with whom joined MacKINNON, Circuit Judge, dissenting:
This litigation is an outgrowth of a Federal Trade Commission (FTC) investigation into the reporting of natural gas reserves by natural gas producers in Southern Louisiana:1 The issues concern the District Court’s action in modifying and enforcing the Trade Commission’s subpoenas.
We see the issues in this case quite differently from the majority. Fundamentally, our colleagues state the issues and write as if the court were reviewing directly the action of the Trade Commission in issuing these subpoenas. We are not. We are reviewing the action of a District Court in modifying and enforcing those subpoenas. It is the validity and reasonableness of the District Court’s action which we judge. What we review and the standards we employ are determined by that?
The District Court modified the subpoenas in two principal ways: (A) limiting the material to be produced to any and all material containing data on proved reserves and their reporting; and (B) limiting the use of the data to purposes other than establishing the accuracy of the proved reserve figures, such estimates having been *423established for the same time period by three separate investigations of the Federal Power Commission.
In so modifying and enforcing the subpoenas the District Court acted on three separate and independent grounds: (1) relevance of the data sought to the purpose of the Trade Commission investigation, the reporting of proved gas reserves (applicable to the content or production limitation); (2) burdensomeness of producing material already twice or thrice furnished before (applicable to both the content and use limitations); (3) administrative collateral estoppel as to one finding of fact, the accuracy of the proved reserve figures, already made by the agency primarily responsible, the Federal Power Commission (applicable to the use limitation).
The first two grounds are essentially factual, within the sound discretion of the District Court, and must be respected unless clearly erroneous. The third ground is a question of law, initially for determination by the District Court.
I. THE FACTS, THE ISSUES, AND THE ARGUMENTS
A.
The American Gas Association (AGA) is a trade association of producers, distributors, and marketers of natural gas. Through its Committee on Natural Gas Reserves, the AGA has since 1946 been providing the industry, the Government, and the general public with annual estimates of the proved natural gas and natural gas liquid reserves of the United States.2
In May 1969 when the AGA reported its 1968 figures, they indicated a decline in proved reserves nationally, the first such decline ever reported. Before the year 1969 was out, this and other information reaching the Federal Power Commission (FPC) prompted a reopening of its just-concluded Southern Louisiana Area Rate Proceeding.
In late 1970 the Federal Trade Commission began an investigation into the reporting of natural gas reserves in Southern Louisiana. On 3 June 1971 the investigation achieved more formal status when the Trade Commission issued a resolution authorizing the use of compulsory process in furtherance of a nonpublic investigation. From the beginning of the investigation the AGA cooperated with the Trade Commission on a voluntary basis. As a result the Commission was able to obtain the field-by-field estimates of proved reserves made by *424each Southern Louisiana subcommittee member for the years 1966 through 1970.3 The Commission also obtained reserve information from Form 15 reports filed with the Federal Power Commission. These reports are filed by interstate natural gas pipelines and list recoverable saleable gas reserves committed to, collected by, or held by reporting pipelines.
Approximately one year after beginning its investigation, on 24 November 1971, the Commission staff issued identical administrative subpoenas duces tecum to eleven natural gas producers. All eleven producers moved to quash the subpoenas. Following the Commission’s denial of the motions, the Trade Commission’s staff, after negotiations with the gas producers, offered additional safeguards for confidentiality of the information to be supplied. As a result two producers agreed to comply fully with the subpoenas and one agreed to comply in part. Soon after petitions for enforcement were filed in the District Court, one more firm agreed to comply with the subpoena.
Petitions for enforcement of the remaining subpoenas were filed in the District Court on 4 June 1973. The two orders here under review were filed on 22 March 1974. One order covered the subpoenas issued to appellees Texaco, Inc., Standard Oil Co. (Indiana), Shell Oil Co., Exxon Corp., Standard Oil Co. of California, and Mobil Oil Corp. (hereinafter the “Six-Company Order”) and enforced specifications A through F and K through L in full.
During the hearings held on 30 July and 13 December 1973, the District Court found the subpoenas to be overly broad and unduly burdensome because they sought to duplicate investigations of the Federal Power Commission which had already resulted in a finding that AGA proved reserve estimates were valid and accurate. As a result, specifications G through I were modified so that raw field data, bid calculation data, and bid calculation files need not be produced.4 However, all documents, wher*425ever located, containing or underlying proved reserve estimates in the Southern Louisiana area are to be produced under the District Court’s order.
The court, in an attempt to make the subpoenas less burdensome, limited production of the documents called for by specifications G through I to a random sample of 100 out of the approximately 225 relevant fields and to the years 1969, 1970, and 1971. Specifications J, K, and L were also modified so that only documents relating to proved natural gas reserve estimates in the offshore Southern Louisiana area need be submitted. The court, however, enforced the subpoena as regards any documents prepared between 1966 and 1971, inclusive, “which were exchanged between or among, or constitute, contain or refer to any agreement, arrangement or communication between or among, respondents or others, including the American Gas Association.” Finally, the subpoenas were modified so that additional protections were afforded to confidential information and the producers were accorded the option of producing records for inspection where they were stored.
B.
Because the producers have not cross-appealed, the issues before us relate solely to those limiting modifications fashioned by the District Court which were not subsequently accepted by the Trade Commission and the producers in their field stipulations. Those issues, in the order we will deal with them, are as follows:
1. Did the District Court abuse its discretion under the relevance standard of United States v. Morton Salt Co.,5 in refusing to order the production of documents which did not relate to estimates of proved reserves? In other words, was the District Court’s finding that documents related to non proved reserve estimates (e. g., raw field data and bid files) were not “reasonably relevant” to the investigation described in the FTC’s resolution of 3 June 1971 a clearly erroneous finding? (a relevance issue — an essentially factual determination for the District Court)6
2. Considering the purpose of the Power Commission’s investigation and the FPC’s previous factual determination that AGA proved reserve data was reasonably reliable for ratemaking purposes,7 would the effort and expense involved in producing documents so that the FTC could determine for itself the validity and accuracy of all natural gas reserve estimates (proved or unproved) have imposed an unfair and unreasonable burden on the producers? If so, did the District Court abuse its discretion in (1) *426limiting enforcement of the subpoenas to documents containing or underlying proved reserve estimates and (2) restricting use of the documents to an investigation of an alleged conspiracy in the reporting of natural gas proved reserve estimates? (a burdensomeness issue —a determination within the sound discretion of the District Court)
3. Was the District Court in error in refusing to permit the Trade Commission to use the documents called for by specifications G through I “to investigate or determine the amount of proved natural gas reserves,” i. e., in ordering this production “for the sole purpose of permitting the Trade Commission to investigate whether there is a conspiracy in the reporting of natural gas proved reserve estimates. . ”8 (the administrative collateral estoppel issue — a question of law initially for the District Court)
4. Did the District Court abuse its discretion in:
a. limiting the production of documents called for under specifications G through I to a random sample of fields in which the producers “had an ownership interest as of the date reports were submitted by the Southern Louisiana Subcommittee of the [AGA]” for the years 1969,1970, and 1971?9
b. limiting the production of documents called for under specifications J and K to “. . . documents relating to . estimates . . . which were included in reports for 1971 by the Southern Louisiana Subcommittee of the [AGA]” and “. prepared or dated during the years 1966 through 1971, inclusive, [and] which were exchanged between or among, or constitute, contain or refer to any agreement, arrangement or communication between or among, [the producers] or others, including the [AGA]”?10
c. permitting documents to be produced for inspection at their situs?11
d. attaching conditions to disclosure to insure confidentiality?12
e. allowing the producers “. . 180 days after the date upon which they are advised of the sample fields selected . . ” in which to comply with the subpoenas? 13
5.Was the District Court in error, under the relevance standard of United States v. Morton Salt Co.,14 in affording Superior Oil Co. differing treatment?15 (a relevance issue — an essentially factual determination for the District Court).
C.
Before embarking on our analysis of these issues we will briefly summarize the arguments of the parties. The producers have never quarreled with the power or the right of the Federal Trade Commission to investigate the natural gas industry to uncover violations of the antitrust laws or unfair trade practices. However, they do argue that there can be no possible reason for wanting documents that do not relate to proved reserve estimates simply because it is only proved reserve estimates that are taken into account by the Federal Power Commission in setting area rate ceilings and it is only proved reserve estimates that are reported to the AGA and then, through the AGA, to the public. Alternatively, the producers argue that the Federal Power Commission, the only agency possessing the requisite expertise, has determined that AGA proved reserve data are accurate. Therefore, the FTC is collaterally estopped from relitigating this one issue of fact.
The Trade Commission, on the other hand, points out that there is no provision in the Federal Trade Commission Act (FTCA) excepting gas producers from the *427coverage of the Act, as there is for banks and certain common carriers, and that therefore jurisdiction to investigate exists. They argue that such jurisdiction is broad, “reaching not only existing violations of . [the Sherman and Clayton Acts], but trade practices which conflict with their basic policies.”16 The Trade Commission goes on to argue that, as a factual matter, its investigation does not duplicate studies made by the Federal Power Commission and that, even if it did, collateral estoppel would be inapplicable because its purpose in determining the accuracy of reserve data is different from the Power Commission’s purpose in determining accuracy.
