joins, dissenting.
With all due respect, I find myself in disagreement with a substantial portion of the opinion of the majority in this case. I do agree with that portion of the majority opinion which finds Krenzer’s first and third contentions in his petition completely groundless. If these were the only contentions in the petition, I would have no trouble joining my colleagues in denying it. However, I am most troubled by Krenzer’s second contention, to wit, that Stoffel et al. (Stoffel) committed fraud on the PTO and this court in withholding evidence of certain unsuccessful attempts to make compounds closely related to those recited in the count using the Zinner technique. In *1218my view, that evidence directly bears on and is highly relevant to the question of whether one skilled in the art could make the compounds recited in the count by following the Stoffel disclosure.
I should like to begin by focusing on the role that the four Zinner publications (collectively termed Zinner) have played both in and before this interference proceeding.
The specific compounds recited in the method of the count were, as far as this record shows, new compounds of the class 2-phenyl or substituted phenyl-4-lower alkyl-3,5-dioxo-l,2,4-oxadiazolidines and 2-phenyl or substituted phenyl-4-lower alkyl-3 thiono-5-oxo-l,2,4-oxadiazolidines. The Zinner publications disclose how to prepare, 2.4- disubstituted-3,5-dioxo-l,2,4-oxadiazolidines and 2,4-disubstituted-3-thiono-5-oxo-1.2.4- oxadiazolidines. Preparation is accomplished by reacting an isocyanate with a hydroxylamine to yield an N-hydroxyurea and then reacting tb j N-hydroxyurea with a chloroformate to yield the oxadiazolidine. Alternatively, Zinner discloses that the same compounds can be obtained by reacting a hydroxylamine with a chloroformate to yield a carbamate and then reacting the carbamate with an isocyanate to yield the oxadiazolidine.
Zinner was inserted into Stoffel’s specification by amendment during ex parte prosecution to overcome a § 112, first paragraph, “how-to-make” rejection by the examiner. The examiner stated that the original disclosure was deficient in disclosing the preparation of the compounds recited in', the claims. The amended specification recited:
The useful compounds are prepared by the reaction of an isocyanate with a hydroxylamine and ethvlchloroformate. Preparation of these compounds is described in detail by Zinner and Webber. Arch. Pharm. 298, 805-809. November 1965. [Underlined matter added by amendment.]
The examiner accepted the addition of Zinner to the specification as overcoming the rejection. In his next action he suggested that Stoffel copy a claim for interference purposes. This claim became the count in issue.
Zinner, however, did not discuss the specific preparation of any of the disubstituted compounds recited in the count. Indeed, there was only one mention of a disubstituted compound which, if made, would have been one recited in the count, namely, 2-phenyl - 4 - butyl - 3,5 - dioxo -1,2,4- oxadiazolidine; and, Zinner was silent on whether or not this compound was made. Nevertheless, Stoffel contended throughout the interference that Zinner, as a general teaching for making 2,4-disubstituted-3,5-dioxo-1,2,4-oxadiazolidines and 2,4-disubstituted-3-thiono-5-oxo-l,2,4-oxadiazolidines, when coupled with the Stoffel disclosure of the class of compounds recited in the count, was sufficient to show one skilled in the art how to make that class of compounds.
To this end Stoffel urged:1
Certainly anyone having the slightest skill in the art, having the above teachings [Zinner] before him and desiring to prepare a compound within the class of the instant count, is capable of selecting the proper isocyanate and hydroxylamine to prepare the compounds using the procedures set forth in the examples of Zinner et al.
In support of the proposition urged, Stoffel repeatedly referred to specific compounds actually synthesized by Zinner which, although not within the class of compounds recited in the count, were related to them in that they were either analogs or isomers thereof. Thus, in his brief before the board Stoffel specifically referred to Zinner’s procedure for the synthesis of 2-methyl-4-phenyl-3-thiono-5-oxo-l,2,4-oxadiazolidine and 2-methyl-4-ethyl-3-thiono-5-oxo-l,2,4-oxadiazolidine as almost identical to the procedure used by Stoffel to prepare 2-phenyl-4-methyl-3,5-dioxo-l,2,4-oxadiazo-*1219lidine, a compound falling within the class of compounds recited in the count. References to other specific compounds synthesized by Zinner occurs throughout the Stoffel brief before the board and the Stoffel brief before this court.
