecution by or on behalf of an applicant, knowing that misrepresentations have been made to the examiner of his application, can ever alleviate its effect. Taking into account human frailty and all of the objectives of the patent system, we think it desirable to permit misdeeds to be overcome under certain limited circumstances.

The first requirement to be met by an applicant, aware of misrepresentation in the prosecution of his application and desiring to overcome it, is that he expressly advise the PTO of its existence, stating specifically wherein it resides. The second requirement is that, if the misrepresentation is of one or more facts, the PTO be advised what the actual facts are, the applicant making it clear that further examination in light thereof may be required if any PTO action has been based on the misrepresentation. Finally, on the basis of the new and factually accurate record, the applicant must establish patentability of the claimed subject matter. Considering the overall objectives of the patent system, we think it desirable that inventions meeting the statutory requirements for patentability be patented and, therefore, we also think it desirable to reserve the possibility of expiation of wrongdoing where an applicant chooses to take the necessary action on his own initiative and to take it openly. It does not suffice that one knowing of misrepresentations in an application or in its prosecution merely supplies the examiner with accurate facts without calling his attention to the untrue or misleading assertions sought to be overcome, leaving him to formulate his own conclusions. We therefore answer question "2", as posed above, in the affirmative.

The final question, therefore, is whether R & H's 1973 "disclosure" sufficed to remedy its previous misconduct. We hold that, as a matter of law, it did not.

This court and others have held that the facts which may lead to a holding of "fraud" in the PTO must be proved by "clear, unequivocal, and convincing evidence." Orthopedic Equipment Co. v. All Orthopedic Appliances, Inc., 707 F.2d 1376, 1383, 217 U.S.P.Q. (BNA) 1281, 1286 (Fed.Cir.1983). Likewise, we now hold that where intentional material misrepresentations have been made, as here, a complete "cure" must also be demonstrated by clear, unequivocal, and convincing evidence. Because all business with the PTO is to be transacted in writing and its actions must be based exclusively on the written record, 37 CFR 1.2, this question of fact should never be difficult to resolve.

R & H relies almost wholly upon the fact findings by the trial court, but those findings make no mention of the misrepresentations made by R & H. They refer, for example, to Monsanto's contention regarding "withheld" data and the Monsanto court's "disingenuous comment." Finding 114. This may account for R & H's submission of all the propanil-related herbicidal test data it could discover, i.e., its attempt to cure any Monsanto -type problems by disclosing that "there were data in its records that had not been initially reported to the PTO." Finding 119. In view of this, the trial court found that R & H told examiner Thomas of the "omissions" of test data from the affidavits and supplied to the extent possible the "omitted" data and all other data it possessed. Finding 128.

Notwithstanding the court's Finding 118 that "to the extent that such base records could be reconstructed, they were identified, keyed to the prior affidavits, compared side by side with each of the affidavits and discussed with the patent examiner," the next Finding 119 emphasizes that "the patent examiner was informed that Rohm and Haas had been unable to locate all of the test data in its records corresponding to data recited in the affidavits filed earlier with the Patent Office." Which affidavit data could not be discovered is not mentioned by the trial court.

The trial court's Finding 125 that R & H was of the "good faith" belief that examiner Thomas appreciated the information supplied and its implications is irrelevant. In this matter, the PTO had to rely on the applicant but, despite R & H's lengthy interview summaries, the record is devoid of any documentation showing that the PTO was told that any misrepresentations had been made or precisely where they had been made. Such express disclosure was necessary to meet R & H's duty, and would have alerted the PTO to the necessity of reconsidering rejections which had been withdrawn as a result of those submissions.

Considering the realities of the situation and all the items said to have been discussed with Mr. Thomas in the pivotal 2 1/2 hour interview, we regard the court's conclusion of law that he was "fully informed of all the factors material to the patentability of the methods claimed," (Conclusion of Law 33) as unrealistic when the whole record is considered. The conclusion that he was "fully informed" rests solely on the presentation to him of a mountain of largely irrelevant data from which he is presumed to have been able, with his expertise and with adequate time, to have found the critical data. It ignores the real world conditions under which examiners work.

Although Thomas' deposition testimony indicates that he had forgotten nearly all of what happened at the interview, he was quite sure of what did not happen, stating that never in his twenty-seven years of experience with the PTO had anyone ever brought their prior misconduct to his attention. Therefore, and in view of the fact that there is no written evidence of any disclosure that there had been previous misrepresentations, to say nothing of their specific correction, we hold that the asserted "cure" was insufficient as a matter of law. Question "3" is answered in the negative.

The Wilson and McRae patent No. 3,816,092 is held invalid because of fraud in the Patent and Trademark Office in its procurement. The decision of the trial court holding it valid is therefore reversed, and the case is remanded for further proceedings consistent herewith.

We have considered appellants' "Motion to Tax Printing Costs" filed June 16, 1983, and appellee's opposition thereto filed June 22, 1983. These two papers, filed before decision, deal only with the printing of the Combined Appendix of ten volumes said to contain a total of 9,307 pages, 5,743 of which were designated by appellee after appellants had initially designated the balance. There is an excessively detailed dispute over how many of the pages designated by appellee were necessary. Appellee appears not to have contributed to the cost of printing the appendix, which cost, paid by appellants, is said to have been $31,850.15. Federal Rule of Appellate Procedure (FRAP) 30(b) authorizes us to impose upon a party the cost of printing material "unnecessarily" included. FRAP Rule 39(a) provides that, in the absence of an order by the court, in the case of a reversal "costs shall be taxed against the appellee."

Having considered the foregoing, we have a clear appreciation of the impossibility of determining, within reason, exactly what was or was not necessarily included in the appendix in this extended and complex litigation. We conclude that, under all of the circumstances with which we have necessarily become familiar in deciding this case, it is fair and equitable that the parties share equally the cost of printing the Combined Appendix and that otherwise each party bear its own costs. In accordance with Rule 39(a), it is so ordered.

REVERSED AND REMANDED.

MILLER, Circuit Judge, concurring in part.

Although I agree with the majority's analysis and holding on the fraud issue, it seems appropriate to state my conclusion that, because of the unpredictability of propanil in 1957, the district court erred in granting R & H an April 4, 1957, date of conception rather than a date, concurrent with a reduction to practice, in the summer of 1957--after the May 27, 1957, date of filing of Monsanto's application, so that the Monsanto patent constitutes a 35 U.S.C. Sec. 102(e) bar to the R & H patent. Alpert v. Slatin, 305 F.2d 891, 896, 134 U.S.P.Q. (BNA) 296, 301 (CCPA 1962).