Although the producers are correct in arguing that data relating to any reserves other than proved reserves would not be “reasonably relevant”17 to the investigation defined by the Trade Commission’s resolution of 3 June 1971 (and that this part of the District Court’s order could properly be supported on relevance grounds alone), it is not clear from the transcript of the 13 December 1973 hearing whether the District Court reached the issue of relevance in regard to all documents subpoenaed. Since the Trade Commission had subpoenaed all reserve records so that it could determine independently the total gas reserves of the area, and since the Power Commission had already determined that the AGA figures were accurate, the court apparently was of the view that the Trade Commission was collaterally estopped from forcing the producers to relitigate this matter, and that therefore it was unnecessary to reach the issue of relevance for all documents. In any event, it seems clear that the District Court also felt that it would be unduly burdensome to permit yet another plenary investigation (for the third time in two years) of natural gas reserves.18
Thus, there are three independent grounds for affirming the District Court’s most important limiting modifications:19 relevance (issue 1 supra), burdensomeness (issue 2 supra), and administrative collateral estoppel (issue 3 supra). For the sake of clear analysis, we will attempt to keep these three grounds separate as we now proceed to a discussion of the issues.20
II. THE FEDERAL POWER COMMISSION INVESTIGATIONS
First, we turn to the District Court’s determination that, for the years covered by the Trade Commission’s subpoena, the Federal Power Commission had already (1) considered and ruled upon the validity and accuracy of natural gas reserve data, and (2) determined natural gas reserves.21 The *428majority opinion implicitly agrees with this District Court finding, but some development of this point is necessary, as it forms the background of the whole case.
Although the Power Commission began area rate proceedings in 1961 for the Southern Louisiana area, it was not until 1968 that the Commission rendered a final decision, which in turn was modified in early 1969. Even before oral argument could be heard before the Fifth Circuit on petitions for review sought by producers, pipeline companies, and consumers, the FPC had instituted new proceedings (So La II) “to reconsider all major actions it had taken” in the prior proceedings.22 As the FPC stated in its order instituting So La II, Phase I of its new proceeding solicited “. . . evidence with respect to the adequacy of gas supply and adequacy of service to consumers, the demand for gas, the gas shortage, if any, the effect of price on gas supply and demand, and other relevant economic evidence. . . . ”23
The FPC was concerned in So La II with complaints that adequate supplies of natural gas were not being produced and would not be produced under the recently ordered area rate ceilings. More important for present purposes, the FPC during So La II was presented with the argument by Municipal Distributors Group (MDG), an intervenor which represented the interests of municipal and other publicly owned gas distribution systems, that the supply shortage was more apparent than real. It was argued that “the sharp decline in the supply picture in 1968-69 is revealed by the above record evidence to be caused largely by revisions in the estimates” of proved reserves reported by the AGA.24 Such a contention went to the heart of the Power Commission investigation. If it were true that the decline in reserves was a matter of definition and not of economics, the concern of the FPC that new area rates might be required to encourage production would be obviated.
In its final opinion in So La II the FPC discussed testimony which was used by MDG to impeach AGA data. The testimony outlined several methods by which producers could withhold reserves from the AGA. Discussed also were MDG’s arguments relating to discrepancies between figures gathered by the FPC and those submitted by the AGA. Additionally, the FPC referred in some detail to the testimony and exhibits supporting the reliability of *429AGA data.25 As a result, it reached the following conclusions:
AGA and Form 15 data show similar trends of reserves and reserve-to-production (R/P) ratios. AGA data indicates a steady decline in the national R/P ratio from 19 in 1963 to 13 in 1969. Form 15 data indicates a similar decline in the national R/P ratio from 20 in 1963 to 14 in 1969. The American Gas Association reserve data is not impeached, in our opinion, in this discrepancy.26
For the reasons stated herein, we find the AGA reserve data is reasonably reliable for the purposes used herein. Accordingly, and because petitioner has not raised any new evidence, we deny the petition to reopen.27
In other words, while in 1963 AGA data started with nineteen years of reserves and the Form 15 data started with twenty years of reserves, by 1969 each calculation had dropped 6 years of reserves off its proved reserve figure, so the two calculations were comparable. Other substantial corroborating evidence, including the fact and extent of pipeline curtailments of gas service, was received and considered by the FPC during So La II28 Thus, AGA data was shown to be reliable.
The underreporting issue was again raised on review before the Fifth Circuit and received the following extensive rebuttal in a footnote:
Standing virtually alone against the National (and record) judgment of a near energy calamity, the American Public Gas Association (APGA) contends that the current critical shortage of natural gas is but a pretextual “cry of wolf” calculated to mislead FPC into establishing artificially high rates in the producers’ behalf. APGA would have us believe that the energy crisis is a mirage — indeed, a hoax! APGA claims that “there appear to be adequate supplies of gas in the domestic United States to satisfy the projected demands of U.S. consumers well into the 21st Century.” APGA Supp. Brf. at 5 n. 9. But to talk *430of “Supplies” of gas is a misleading oversimplification. Obviously, the gas is not presently available. At most, if there is appropriate exploration, the demonstrable reserves may be exploited to meet the needs. Given a system which depends on private stewardship and marshalling of natural resources, there is a supply shortage if the producers do not produce. PPC has the statutory duty, not only to guard the consumers against super-profits reaped from artificially inflated rates, but also to protect consumer interests by making sure that the rate schedule is high enough to elicit an adequate supply. It is a delicate balancing test. PPC must fix its course to attain the utopian “optimum” rate schedule. Given the current shortage of available supply FPC must swing the pendulum towards the incentive, supply-eliciting side of rates. And so it has done.29
In addition to the examination undertaken in relation to the So La II proceedings, the Power Commission undertook in early 1971, at the direction of Congress, a National Gas Survey, a portion of which was the National Gas Reserves Study (NGRS). The NGRS was a completely independent survey of reserves and did not rely on AGA figures at any point. However, in the final staff report of the NGRS (issued May 1973) a comparison was made with AGA figures:
The NGRS estimate is lower than the estimate by A.G.A.; however, the difference is less than 10 percent. The difference of 23.5 Tcf between the estimate of the non-associated and associated gas reserves for the 6,358 entries in the reported fields category (a) is the primary difference between the total estimates. The gas reserves in the “A.G.A. omitted fields” are a relatively insignificant part in the total NGRS estimate, and it seems evident that the 62 entries in the “omitted” category (b) are small fields. The two dissolved gas estimates differ by 1.7 Tcf or by about 5 percent.30
Thus, this independent study concluded that, if anything, AGA proved reserve estimates overstated the true picture.
In these Power Commission investigations the FPC necessarily analyzed and checked the AGA reserve data in So La I and So La II, the FPC considered the data offered by the Municipal Distributors Group to impeach the AGA data and found that this data (derived from the regular Form 15 reports) confirmed rather than contradicted the AGA data, all of which was reviewed by the Fifth Circuit and the Supreme Court. Then separately and in addition to the above, the Power Commission made, in 1971-73 at the request of Congress, a completely independent analysis, not relying on AGA data in any way, called the National Gas Reserves Study, which concluded that proved gas reserves were actually somewhat less than calculations based on the AGA figures showed.
While the analyses of the FPC in the So La I and So La II proceedings were made for ratemaking purposes, the National Gas Reserves Study was not. And in both instances the Power Commission made findings of fact on the precise issue (accuracy of reported reserves) which the Trade Commission now seeks to relitigate. Hence the District Court made the factual determination that the Power Commission had already (1) considered and ruled upon the validity and accuracy of natural gas reserve estimates and (2) determined natural gas reserves. The District Court’s finding of fact should not be confused with the question of law which necessarily followed, whether it was appropriate to give this finding of fact collateral estoppel effect, which we explore in section VI, infra.
*431III. COURT ENFORCEMENT OF ADMINISTRATIVE SUBPOENAS
A. Court Enforcement of Administrative Subpoenas
The Trade Commission and a majority of this court apparently would have us proceed as if there were on appeal here an order of the Trade Commission, entitled to deference as an exercise of that agency’s expertise.
To the contrary, this is an appeal from orders of the District Court enforcing the Commission’s subpoena with some limiting modifications. Accordingly, it is not the views of the Commission staff which must be accorded deference, but the determinations of the District Court which must be upheld unless clearly erroneous or an abuse of discretion.
There are limits to the subpoena power of an administrative a'gency, and the duty and authority to enforce those limits rests in our federal district courts. As the Ninth Circuit has stated,
There is no rule requiring a court to act against conscience. The proceeding [judicial enforcement of administrative subpoenas] is equitable in character. Equitable considerations should prevail. There is no power to compel a court to rubber-stamp action of an administrative agency simply because the latter demands such action.31
By arguing as if it had been denied the ability to proceed with its investigation, and by arguing that its subpoena must be enforced unless “the evidence sought is plainly irrelevant to any purpose within [its] statutory authority . . .,”32 the Trade Commission demonstrates that its real purpose is to strip the federal judiciary of any discretion in subpoena enforcement proceedings. Ironically, the District Court’s order here enforces the FTC’s subpoena as to the great preponderance of the documents sought, and even specifically preserves the FTC’s ability to obtain further documents and information.33 Consequently, the majority’s failure to sustain the District Court in its reasonable limitations of the FTC’s subpoena may have exactly the precedential effect desired by the Trade Commission. By stripping the District Court, which has reviewed countless submissions and has held two full days of hearings, of any discretion in enforcing this FTC subpoena, our colleagues have totally undermined the concept of judicial enforcement of administrative subpoenas.
The Supreme Court has announced repeatedly the principles which should guide a district court in deciding the lawfulness of investigative subpoenas issued by administrative agencies:
The gist of the protection is in the requirement, expressed in terms, that the disclosure sought shall not be unreasonable. .