One problem with Stoffel’s contention was that Zinner itself recognized instances where compounds could not be synthesized using the disclosed Zinner reaction procedure. Thus, the board’s opinion, in addition to concluding that Zinner could not effect ring closure with 2-phenyl-4-butyl-3,5-di-oxo-1,2,4-oxadiazolidine, noted the expressed failure of Zinner to make the compound 2-cyclohexyl-4 ethyl-3,5-dioxo-l,2,4-oxadiazolidine and the compound 4-cyclo-hexyl-3,5-dioxo-l,2,4-oxadiazolidine. It was because of these failures, which cast doubt on the ability of one skilled in the art to extrapolate Zinner to make the compounds recited in the count, that the board concluded that Zinner was in fact insufficient to teach one skilled in the art to make the specific compounds of the class recited in the count.
On appeal to this court, Stoffel continued to contend that Zinner was sufficient to teach one skilled in the art how to make the compounds recited in the count. They argued the general applicability of Zinner’s reaction mechanism as well as pointing to the specific compounds actually made by Zinner, On this point, the Stoffel brief before this court listed 23 different 2,4-di-substituted - 3,5 - dioxo -1,2,4 - oxadiazolidines and 2,4-disubstituted-3-thiono-5-oxo-l,2,4-oxadiazolidines made by Zinner noting that seven of the 23 were “simple mirror isomers” of the compounds of the count. Stoffel also pointed out that the remaining 16 were analogs of the compounds recited in the count. The whole point to Stoffel’s argument, of course, was to minimize any adverse impact the abortive or unsuccessfully attempted synthesis of compounds by Zinner might have on the issue of whether Zinner’s reaction could be straightforwardly applied to make the compounds recited in the count in issue. To this end the Stoffel brief before this court stated:
As opposed to the two unsuccessful attempts of Zinner et al I, in each of which either R or R' was a cyclohexyl radical, the four [Zinner] articles teach the successful preparation of almost two dozen 1,2,4-oxadiazolidines in accordance with the synthesis methods described in detail.
In our decision we agreed with Stoffel’s contention. Although the board concluded that “no ring closure was effected” by Zinner with respect to 2-phenyl-4-butyl-3,5 dioxo-1,2,4-oxadiazolidine, we concluded that no attempt was made by Zinner to synthesize this compound. We viewed the evidence of record as establishing the fact that “no technical reason was apparent” why this compound could not be made using the Zinner reactions. We did not find Zinner’s failure to make two of .his attempted compounds to be significant and controlling. We concluded that Stoffel’s specification, together with the knowledge found in Zinner, was sufficient to enable any person skilled in the art to make and use the compounds recited in the count.
While Stoffel were making their arguments to the examiner, the board, and this court, they were in possession of evidence which, in my view, may have altered the decision of this court (and perhaps that of the examiner in withdrawing his § 112 rejection) on the “right-to-make” issue. That evidence, primarily taken from joint inventor Stoffel’s laboratory notebook, indicates that Stoffel, prior to their filing date, attempted to make various compounds of the generic formula 2,4-disubstituted-3,5-dioxo-1,2,4-oxadiazolidine and 2,4-disubstituted-3-thiono-5-oxo-l,2,4-oxadiazolidine using the Zinner reaction scheme. If Stoffel’s contentions to the PTO and this court were valid, it should have been a straightforward procedure for Stoffel to make all of the compounds attempted, since all that was necessary was the “selecting [of] the proper isocyanate and hydroxylamine to prepare the compounds using the procedures set forth in the examples of Zinner et al.”2 *1220Stoffel’s notebook entries, however, appear to refute the validity of their contention.