[T]he requirement of reasonableness . comes down to specification of the documents to be produced adequate, but not excessive, for the purposes of the relevant inquiry. Necessarily, . this cannot be reduced to formula; for relevancy and adequacy or excess in the breadth of the subpoena are matters variable in relation to the nature, purposes, and scope of the inquiry.34
* * * * * *
It is now settled that, when an administrative agency subpoenas corporate books or records, the Fourth Amendment requires that the subpoena be sufficiently limited in scope, relevant in purpose, and specific in directive so that compliance will not be unreasonably burdensome.35
*432More specifically, in United States v. Morton Salt Co., in the context of a Trade Commission investigatory proceeding, the Supreme Court wrote,
Of course a governmental investigation into corporate matters may be of such a sweeping nature and so unrelated to the matter properly under inquiry as to exceed the investigatory power. But it is sufficient if the inquiry is within the authority of the agency, the demand is not too indefinite and the information sought is reasonably relevant.36
In addition, the Court has made clear that these “issues of . relevancy of the materials sought, and breadth of the demand are neither minor nor ministerial matters.”37 They are, instead, matters which require the exercise of sound discretion on the part of a district court and matters which raise questions essentially factual in nature. Accordingly, we, at the appellate level, must defer to the determinations of the District Court unless we find those determinations clearly erroneous or an abuse of discretion.38
In a recent opinion affirming the order of a district court enforcing another Trade Commission subpoena, this court correctly stated the standard of appellate review applicable to a district court’s determinations of relevance:
A finding by the district court that documents are relevant and necessary to an inquiry by the FTC is essentially factual in nature and cannot be overturned unless the district court’s finding is clearly erroneous.39
We point out in Part IV.A., infra, how the majority has ignored and evaded this established standard — and any other appellate standard — of judicial review. Similarly, in NLRB v. Northern Trust Co., the Seventh Circuit has written,
We agree that the courts are not mere rubber stamps and that subpoenas are in the District Court’s discretion, but we fail to see how that helps appellants. It seems to us to have the opposite effect, for here the court after a very thorough and careful inquiry granted the Board’s request for enforcement. Thus appellants are in the unenviable position of sustaining the great burden of showing an abuse of discretion. They have failed to do so. The facts of the case insofar as the Board knew them at the time of the application and the materiality of the evidence sought were brought out before the trial judge, who concluded that the documents which the Board wanted to be produced were relevant. We are not willing to disturb his holding, because we do not think any abuse of discretion was involved.40
*433Thus, we are left to decide in the instant case whether the District Court abused its discretion, under the reasonably relevant standard of United States v. Morton Salt Co., by refusing to enforce those portions of the subpoenas that called for documents which did not relate to estimates of proved reserves. In other words, was the District Court’s finding that only documents related to proved reserves were “reasonably relevant” to the FTC’s investigation a clearly erroneous finding? We conclude that it was not and that the District Court acted well within the bounds of its discretion.
IV. RELEVANCE
A. Purpose of Inquiry, Test of Relevancy, and Standard of Judicial Review
The requirement that documents sought by an agency must be “reasonably relevant’ to a lawful investigative purpose constitutes the fundamental limitation on administrative subpoena power. The most astonishing point about the majority opinion’s discussion of “relevance” is that nowhere therein does it define the FTC’s “purpose” in the investigation, nor does it purport to employ a standard of appellate review which is legally definable.
“Relevance” simply cannot be determined in the absence of defined “purpose,” whether that purpose be as sharply defined as in a criminal trial, less precisely delineated as in a civil proceeding, or more generally defined as in a grand jury inquiry or in an administrative agency investigation as here.41 In all situations, purpose in some degree must be defined in order that notice be given and relevance thereafter may be assessed.
This is the reason for the Administrative Procedure Act requirement42 that an agency state the purpose of its proposed investigation at the very outset, to give notice to objects of the investigation and all other interested parties, and to set a standard by which the relevance of the agency’s demands for tangible evidence and testimony, as well as other actions, may be judged. Without such a requirement for a defined and announced purpose the Federal Trade Commission — or any other of our regulatory agencies — would speedily metamorphosize into the “Roving Commission.” In seeking to ascertain the actual purpose of the FTC’s investigation, the District Court here understandably looked beyond the FTC’s authorizing resolution, for, as written, this broad resolution could have authorized a subpoena covering virtually any information relating to appellees’ oil and gas business. And the majority here never came any closer to defining the purpose and scope of the FTC inquiry than this.
Where the agency’s statement of purpose is not sufficient to permit a relevance determination, the courts may refuse enforcement.43 Or, even if the statement is somewhat obscure, the court may seek to clarify and determine for itself the agency’s actual purpose, as the District Court did here.44 *434As written, the resolution certainly furnished no meaningful criteria for measuring the relevance of the data sought.45 Forced to seek clarification from other sources the District Court asked Trade Commission counsel to define the purpose of the Agency’s investigation. He responded as follows:
[FTC COUNSEL]: . . . [W]hat we are investigating is possible collusive conduct by the natural gas producers in the reporting of these reserves.
[FTC COUNSEL]: . . . What we want to find out is whether or not in reporting natural gas reserves there has been collusive conduct in the way these estimates are prepared.
THE COURT: Reporting them to whom.
[FTC COUNSEL]: All right. Reporting them to the American Gas Association, because the American Gas Association data is the only available published data on these reserves.46
In addition, the FTC previously had made the following significant representation to Congress:
The subpoena [issued on 24 November 1973] focused on Offshore South Louisiana and revolved around one central premise which is that natural gas producers, for a variety of reasons, make estimates of proven gas reserves.
The theory of the investigation is to compare the reserve estimates used by the companies in their internal purposes with the estimates submitted to the American Gas Association that were relied on by the Federal Power Commission. In other words the investigation was not attempting to make new estimates of the proved reserves in South Louisiana but was instead attempting to get at reserve estimates already prepared.47
It seems clear from these sources that the central purpose of the FTC investigation was to compare proved reserve estimates kept by the companies with the proved reserve estimates reported to the AGA. Moreover, since only proved reserves are reported to the AGA and since AGA data is the only published reserve data available, it followed that speculative estimates of possible, potential, or other unproved reserves (e. *435g., the reserve estimates typically found in raw field data or bid files) could have no relevance to the FTC’s investigation into whether in the reporting of natural gas reserves there had been collusive conduct in the way these proved reserve figures were prepared.
The District Court was clearly of this opinion, and because this factual determination was well-founded in both logic and evidence, we do not understand how a majority of this court could find it to be clearly erroneous. Likewise, we do not understand how the District Court could have abused its discretion in limiting enforcement of the subpoenas to documents containing or underlying proved reserve estimates. By the Trade Commission’s own admissions, before the District Court, in briefs to the court,48 and in representations to Congress, these were the only documents relevant to the Trade Commission’s investigation.49
Of course, the majority does not determine the District Court’s finding to be “clearly erroneous,” nor did it find an “abuse of discretion.” By these accepted tests the trial court action could not possibly be upset. What the majority does do is to brush aside the accepted legal standards of appellate review. In so doing it wriggles and twists to avoid applying the proper standard of judicial review by claiming “[h]ere, however, the district court’s relevance determinations rested upon — and, indeed, were inseparable from — its view of the applicable law with respect to the proper scope of the FTC’s investigation.”50 After the factual inquiry made by the District Court of FTC counsel, after the FTC’s rep-*436reservations to Congress, it is clear that the trial court’s determination of the purpose of the FTC investigation was made as a factual matter, not a legal determination. (“Tell me what you are trying to do. I don’t understand the resolution.”)51 The FTC resolution, the FTC counsel’s answers to questions, the FTC statements to Congress (notes 47 and 49, supra), all are evidentiary facts as to the Commission’s stated purpose of its inquiry.
A most basic flaw in the majority’s analysis concerns their characterization of this inquiry by the trial court into the purpose of the FTC’s investigation. A close examination of this characterization is in order, since the majority relies on the characterization as the foundation for its belief that it should completely substitute its judgment for that of the District Court in this case. The majority states that the District Court’s “relevance determinations rested upon — and, indeed, were inseparable from — its view of the applicable law with respect to the proper scope of the FTC’s investigation.”52 At another point the majority emphasizes that the trial court used “its own conception of the proper scope of the FTC’s investigation”53 in making its decisions as to relevance. In other words, the majority believes that the District Court imposed on the parties its own view of the proper scope of the investigation and, in effect, decided that the FTCA allowed the FTC to investigate only proved reserves of natural gas.
The majority is clearly wrong in its artificial attempt to cast the actions of the trial court as a determination of law. Rather, the trial judge conducted a factual inquiry in order to determine what, in actuality, the FTC was attempting to do in this case. Indeed, at the hearing on 13 December 1973, the District Court, in making its relevance determination as to the bid files, stated that the files were not relevant “to the investigation or inquiry that you tell me you are making. . . ,”54 The trial court recognized that the breadth of the investigation was primarily for the investigators (the FTC) to determine; in this case, the court was merely trying to develop the necessary factual foundation to make a ruling on relevance. The District Court was clearly not trying to mold the scope of the investigation, but rather to define it; this was not an imposition of the court’s view but an inquiry made of those conducting the investigation. In making its relevance determinations the trial court proceeded on the basis of the FTC counsel’s representations to the court and not on its own view of the FTC’s statutory authority in this area.
Not only does the majority repeatedly offer an inaccurate characterization of the trial court’s factual inquiry into the FTC’s purpose, but they further confuse the issue by failing to be consistent even in their own misconceptions. The majority at one point assert that the trial court’s erroneous definition of purpose resulted from its acceptance of the gas producer’s concept of purpose in this case.55 The majority thus switches its position, at various points asserting that the District Court highhandedly shaped the investigation according to its own view of the law, and at another stating that the court totally accepted the view of the private parties in this case. Which is it? Imposition or total acceptance? The truth is that neither characterization is accurate; rather, the trial court conducted a factual inquiry directed to the FTC in order to lay the analytical foundation for the relevance determinations. There is such a basic difference between shaping an inquiry and defining it; the sequence of events as revealed in the record in this case simply does *437not lend itself to inclusion in the former category.
Nor is there any place in the record where the District Judge “speculate[d] about the possible charges that might be included in a future complaint, and then . determine[d] the relevance of the subpoena requests by reference to those hypothetical charges.”56 Quite to the contrary, the District Judge acted on the basis that “[t]he relevance of the material sought by the FTC must be measured against the scope and purpose of the FTC’s investigation, as set forth in the Commission’s resolution”57 — the standard as defined by the majority, supplemented by factual inquiries of FTC counsel and FTC statements to Congress.
The majority’s mistaken speculation that the District Court equated its definition of purpose with “a particular theory of violation”58 is grossly deficient in several respects. First, there is no evidence that the trial court required a “narrowly focused theory of a possible future case”59 as the only acceptable definition of purpose. Indeed, the District Court did not require it; we in the dissent do not suggest it; the precedents do not mention it; and logic, reason, and clear thinking do not dictate such a narrow focus. What the District Court, we in the dissent, the precedents, and logic do suggest is that some refinement of the FTC’s extraordinarily broad investigatory resolution is in order. What the majority has done in suggesting that “a particular theory of violation” is the only acceptable refinement of purpose is to set up a straw man against which to argue its case for complete substitution of judgment on factual matters. Straw men are easy to knock down; it is far more difficult to provide a response to the eminently sensible proposition that some refinement of purpose is needed if the court is to have any role in considering the rights of all the parties who stand before it in a subpoena enforcement proceeding.
Returning to the question of the proper standard of review of the trial court’s relevance determination, after the majority rejected the “clearly erroneous” and “abuse of discretion” standards, what standard did the majority set for itself? The standard is substitution of judgment — about that there is no doubt or disagreement. The truly novel element of this substitution standard is the justification put forth by the majority for its use in this case. The majority states that it has substituted its judgment here because “the [relevance] determinations of the trial court were dependent on, and integrally related to, a legal premise.”60 The majority does not venture beyond this conclusory statement in defense or explanation of its standard of review. In what way are the determinations as to purpose related to the legal premise so as to justify this standard of review? Are all fact determinations made in pursuit of a relevance determination (the “legal premise”) subject to being disregarded under the substitution standard? What is the relationship between law and fact in this case which calls for the appellate court to substitute its view for that of a United States district judge? These are but a few of the vital questions left unanswered by the majority in this case.
The majority concedes that “[w]hen relevance determinations are, in essence, factual judgments, they are normally entitled to special deference from appellate courts . .”61 As noted previously, the majority has attempted to cast the actions of the trial judge as determinations of law by inaccurately characterizing the events in the District Court. Once this false characterization is stripped away, we are left with *438a factual determination of purpose made by the District Court. How was this factual determination “clearly erroneous"? The majority dare not make this claim, for it was the FTC which provided the facts for the trial court’s finding related to purpose. And if it is not clearly erroneous, we must accept it; if we must accept it, the modifications fashioned by the District Court in the name of relevance must be seen as sensible. And, further, when the majority’s straw man who advocates “a particular theory of violation”62 as the only acceptable statement of purpose is removed from the scene (and this straw man is the only advocate of such a narrow focus), the scope of the District Court’s refinement of the FTC’s purpose emerges as a moderate and sound method to accommodate the rights of all the parties to this case.
We are well aware that the distinction between questions of law and questions of fact is often an elusive one. We have, however, shown that the District Court’s inquiry into the purpose of the FTC’s investigation was essentially of a factual nature — to elicit from the FTC information (facts) relating to the purpose for investigating the gas industry at this time. The majority believes it is sufficient to say that the relevance determinations involved a “legal premise” and thus are subject to a substitution of judgment. But they have not told us why this is a question of law; instead, the label “legal premise” is supposed to be taken as a sufficient basis for the appellate court’s usurpation of the trial court’s fact-finding role. The failure to explain the circumstances under which the mere existence of a “legal premise” can trigger de novo review in the appellate court is most disturbing. By failing so to explain, the majority has established a standard which can be employed by the simple unexplained step of denominating a trial court’s decision as resting on a “legal premise.” All litigation is premised on the law; until we have further clarification and analysis of this point from the majority, this standardless form of review can be imposed whenever more than half of the judges sitting on a case decide that they want to substitute their own preferences for the careful hearings of the trial judge.
The majority have correctly observed that there is a “limited role assigned to the federal courts in [subpoena] enforcement proceedings.”63 We fail to see how this observation can be reconciled with the complete substitution of judgment which the majority undertakes in this case. The directionless, standardless review employed by the majority in order somehow to vindicate a “legal premise” does no credit to an appellate tribunal.
B. Raw Field Data, Bid Calculation Data, and Bid Calculation Files
We now narrow our focus to two particular categories of documents which the District Court held need not be produced: (1) “[r]aw field data” and (2) “bid calculation data and bid calculation files.”64 We shall first try to clarify what remains in dispute after the stipulations and then analyze the principal controversy remaining.
As to the first category of documents there appears to be no further controversy between the parties — all parties having agreed that “[t]he Order may be affirmed insofar as it excludes raw field data, subject to the FTC’s right to seek access to or production of such data pursuant to paragraph 9 of said Order. . . ’,65 The *439FTC and producers other than Superior also agreed that “[t]he Order may be modified to expressly exclude maps or other documents relating to the suspected location of natural gas in currently unleased acreage.”66 This second stipulation agreeing to exclude such documents pertaining to “currently unleased acreage” seems to eliminate any controversy between the parties over “bid calculation data” or “bid calculation files” (hereinafter referred to collectively as “bid files”)67 relating to acreage not currently leased by one of the appellees,
Unfortunately, as to bid files relating to acreage that is currently leased or owned by one of the appellees, and to the extent that these files contain information other than “raw field data,” a controversy still exists — perhaps the major controversy in this case. These bid files, which would be liable to disclosure under specifications G, H, and I of the FTC’s subpoena (as modified by the “raw field data” and “treasure map” stipulations supra), include information which producers assemble in advance *440of a lease bid. The information is based upon limited geophysical information and almost never upon actual exploration. The producers consider the models they have developed for making lease bids the most valuable trade secrets they own, because of the large outlays which have gone into their development and the ease with which any producer could be outbid on leases if its competitors had access to the model.68 Understandably, the producers argue strongly against disclosure of their bid files, since, according to them, the data contained therein would give a competitor easy access to their bid model.
Most important to the legal issues herein, they also assert that bid files are not relevant to any calculation of proved reserves, because these files only contain speculative estimates of producing capacity based on limited information. Once a producer obtains a lease and thus is able to drill exploratory wells, these bid estimates are no longer used by any company (except presumably to improve its bidding model or to analyze the bidding behavior of opponents). As a result, the producers contend that bid files could not be relevant to an investigation of conspiratorial and other practices aimed at underreporting proved reserves.
Bid files rarely, if ever, contain proved reserve estimates.69 However, if they do, all appellees except Superior have agreed to produce any proved reserve data contained in their bid files.70
The Trade Commission, on the other hand, argues that “[t]he ‘bid files’ contain estimates of reserves, no matter how speculative or untested, and, as such, are plainly relevant to the analysis of gas reserve reporting which is part of the Commission’s investigation.”71 How unproved reserve estimates, which are never reported, can be “. . plainly relevant to the [FTC’s] analysis of gas reserve reporting . . ,” (emphasis added) we leave to our colleagues to explain.
Our reading of the transcript of 13 December 1973 indicates that the District Court denied production of appellees’ bid files on the theory that they were irrelevant to any type of investigation of reserve reporting. After probing FTC counsel to ascertain whether the FTC had any theories of possible relevance, the court stated,
I hold that the bid files, the estimated unproved and unverified estimates, represented by the bid files are in no way relevant to the investigation or inquiry that you tell me you are making, with one exception.
If you can later come back to me with a situation where a company has been awarded a bid on' a property and has delayed an unreasonable time in drilling on it so they could come up with a proper estimate, then you may apply and I will consider giving you the bid file on that particular one.72
Since issuance of the District Court’s orders, the Trade Commission has implicitly conceded the irrelevance of producers’ bid files on at least two occasions. First, in FTC v. Continental Oil Co.73 one of the cases originally consolidated herewith, the FTC initially appealed, but subsequently moved for dismissal, explaining that FTC representatives had held further discussions *441and negotiations with Continental, that Continental had agreed to submit “further testimony relating to their production in compliance with the FTC’s subpoena, as needed,” and that the ease was therefore moot.74 According to appellees, Continental did not, however, agree to produce its bid files.75 Second, as previously mentioned, the FTC has now offered, as a basis of settling its dispute with Superior, to modify specification G “. .to require Superi- or to produce only documents containing proved natural gas reserve estimates.”76 By agreeing that production of all subpoenaed documents other than bid files was satisfactory compliance for Continental and by offering to settle for production by Superior of only those documents containing proved reserved estimates, the FTC necessarily recognized the correctness of the District Court’s determination that bid files are irrelevant.
In its supplemental brief on rehearing en banc, the FTC argues that resolution of the bid file relevance issue somehow hinges upon resolution of the collateral estoppel issue.77 This both distorts the holding of the District Court and misstates appellees’ position. This court’s determination of the bid file issue (as well as its determination of all other relevance issues raised herein) should be entirely independent of its resolution of the collateral estoppel issue. Relevance and collateral estoppel are two of three wholly independent legal grounds for affirming the District Court’s major limiting modifications. (As mentioned earlier, the third ground is burdensomeness.)
The affidavit of H. R. Hirseh, appended to Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974), lucidly explains why data included in the producers’ bid files has no logical or scientific relevance to any investigation of reports made after the drilling and development of a lease has commenced. In response to this affidavit by a qualified expert with twenty years experience in the oil and gas industry, the FTC offers only the argumentative affidavit of an FTC attorney who states that he has seen some proved reserve data in bid files which he has inspected.78 The FTC’s failure to offer rebuttal testimony by another qualified expert leads us to believe that it has been unable to find an expert willing to endorse its unsound, unscientific, and illogical speculations about the possible relevance of bid file data.