Although a successful synthesis was obtained by Stoffel for many of the compounds attempted, there were many failures. All of the compounds which Stoffel tried to synthesize were either analogs or isomers of the compounds recited in the count. The compounds involved, with unsuccessful synthesis indicated by an asterisk, can be tabulated as follows:3
Analogs
1. 2-phenyl-4-(3',4'-dichlorophenyl)-3,5-dioxo-l,2,4-oxadiazolidine
2. 2-phenyl-4-(4'-chlorophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
3. 2-phenyl-4-(4'-methylphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
4. 2-phenyl-4-(4'-nitrophenyl)-3,5-dioxo-
1.2.4- oxadiazolidine
5. 2-phenyl-4-(3'-chlorophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
6. 2-phenyl-4(3'-trifluoromethylphe-nyl)-3,5-dioxo-l,2,4-oxadiazolidine
7. 2-phenyl-4-(4'-methoxyphenyl)-3,5-dioxo-l,2,4-oxadiazolidine
8. 2-phenyl-4-(3'-nitrophenyl)-3,5-dioxo-1.2.4- oxadiazolidine
9. 2-phenyl-4-chyclohexyl-3,5-dioxo-l,-2.4- oxadiazolidine
10. *2-phenyl-4-(2'-chlorophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
11. *2-phenyl-4-(2',5'-dichlorophenyl)-3,5-dioxo-l,2,4-oxadiazolidine
12. *2-phenyl-4-phenyl-3,5-dioxo-l,2,4-ox-adiazolidine
13. *2-phenyl-4-(3'-methylphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
14. *2-phenyl-4-cyclohexyl-3-thiono-5-oxo-l,2,4-oxadiazolidine
15. *2-phenyl-4-(3',4'-dichlorophenyl)-3-thiono-5-oxo-l,2,4-oxadiazolidine
16. *2-phenyl-4-(3'-chlorophenyl)-3-thiono-5-oxo-l,2,4-oxadiazolidine
Isomers
1. 2-methyl-4-(4'-ethoxyphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
2. 2-methyl-4-(4'-chlorophenyl)-3-thiono-5-oxo-l,2,4-oxadiazolidine
3. 2-methyl-4-(4'-chlorophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
4. 2-methyl-4-(3',4',dichlorophenyl)-3,5-dioxo-l,2,4-oxadiazolidine
5. 2-methyl-4-(3'-chlorophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
6. 2-methyl-4-(4'-methylphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
7. 2-methyl-4-(3'-methylphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
8. 2-methyl-4-(3',4',dichlorophenyl)-3-thiono-5-oxo-l,2,4-oxadiazolidine
9. 2-methyl-4-(4'-nitrophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
10. *2-methyl-4-(2-^butyl-6-methylphe-nyl)-3-thiono-5-oxo-l,2,4-oxadiazolidine
11. *2-methyl-4-(2'-methylphenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
12. *2-methyl-4-(2'-nitrophenyl)-3,5-di-oxo-l,2,4-oxadiazolidine
13. 2-methyl-4-(3 '-nitrophenyl)-3,5-dioxo-l,2,4-oxadiazolidine
In addition, Stoffel’s records indicate that the compound 2-(3',4'-chlorophenyl)-4-me-thyl-3,5-dioxo-l,2,4-oxadiazolidine was not made because a well-known starting material necessary for the synthesis could not be made by Stoffel.
From the above, it is evident that with respect to the analogs of the compounds recited in the count, Stoffel was unable to synthesize 44 percent of those attempted. Yet complete success should have been achieved if, as contended by Stoffel, Zinner’s general reactions were reliable. With respect to the isomers a 23 percent failure rate occurred. Stoffel even noted a whole class of compounds which could not be produced using Zinner’s teachings. A progress report authored by Stoffel covering the period March-September 1964 stated:
*1221A series of l,2,4-[oxa]diazolidine-3,5-diones was prepared; see Tables I and II. Phenyl hydroxyureas containing any or-tho substituent (Cl, NO2, CHS, OCH3) could not be cyclized to the oxa[dia]zolidine dione. This “ortho effect” has been noted and reported . . . in 3 other systems in which all substituted phenyl derivatives cyclize except those substituted in the ortho position. [Bracketed material added, emphasis in original.]
Of course, the count in issue recites 1,2,4-oxadiazolidine-3,5-diones containing a phenyl ring which can be substituted in the ortho position.
In my view this evidence, solely in the possession of Stoffel, is highly relevant to the right-to-make issue. It certainly supports the board’s original skepticism of the Zinner teachings. If Zinner is not reliable, a serious question is raised whether one skilled in the art, following the Zinner procedures, could make the subject matter of the counts in issue, particularly with respect to those compounds possessing ortho substituted phenyl derivatives. I therefore cannot so lightly dismiss the evidence withheld as does the majority. I also do not agree with the majority that the general operability of Zinner was not in issue throughout this proceeding. Since not one compound falling within the class of compounds recited in the count was made by Zinner, the general operability of Zinner was in issue all along. Stoffel’s whole argument was premised on an extrapolation of the Zinner disclosure.