The simple fact is that estimates based on drilling — on actual physical penetration of geological structures — are so superior to speculative bid file estimates based on various geophysical tests that they immediately and completely supersede the earlier estimates. Even if the FTC is completely free of collateral estoppel limitations (as a majority of this court holds) and is allowed to pursue the broadest investigatory purpose it or this court can now envision, only reserve estimates and reports based on drilling (where hydrocarbon-bearing structures have been penetrated) will be of any relevance. After drilling, all prior estimates (there are no prior reports since only proved reserve estimates are reported) become nothing but irrelevant, superseded, and misleading.
Of course, if after making a successful bid on a piece of property a producer does not drill promptly, or if after drilling and discovering proved reserves the producer does not promptly report its discovery, then there might be some cause to look into the company’s bid files.79 These possible delays *442in drilling and reporting, although not supported by any evidence before the District Court,80 were exactly what the court had in mind when it made the following provision in both orders here under review:
The Court reserves its ruling as to any and all matters, contentions or issues not specifically disposed of by this Order. Jurisdiction over these proceedings is retained for the purposes of providing other and further relief as necessary.81
Additionally, at the 13 December 1973 hearing, Judge. Hart expressly notified the FTC that he would be receptive to a motion for the production of any bid file where the Commission could show him that “. a company ha[d] been awarded a bid on a property and delayed an unreasonable time in drilling. . . . ”82 Since through its subpoena, as enforced and modified by the District Court’s Six-Company Order, the FTC would have at its disposal all the information necessary to make a showing of delayed drilling or delayed reporting83 (if either such practice exists), the District Court’s approach seems perfectly sound— especially in light of the highly confidential nature of these files.
With regard to bid files, the FTC advances, in its supplemental brief on rehearing en banc, the following three theories of relevance:
(1) The bid files may be relevant to the calculation of proved reserves because they may contain reserve figures for adjacent tracts,84
(2) The bid files may be relevant in establishing a lease history which, by permitting comparison of proved reserve figures for a tract with initial speculative estimates contained in bid files, might indicate that the proved reserve figures were in error or misstated if they were markedly lower.85
(3) Bid files may be relevant to determine whether appellees have a practice of deferring drilling in certain circumstances to minimize the extent of proved reserves even though they regard their untested estimates to be of comparable degree of certainty.86
In building up the informational foundation necessary to understand the FTC’s bid file theories, we have necessarily touched upon these three theories to some degree. Now, without being unnecessarily repetitive, we will briefly examine and show the error of each theory.
1. Possible Existence of a Limited Number of Proved Reserve Estimates in a New Bid Files
It is highly unlikely that a producer’s bid files would ever include proved reserve esti*443mates. In only one situation is this even hypothetically possible. If a company leased a tract adjacent to an open, unleased tract, and commenced drilling on its tract, perhaps that company could develop sufficient data to make proved reserve estimates for some portion of the adjacent open tract. If the company bid successfully on the open tract and obtained it, then these proved reserve estimates for the adjacent tract would be called for under the subpoena, as enforced by the District Court. If, however, the company did not bid successfully on the open tract, then it could conceivably have in its bid files proved reserve estimates for a tract which was still open or leased by another company.87
In any event, the FTC’s adjacent tract argument does not support its demand for appellees’ bid files in toto —particularly the highly confidential and sensitive documents reflecting bid methodology (i. e., appellees’ bid models). At most, the FTC’s argument would support an order directing them to extract and produce all proved reserve data (if any) found in their bid files. But at this point even this order would be a meaningless gesture, since all parties (except Superior) have agreed in their stipulations that the District Court’s Six-Company Order may be modified by adding the following proviso:
Provided that, to the extent otherwise called for, any and all proved reserve data contained in bid calculation files and relating to tracts covered by [this subpoena] shall be produced.88
2. Evaluating the Accuracy of Proved Reserve Estimates by Comparing Them to Earlier, Superseded, Speculative Estimates in Bid Files
We agree with appellees: “The FTC’s argument that comparing proved reserve estimates on a tract with earlier, superseded speculative reserve estimates may reveal something about the accuracy of the proved reserved estimates illustrates dramatically the danger of relying solely on legal imagination [as a majority of this court does] in a highly technical area such as gas reservoir engineering.” 89 As the only qualified expert whose testimony was before the District Court explained,
[T]he speculative reserve estimates which are based purely on raw geological and geophysical data are often grossly inaccurate. Thus to suggest that they should be compared to proved estimates to see if there has been under-reporting is absurd.90
The evidence and argument presented to the District Court permitted it to make the factual determination that the producers’ bid files contained highly speculative estimates, which would be rapidly superseded by much sounder estimates, once the winning bidder could begin drilling. After the first well is drilled on a lease, no company relies on bid file data, nor is such data ever reported to the AGA. Only data relating to proved reserves is reported, and the essential ingredient in the definition of “proved *444reserves” is that the data is obtained by drilling, by penetration of the formation 91
The “proved reserves” definition is accepted by the industry, the Power Commission, and the Trade Commission here.92 Thus, even if the bid file data showed a gross disparity compared to the data on proved reserves submitted to the AGA, this would demonstrate nothing except that the company’s original bid estimates were sadly in error, for by definition the proved reserve data must be based on geological information obtained by drilling at a later time than when the bid file data is assembled.93 The preliminary and superseded data in the bid files has no relevance whatsoever in determining whether preliminary bid file estimates will later be moved into the more strictly defined category of “proved reserves”.94 In the absence of any such correlation, the FTC’s hypothesis that compilation of a “lease history” dating back to the initial speculative estimates might be relevant in ascertaining possible violations of Section 5 is plainly unsound.95
*4453. Delays in Drilling or Reporting
The Trade Commission makes one final argument: that companies sometimes delay drilling in order not to acquire any proved reserve data which would supersede the speculative estimates in the bid files and which they would be obligated to report to the AGA. Even though the likelihood of such behavior (/. e., of a company paying millions of dollars for a lease and then permitting that lease to go undrilled) seems quite remote,96 the District Court, as we have mentioned before, expressly provided for just such a circumstance.97 Considering the highly sensitive and confidential nature of appellees’ bid files, the District Court’s approach seems particularly sound. Indeed, in the case of Continental Oil Co., even the FTC accepted this approach as satisfactory.98
If a few instances of deferred drilling or reporting do actually exist, the FTC would be able to detect such behavior with the information called for by the subpoenas, as enforced by the District Court’s order.99 By no means, however, would a few such instances justify the blanket subpoena of all bid files (or even the subpoena of the contents of one entire bid file) by the FTC. We would affirm the District Court’s ruling, since it amply preserves the FTC’s interest in having access to those speculative reserve estimates in bid files involved in any case of delayed drilling or reporting, while at the same time it affords appellees protection against the forced disclosure of sensitive data where such information is irrelevant to any alleged delay in drilling or reporting.
C. Summary on the Relevance Issue
How wide of the mark is the Trade Commission’s understanding of the differing roles of different type reserve data — a misunderstanding which unfortunately has infected the majority opinion — is capsuled by this statement: “But any estimate of reserves — however defined — on which a company relies in the course of its business is relevant to the company’s practices in estimating and reporting reserves.”100
Of course a company relies on each of the estimates of reserves it makes, but at different stages of the natural gas business, and with different degrees of confidence and far different purposes. The most preliminary and superficial of exploratory techniques may yield some clues as to natural gas presence or absence. Reliance at this stage is expressed simply as a decision to spend more money on additional exploration. At a later stage, with data assembled from much more sophisticated techniques— but short of drilling or penetration of the geologic formation, the company may rely *446on the preliminary data to the extent of bidding for a lease.
In bidding for a lease the company obviously has some preliminary reserve estimate on which it relies, for the purpose of bidding; otherwise, it would never bid or would make a total gamble. Not only is this reserve estimate never reported, not only is it never revealed to a competitor, but even within company walls it is a closely guarded secret. This preliminary reserve estimate is a product of the “bid model” of the company and the scientific data amassed on that individual tract, and constitutes the heart of the “bid file.”
At this point we note that the Trade Commission has never stated, or even hinted, that it suspects and is inquiring into collusive bidding by the gas producers. Nothing in the record suggests that. If that were an object of inquiry, the contents of the bid files obviously could be relevant, e. g., to show that from differing scientific calculations there strangely emerged similar bids, or that based on similar scientific data the companies bid and failed to bid.
But the inquiry of the FTC is into the reporting of proved reserves and at the bidding stage of developing a gas deposit there is no such thing as a PROVED reserve, and there is no reporting (or revealing) of such reserve estimates as the companies do have. Then when proved reserve data is obtained — by penetration of the formation only — all the preliminary reserve data is literally junked, it is relied upon no more by the companies, reliance is placed entirely upon the proved reserve data, which is reported to the AGA, and which is to be produced — totally—by the subpoenas as modified by the District Court. Anything else the FTC seeks in the way of reserve estimates is completely immaterial to the FTC’s purpose of investigating the collusive or false reporting of proved reserves.
One other statement from the majority opinion will serve to illustrate why, in reading the briefs and hearing oral arguments, we have received the impression that the great problem in this case has been and still is the curtain of ignorance between the Trade Commission staff and counsel and the gas producers, sheer ignorance on the part of the investigators as to the science and economics of the natural gas industry which leaves them without any rational concept of relevance or direction to their inquiry. As the majority says (approvingly, we fear):
“The Commission also suggests that bid files could be used to establish lease histories . . . . [I]f [the] ‘speculative’ estimates were consistently higher than the reported ‘proved’ estimates, and by a roughly equivalent amount — this might well be indicative of anticompetitive practices.” 101
In this statement the Trade Commission exhibits a basic misconception concerning the significance of any disparities between the estimates of preliminary and proved reserves. Repeated disparities of this variety are significant- — indeed, they are precisely the type of error which would cost a company millions of dollars and get a group of geologists fired — and which have absolutely nothing to do with the kind of inquiry the Trade Commission is making here. If the preliminary or “speculative” estimates turn out in repeated instances to be widely different from later proved reserve estimates, there will be considerable soul-searching on the part of the company geologists and higher officials. If time after time the proved reserves delineated by drilling are different from the gas reserve estimates based on preliminary scientific data, what was wrong with the preliminary estimates? What was wrong with the bid model? The company will seek to find out, to correct its bidding on the next available leases, whether the error was in bidding too much or in bidding too little and seeing a profitable tract go to a competitor. But in doing so, we may be sure that the company will not breathe a word of it to a competitor or to the AGA.
*447This is a matter of accuracy of bidding calculations, it has nothing to with the accuracy of proved reserves, but it is the guts of competition in the gas industry. If bid files, bid models, bid calculation techniques are brought out into the public domain 102 by enforcement of the FTC’s unmodified blunderbuss subpoena, then effective competition in both exploration and bidding will be discouraged. We had always assumed that the principal duty of the Trade Commission was to preserve effective competition; it appears in this case not to know the effect of what it is doing.
The most frightening aspect of this revealed lack of comprehension by the Trade Commission is to contemplate what the FTC will do with the bid file data, if the subpoenas are allowed to stand unmodified. Obviously, from the instances publicly known and cited to us by appellees herein, the FTC will discover in the files of every company many horrible examples of the grossest errors in preliminary calculation of reserves as compared to the proved reserve data later developed by drilling. Given the FTC’s concept of relevancy as argued herein, what conclusions will the FTC draw? If a company originally estimated a hundred million cubic feet of recoverable gas, and made an appropriate bid thereon, and now reports only ten million cubic feet in proved reserves, will the FTC trumpet that the company is falsely underreporting proved reserves, because the original estimate was so different? If a company estimated and bid the same amount, drilled twenty dry holes, and abandoned the lease — with proved reserves of zero — the same logic would hold, although we would expect the FTC to have enough good sense not to claim on these facts that the company was deliberately underreporting proved reserves. If Company A estimates a hundred million cubic feet of recoverable gas on a tract, but decides not to bid on it because of a preference for other tracts in the same auction, and Company B is the successful bidder but later reports only ten million cubic feet of proved reserves, will the FTC accuse Company B of false underreporting of proved reserves?
The examples could be multiplied endlessly, with but only one question to be put to the FTC: Can the FTC draw a conclusion of false or collusive reporting from discrepancies between the proved reserve data and the preliminary reserve estimates? If the FTC thinks it can, then it is in the process of perpetrating a terrible hoax on the American people. If the FTC admits it cannot, then it concedes the irrelevancy of the preliminary reserve estimates for the purpose of its investigation.
The conclusion is inescapable: the preliminary estimates of reserves before drilling on which bids are made have no earthly relevance to proved reserve calculations based solely on drilling into the geological formation. Actual production from drilling is the proof, hence the term “proved reserves”; every estimate, no matter how much scientific thought has gone into it, made before is nothing but an educated guess — -it is not proof, and it has no bearing on the accuracy of the proved reserve figures, which alone are reported.
In sum, none of the theories of bid file relevance advanced by the FTC withstands analysis. The leading Supreme Court case on FTC investigations, United States v. Morton Salt Co.,103 makes clear that the burden is upon the FTC to prove that the bid files (and all other documents not related to proved reserves)104 are “reasonably *448relevant” to its investigation,105 This the FTC has not done, either here or in the District Court. The finding on relevance by the District Court is “essentially factual in nature” and cannot be overturned unless it was “clearly erroneous” or “an abuse of discretion.” 106
V. BURDENSOMENESS
We turn now to the second independent ground supporting the District Court’s order — burdensomeness. In formulating its order enforcing the FTC’s subpoena, the District Court had to consider the reasonableness of the burden of compliance placed upon the respondents.107 And where, as here, this consideration leads to a finding that the subpoena is unduly burdensome, the District Court may act to alleviate this burden by modifying or partially enforcing the subpoena.108 Such modification or partial enforcement is an exercise of the District Court’s sound discretion and may be overturned on appeal only for an abuse of discretion.109
*449In FTC v. Loaning this court recognized the appropriateness of an abuse of discretion standard when it recently had to evaluate the burden placed on a cereal manufacturer by a district court’s order enforcing another FTC subpoena. The cereal manufacturer took the position that certain data called for under the subpoena “. should be disclosed only to counsel of record who are ‘outside’ counsel and should not be revealed to any counsel of record who is inside or house counsel for a respondent.”110 This court disagreed and sustained the District Court which had adopted the protective order issued by the administrative law judge permitting disclosure to all counsel of record. The court concluded,
We see no abuse of discretion in limiting the disclosure of the individual brand cost data to counsel of record. Although this Court has previously commented that there may be some distinction between disclosure of trade secrets to house counsel, and outside counsel, we find no abuse of discretion in refusing to limit such disclosure in this case. . . . The decision as to the type and scope of any protective order rests within the sound discretion of the trial judge and must be determined on a case by case basis. There is nothing in this record which would support a determination that the district court abused its discretion by its adoption of the protective order issued by the Administrative Law Judge.111
Thus, like appellants in NLRB v. Northern Trust Co., our colleagues and the FTC “. . . are in the unenviable position of sustaining the great burden of showing an abuse of discretion.”112 We conclude that they, too, have failed to do so.
In determining the reasonableness of the burden of compliance cast upon appellees, the District Court was, of course, free to consider all pertinent facts and circumstances. The repetitive and cumulative nature of the Trade Commission’s subpoena (in view of the Power Commission’s previous factual determination that AGA proved reserve data was reasonably reliable for ratemaking purposes)113 was, and should have been, an important factor in the District Court’s assessment of burden.114 The repetitiveness of an agency’s demands cannot be ignored or excused merely because the repetition and cumulativeness are the result of similar demands made by other agencies of the same federal government.115
The majority’s position in this regard appears to be that repetition and cumulativeness must be excused where agency jurisdictions overlap or where, as has become increasingly common, different agencies or officers of the federal government assume an adversarial or competing posture toward one another. We strongly reject this position, a position which will leave district courts — even in the most extreme cases— powerless to ameliorate the effects of repetitive agency demands. In the words of one of the appellees, “Such a result would create bureaucratic competition having the characteristics of a Kafka nightmare in which no response to an agency is ever sufficient because of the needs of a competing agency to show the insufficiency of a *450prior response to the former agency.”116 In such cases of competing agencies or overlapping jurisdiction, we would leave it to the sound discretion of the district courts to determine whether the effort and expense involved in responding to repetitive agency demands imposes an unfair and unreasonable burden on the responding parties. Being another essentially factual determination, this has always been, and should remain, the province of the district courts.
Here there was a full showing of the immense burden which even partial compliance with the FTC’s subpoena would impose on appellees. For example, Standard Oil Co. of California filed in the District Court an affidavit of one of its vice-presidents setting forth a preliminary estimate of the time and cost of compliance, as he understood the subpoena’s specifications. He stated that in order to locate the responsive documents, more than four million documents would have to be reviewed requiring approximately sixty-two man-years. The total cost of searching, reviewing, reproducing, and refiling was estimated at approximately four million dollars.117
Considering the purpose of the Trade Commission’s investigation118 and the factual determination of the prior Power Commission investigation into reserve reporting,119 the District Court was eminently fair and reasonable (1) in limiting the disclosures required under the FTC’s subpoena to those documents relating to the reporting of proved reserves and (2) in not permitting the use of any such documents to reinvestigate or redetermine the amount of proved gas reserves in Southern Louisiana. By no stretch of the imagination can either of these limitations be characterized as an abuse of discretion.
On burdensomeness issues, the majority opinion recognizes the “abuse of discretion” standard (there’s too much binding authority on this point), but attempts to escape from the teeth of the abuse of discretion standard by arguing that the District Court’s determinations of burden were intimately tied to and colored by improper applications of collateral estoppel and relevance. Therefore, says the majority, we shall review these modifications for “mere error.” 120 The majority cannot escape their obligatory standard so easily. In trying to do so, they have confused the overlap of some facts basic to more than one legal theory with an overlap of the legal theories themselves.
Of course there is some overlap in the factual foundations underlying burdensomeness and collateral estoppel; to some extent both grounds rely upon the finding of fact discussed in Section II supra. (However, the third ground affirmance— relevance — is entirely independent of collateral estoppel and burden, both in terms of legal theory and underlying factual foundation.) Notwithstanding the overlap between FACTUAL foundations, burdensomeness and collateral estoppel are independent LEGAL grounds for affirming the District Court’s “use” restriction. Similarly, burdensomeness and relevance are independent legal grounds for affirming the District Court’s PROVED reserve limitation. Each of these grounds would be sufficient without the others. Accordingly, there is no way, as the majority opinion puts it, that the District Court’s views on collateral estoppel and relevance could have somehow “infected” its determinations based on burden. Even if the legal theories *451of collateral estoppel and relevance had never been raised by appellee’s counsel or ever thought of by the District Court, the major modifications of the court’s order would be fully sustainable on burdensomeness grounds alone.
There can be no question that the Trade Commission has jurisdiction to determine whether the antitrust laws or the FTCA has been violated. However, the District Court was entitled to conclude on the present record (1) that the Trade Commission desired all reserve data in the possession of appellees in order to recompute independently Southérn Louisiana reserves, and (2) that the Power Commission, on the basis of a contested evidentiary hearing, had found that the industry’s figures for Southern Louisiana reserves for the relevant years were accurate. Based on these findings, the District Court could reasonably have concluded that it would be unjust, unreasonable, and unduly burdensome to require the production of every scrap of data which related to reserves of any kind. By limiting production to “documents containing or underlying proved natural gas reserve estimates” and by restricting the use of these documents to “the sole purpose of permitting the Trade Commission to investigate whether there is a conspiracy in the reporting of natural gas proved reserve estimates,” the court drew a reasonable balance between the investigatory needs of the Trade Commission and the producers’ burdensomeness claims. Similarly, through its “use” restriction the court struck an equally reasonable balance between the investigatory needs of the FTC and the producers’ collateral estoppel claims.