Although I disagree with the majority’s evaluation of the evidence submitted by Krenzer in his petition, I do not believe it necessary at this juncture to pass on whether Stoffel’s conduct in not bringing their unsuccessful experiments to the attention of this court and the PTO amounted to fraud. In my view there need not be clear and convincing evidence of fraud for this court to recall and vacate its mandate, although such evidence would unquestionably warrant such action. Hazel-Atlas Glass Co. v. Hartford-Empire Co., 322 U.S. 238, 64 997, 88 L.Ed. 1250, 61 USPQ 241 This court has the power, in the interests of justice, to recall and vacate its mandate in an appropriate case. In re Willis, 537 F.2d 513, 190 USPQ 327 (Oust. & Pat.App.1976). As stated in Willis and In re Jones, 542 F.2d 65, 191 USPQ 249 (Oust. & Pat.App.1976), that power should be exercised sparingly and only upon a showing of good cause. Some evidence of fraud is good cause. Kravig v. Henderson, 393 F.2d 1017, 55 CCPA 1182, 157 USPQ 564 (1968). In Kravig we were presented with an argument by one of the parties to an interference that the other party had committed fraud by submitting false evidence in the interference. At the time of the allegation, our mandate had issued remanding the case to the Board of Patent Interferences for action on another matter. We found that the allegations and supporting depositions raised “questions of sufficient import to warrant enlarging the scope of our remand.”4 Presumably, if we found those allegations completely meritless we would have merely denied the motion. Cf. Smith v. Hayward, 193 F.2d 198, 39 CCPA 748, 92 USPQ 126 (1951). We did not pass on the merits of the allegation in Kravig because we preferred, as a court of review, not to decide any matters relating to priority, or the evidence relating thereto, which had not first been considered by the board. Kravig v. Henderson, 393 F.2d at 1018, 55 CCPA at 1184, 157 USPQ at 565. I believe a similar rationale should apply here. S.Ct. (1944).
The remedy ultimately sought by Krenzer is the striking of the Stoffel application from the PTO files under 37 CFR 1.56. Indeed, this petition is before us only because of an interlocutory ruling by the Commissioner on Krenzer’s petition to the Commissioner to strike the Stoffel application. A decision by this court on the substantive fraud issue would, for all intents and purposes, dictate the outcome of Krenzer’s petition to strike. However, a decision to strike a party’s application should be made “in the first instance” by the Commissioner. Vandenberg v. Reynolds, 242 F.2d *1222761, 44 CCPA 873, 113 USPQ 275 (1957). Accordingly, I believe we should follow the procedure in Kravig, i. e., remand to the board, thereby allowing for a decision of the Commissioner on Krenzer’s petition to strike.
I am also troubled by another aspect of this case. As noted earlier, I find the withheld evidence to be highly relevant to the right-to-make issue. This is true regardless of whether the withholding of it amounts to fraud. Of course, if Krenzer ultimately prevails in his petition to the Commissioner, the right-to-make issue becomes moot because the Stoffel application will be stricken from the files. However, if Krenzer does not prevail, the right-to-make issue would stand without any tribunal having decided that issue in view of all of the evidence now available.
In view of the above considerations and the strong public policy favoring the issuance of a valid patent to the proper party, I believe that justice would be best served in this case by granting the Krenzer petition to recall and vacate our mandate to the extent that we reassume jurisdiction over the case and remand the same to the board. The board could then reconsider the right-to-make issue in light of the evidence of the Stoffel unsuccessful experiments. The board could also consider the fraud issue (presumably in conjunction with the present petition before the Commissioner), since that issue is ancillary to priority. Norton v. Curtiss, 433 F.2d 779, 57 CCPA 1384, 167 USPQ 532 (1970). In this manner a full and fair hearing of all matters raised by the second contention in the Krenzer petition can be obtained.
. Stoffel opposition to Krenzer’s motion to dissolve interference No. 96,419 with respect to the party Stoffel, filed April 21, 1969.
. Note 1, supra.
.This tabulation is from a summary compiled by Krenzer. Stoffel has not objected to its accuracy.
. Order entered Feb. 6, 1967 in Patent Appeal No. 7862.