The majority opinion has not only managed to confuse overlapping facts with overlapping law, in their effort to escape applying the abuse of discretion standard, but the opinion manages to confuse in which forum a burden of proof concept is applicable. The majority claims that “[t]he burden of showing that the request is unreasonable is on the subpoenaed party.”121 Of course it is — but in the District Courtl On appeal we as an appellate court review the issue of burdensomeness with an abuse of discretion standard.
Why the majority falls into this error is obvious. From the start, and continuing throughout, the majority opinion proceeds as if this court were reviewing directly an administrative agency decision.122 The majority ignores the intervening action of the District Court in modifying and enforcing the administrative subpoenas, which under the law requires us to review the action of the District Court — not the FTC — under certain well defined standards. Having ignored the District Court and cast themselves in that role, the majority unwittingly adopt a standard fitting for — and applied by — the District Court. A very revealing error. An error which goes a long way to support the Trade Commission’s effort to eliminate effective judicial review of administrative subpoenas.123
Having stated, from the standpoint of burdensomeness, our position on the District Court’s two most important modifications, approving the court’s other minor modifications becomes almost an a fortiori exercise. Consequently, we will forego any detailed analysis of these modifications, and comment on them only very briefly in section VII, infra.
VI. COLLATERAL ESTOPPEL
We turn now to an analysis of the third independent ground supporting the District Court’s order — collateral estoppel. Initially, it is important to understand that appellees have never sought, and the District Court did not issue, an order halting the FTC’s inquiry in this proceeding. As stated in their initial brief to this court, “Respondents do not contend that the FTC has no authority to investigate the allegation of collusive reporting or fabrication of reserve *452estimates.”124 Appellees do contend that under the principles of collateral estoppel, and alternatively, in order to prevent the imposition of an unfair burden of compliance, the FTC should not be permitted to redo for a third time what has already been done by another equally competent agency of the federal government.125
Having already concluded in Section II supra that the District Court’s factual determination that the FPC had already (1) determined natural gas reserves and (2) considered and ruled upon the validity and accuracy of such reserves was correct (indeed, implicitly assumed by the majority opinion), we focus now on the question of law confronting the District Court after it made this important finding of fact: Under the circumstances of this case, is it appropriate to give this finding of fact collateral estoppel effect?
The Supreme Court and this court have clearly stated that an agency or a private party can be collaterally estopped in a later court proceeding if a relevant factual issue has already been resolved in a contested hearing before the agency.126 The Supreme Court recently affirmed this principle in United States v. Utah Construction Co.:
Occasionally courts have used language to the effect that res judicata principles do not apply to administrative proceedings, but such language is certainly too broad. When an administrative agency is acting in a judicial capacity and resolves disputed issues of fact properly before it which the parties have had an adequate opportunity to litigate, the courts have not hesitated to apply res judicata to enforce repose. Sunshine Coal Co. v. Adkins, 310 U.S. 381, 60 S.Ct. 907, 84 L.Ed. 263; Hanover Bank v. United States, 285 F.2d 455, 152 Ct.Cl. 391; Fairmont Aluminum Co. v. Commissioner, 4 Cir., 222 F.2d 622; Seatrain Lines, Inc. v. Pennsylvania R. Co., 207 F.2d 255. See also Goldstein v. Doft, D.C.N.Y., 236 F.Supp. 730, aff’d, 2 Cir., 353 F.2d 484, cert. denied, 383 U.S. 960, 86 S.Ct. 1226, 16 L.Ed.2d 302 where collateral estoppel was applied to prevent relitigation of factual disputes resolved by an arbitrator.127
The Supreme Court has also held that collateral estoppel can be applied when the prior proceeding involved a different government agency because, for collateral estoppel purposes, agencies of the same government are in privity with one another:
Where the issues in separate suits are the same, the fact that the parties are not precisely identical is not necessarily fatal. As stated in Chicago, R. I. & P. Ry. Co. v. Schendel, 270 U.S. 611, 620, 46 S.Ct. 420, 423, 424, 70 L.Ed. 757 “Identity of parties is not a mere matter of form, but of substance. Parties nominally the same may be, in legal effect, different, . and parties nominally different may be, in legal effect, the same.” A judgment is res judicata in a second action upon the same claim between the same parties or those in privity with them. Cromwell v. County of Sac, 94 U.S. 351, 24 L.Ed. 681. There is privity between officers of the same government so that a judgment in a suit between a party and a representative *453of the United States is res judicata in relitigation of the same issue between that party and another officer of the government.128
Other Circuits have applied both these principles in appropriate cases. The Eighth Circuit in George H. Lee, Co. v. FTC129 has held that the PTC was collaterally estopped from claiming use of unfair methods of competition where the claim was based on factual issues resolved favorably for defendants in a prior proceeding instituted under the Food and Drug Act. Reciprocally, the Seventh Circuit in United States v. Willard Tablet Co.130 upheld the defense of collateral estoppel in a proceeding under the Food and Drug Act where a prior proceeding before the FTC held that defendant’s labeling claims were not deceptive. More recently, in Safir v. Gibson131 the Second Circuit held that the Maritime Administration, an agency within the Department of Commerce, was estopped from re-investigating and redetermining an issue previously decided by the Federal Maritime Administration.
In Safir the Second Circuit gave collateral estoppel effect to the FMC’s prior determination that the rates of a conference of water common carriers were unjustly discriminatory and unfair to plaintiff Safir’s company. Writing for the court, Judge Friendly explained,
. While the issues here may not have been purely factual, they were fully litigated before the agency designated to determine them. . . The Restatement of Judgments, § 70, says that even determinations of questions of law are conclusive between the parties on a different cause of action unless injustice would result. .
It is the FMC, not the Maritime Administration, that has the expertise to pass on whether rates are unfair or unduly discriminatory, . . . and it would be quite unseemly for the Maritime Administration to conclude that its sister agency had been wrong on a fully litigated issue the decision of which Congress had confided to it. .
We recognize there is a general rule against judicial interference with administrative proceedings prior to the issuance of a final order . . .. Nevertheless we believe ... it appropriate for us to direct conformity with our views on the preclusive effect of the FMC decision here and now. . . . [T]he reason for applying res judicata to administrative agencies is not only to “enforce repose” but also to protect a successful party from being vexed with needlessly duplicitous proceedings. ... If the latter interest is not protected at the outset of the second proceeding, it will be lost irreparably. . . .132
We conclude that the Trade Commission is now in the same position as the Maritime Administration was in Safir, as the United States was in Willard Tablet, and as the FTC itself was in George H. Lee.
Regardless of the general applicability of res judicata and collateral estoppel principles between administrative adjudications involving different agencies, on the facts of this case application of the doctrine of collateral estoppel is particularly warranted. The Power Commission is the federal agency with special technical and scientific ex*454pertise on factual issues involving the oil and gas industry (e. g., the accuracy of AGA proved reserve estimates), and as such its findings of fact on oil and gas matters are entitled to the respect of other government agencies. The Trade Commission knew of the Power Commission’s adjudicatory proceeding in So La II when it initiated its own investigation into gas reserve reporting and could have intervened in the Power Commission’s ongoing investigation. This way, perhaps the Trade Commission could have avoided the charge leveled so persuasively by the producers in the District Court — that some of the documents called for under the Trade Commission’s subpoena were sought for the sole purpose of collaterally attacking a Power Commission finding of fact on an oil and gas matter.
The District Court, although it found the producers’ charge to be perfectly valid, decided not to restrict production to only those documents which could never be used to attack collaterally the FPC’s prior finding. Instead, the court took the less restrictive approach of placing the following “use” restriction on all data produced pursuant to specifications G, H, and I:
All production . . shall be made for the sole purpose of permitting the Trade Commission to investigate whether there is a conspiracy in the reporting of natural gas proved reserve estimates, and not for the purpose of permitting the Trade Commission to investigate or determine the amount of proved natural gas reserves.133
Thus, even though some data containing or underlying proved reserves, and ordered to be produced by the court, could conceivably be used by the Trade Commission in a collateral attack on the Power Commission’s reserve estimates finding, this “use” was strictly proscribed by the court’s order. It is this “use” restriction, soundly based on either collateral estoppel or burdensomeness grounds, which the Trade Commission attacks so vigorously in this appeal.
Having briefly explained why it is appropriate to give the Power Commission’s reserve estimates finding collateral estoppel effect, we now direct our attention to the reasons which the FTC and a majority of this court give for reaching an opposite result. Generally, they conclude that the doctrine of collateral estoppel has no applicability here because (1) it is inappropriate to consider the issue of collateral estoppel at the subpoena enforcement stage of an investigatory proceeding, (2) the Power Commission’s reserve estimates finding was made in the context of a quasi-judicial rate-making proceeding, and (3) the Power Commission found the AGA reserve figures “reasonably reliable” for ratemaking purposes only. To our mind, none of these reasons are valid or should preclude application of administrative collateral estoppel.
A. The Appropriateness of Considering the Issue of Collateral Estoppel in this Subpoena Enforcement Proceeding
The issue of collateral estoppel is not an issue on the merits that must be deferred until an administrative complaint has issued; far from it, deferring consideration does violence to the sound policy justifications underlying the doctrine. Thus collateral estoppel is appropriately at issue in this enforcement proceeding. As the Trade Commission itself recognizes, “Collateral estoppel is now seen as an immensely practical doctrine, rooted in considerations of fairness and the public policy favoring finality in litigation.” 134 These fairness and public policy considerations, including “. finality to litigation, prevention of needless litigation, [and] avoidance of unnecessary burdens of time and expense[,] are as rele*455vant to the administrative process as to the judicial.”135
One important consideration which argues for applying collateral estoppel at this stage of the instant proceeding is the need for a consistent and final determination of a technical oil and gas issue by the expert agency in the oil and gas field. Early consideration of the issue ensures that judicial and administrative resources, as well as the resources of the litigants, will not be wasted needlessly. In addition, the Trade Commission’s inquiry into alleged underreporting of reserves is neither stymied nor prohibited by the District Court’s “use” restriction. The only restriction is that the Trade Commission must accept the Power Commission’s factual finding with respect to the reliability and accuracy of AGA proved reserved estimates. No administrative agency, including the Trade Commission, has carte blanche authority to reinvestigate and relitigate a factual issue simply because it disagrees with another agency’s finding. The collateral estoppel doctrine serves the essential purpose of protecting the agencies, the courts, and the parties from such unnecessary and wasteful expenditures of time and money.
Endicott Johnson Corp. v. Perkins,136 Oklahoma Press Publishing Co. v. Walling,137 and other cases refusing to consider at the investigative stage certain defenses which should be raised on the merits against a formal complaint, illustrate the generally-accepted principle that courts should not interrupt the administrative process except under exceptional and narrowly-defined circumstances.
Collateral estoppel, as applied by the District Court, is easily distinguishable from the jurisdiction and statutory coverage questions dealt with in Endicott Johnson, Oklahoma Press and their progeny.138
First, in this case the main policy goal behind nonintervention has been fulfilled; the expert agency in the oil and gas field (the Power Commission) was allowed initially to apply its expertise and make the factual determination on which ultimate decisions will later be based. Second, in contrast to the Endicott Johnson line of cases, here the District Court’s limited application of collateral estoppel would not preclude the Trade Commission from exercising its full statutory and jurisdictional authority in pursuing its investigation of possible Section 5 violations. The District Court here did not intrude on the jurisdiction of the Trade Commission or decide a jurisdictional question of statutory coverage.139
This simply is not an ENDICOTT JOHNSON-type case. Endicott itself dealt not with the issue of collateral estoppel, but with the question of whether the Secretary of Labor could subpoena data for the purpose of determining if a company’s activities were covered by (/. e., within the jurisdiction of) the Walsh-Healy Act. Denying *456enforcement of the Secretary’s subpoena would have had the effect of preventing a determination of the coverage issued by the very person (the Secretary) authorized by the statute to make that determination. Specifically, the Walsh-Healy Act directed the Secretary to make
findings of fact after notice and hearing, which findings shall be conclusive upon all agencies of the United States, and if supported by the preponderance of the evidence, shall be conclusive in any court of the United States; and the Secretary of Labor shall have the power, and is hereby authorized, to make such decisions, based upon findings of fact, as are deemed to be necessary to enforce the provisions of this Act.140
With this language in mind, the Supreme Court found that “. . . under the statute determination of [the coverage] issue was primarily the duty of the Secretary . ”141 and not the district court.142
Here we have no such statutory authorization; the Federal Trade Commission Act clearly does not make the estimation of natural gas reserves primarily the duty of the Trade Commission and cut out the Power Commission; nor does it make the Trade Commission’s findings of fact on oil and gas matters “conclusive upon all agencies of the United States” (i. e., upon the Power Commission). Quite to the contrary, the conceded expert on all technical aspects of the oil and gas industry is the Power Commission, not the Trade Commission.143 In So La II the information necessary to reach a determination on the gas supply question was placed before the Power Commission, the agency most qualified to make such a determination, and that agency made specific findings of fact based on a lengthy (over 30,000 pages) and comprehensive record. Those findings are now fully available to the Trade Commission and are entitled to its respect as a matter of law and logic.
The Second Circuit’s opinion in Safir strongly supports our position. There, in concluding that collateral estoppel was applicable to an administrative proceeding at the investigative stage, Judge Friendly pointed out that the concepts of collateral estoppel and res judicata differ from defenses raised in cases like Endicott Johnson and Oklahoma Press, where the respondents contended that the agencies lacked jurisdiction:
[T]he reason .for applying res judicata to administrative agencies is not only to “enforce repose” but also to protect a successful party from being vexed with needlessly duplicitous proceedings. . If the latter interest is not protected at the outset of the second proceeding, it will be lost irreparably. ... In this respect, a claim of res judicata differs from a claim that an administrative agency has no jurisdiction over the subject matter of the investigation . an issue which Congress meant to be decided in the first instance by the agency itself.144
In'accusing the writers of this dissent and the District Court of (1) attempting to discern the shape of the merits, (2) trying to divine all possible issues that could arise in the course of an FTC investigation, (3) at*457tempting to forecast the ultimate conclusion of the FTC’s proposed investigation, and (4) generally putting our cart before our horse, the Trade Commission and our colleagues repeatedly overlook the difference between res judicata and collateral estoppel, i. e., the difference between claim preclusion and issue preclusion. When this difference is understood it becomes apparent how both we and the District Court are able to apply the doctrine of collateral estoppel to the facts of this case without so much as a passing glance into' our crystal ball.
The FTC is fully aware of the difference between issue and claim preclusion, and it has even had occasion to admonish parties appearing before it that collateral estoppel precludes the redetermination of factual issues while res judicata forecloses entire claims or ultimate issues:
Strictly speaking the doctrine of res judicata refers to the merger or bar of a subsequent action based on the same cause of action as opposed to the doctrine of collateral estoppel under which the determination of a question of fact essential to a judgment is conclusive between the parties (and their privies) in a subsequent action on a different cause of action.145
Thus, to apply the doctrine of collateral estoppel (not the doctrine of res judicata) to the facts of this case, we need not attempt to forecast the ultimate conclusion, or even the ultimate issue of the FTC’s proposed investigation; we need only define ONE FACTUAL ISSUE (the accuracy of AGA proved reserve estimates) which has already been determined by the Power Commission. The Trade Commission has a free investigative rein on ALL ULTIMATE issues (e. g., have appellees committed or conspired to commit an unfair trade practice; have appellees engaged or conspired to engage in a price-fixing scheme; are appellees guilty of an attempt to monopolize) and on all relevant factual issues except the accuracy of AGA proved reserve estimates.
Thus two recent decisions of the Sixth and Seventh Circuits, relied on by the Trade Commission, in no way conflict with our position here, nor with the decisions in Safir and others which we believe govern this case. In FTC v. Feldman146 and FTC v. Markin147 two taxicab companies argued that a 1947 holding that they had not been engaged in interstate commerce barred the FTC from now investigating the industry. First, it is obvious that changed facts and changed law might justify an investigation after 29 years. In our ease the Trade Commission proposes to examine the reporting of proved reserves in the same time frame as the Power Commission did. It was on precisely this point that the Seventh Circuit distinguished Feldman from the panel opinion in our case, with which this dissent agrees. Second, in both Feldman and Mar-kin the private litigants sought to stop the whole investigation, i. e., claim preclusion. Here the appellee producers have never sought to bar the Trade Commission investigation, they have never asserted that the Power Commission’s inquiries immunized them against further investigation; they have only asserted that the Power Commission’s determination of one factual issue, the accuracy of the proved reserve reporting during the same time span, must be taken as established under the doctrine of collateral estoppel, i. e., issue preclusion.
Unlike the situation in Endicott Johnson and similar cases, the District Court here was not making a jurisdictional determination 148 or arrogating to itself a factual de*458termination that was primarily the duty of the Trade Commission. In accordance with well-established principles of collateral estoppel, the District Court simply applied an already-found fact to the proceedings before it. The court did not try the issue of underreporting itself; nor did it delve into the same conflicting materials that the Power Commission had previously considered in So La II. The Trade Commission was foreclosed on collateral estoppel grounds from reinvestigating and redetermining this one question of fact, not by the District Court, but by the Power Commission, a sister agency of at least equal competence on oil and gas matters.149
Likewise, the Power Commission was not arrogating to itself a factual determination that was primarily the duty of the Trade Commission. On this one question of fact, the Trade Commission simply began its investigation too late. On 24 November 1971, when these subpoenas issued, the FTC was estopped from collaterally attacking any finding of fact essential to the FPC’s rate-making determinations in So La II (issued on 16 July 1971). Having to accept this one finding of fact can have the effect of conferring antitrust and Section 5 immunity on appellees only insofar as the FTC’s antitrust theories necessarily require an opposite finding. If it turns out that all of the FTC’s antitrust and Section 5 theories do require a finding that AGA proved reserve estimates for the years 1962-1970 were in accurate, then the problem here is not an overly restrictive enforcement order, but a poorly aimed investigation.150
B. Giving Collateral Estoppel Effect to a Finding of Fact Made in the Context of a FPC Ratemaking Proceeding
As the Supreme Court made clear in United States v. Utah Construction & Mining Co.,
When an administrative agency is acting in a judicial capacity and resolves disputed issues of fact properly before it which the parties have had an adequate opportunity to litigate, the courts have not hesitated to apply res judicata to enforce repose.151
By its use of the words, “[w]hen acting in a judicial capacity,” we do not believe the Court intended to engraft on the doctrine of administrative collateral estoppel the drastic restrictions advocated by the Trade Commission. The fact that the Power Commission’s ratemaking proceeding in So La II was “quasi-judicial” in nature, rather than purely “judicial”, does not mean that the facts found therein are not entitled to collateral estoppel effect.
The crucial, threshold requirement for any application of collateral estoppel is a truly adversary proceeding, i. e., a proceeding sufficiently adverse to guarantee that opposing interests are adequately repre*459sented and that the facts established can withstand the test of opposition. When facts are established as a result of the adversary process (i. e., when they have been challenged and controverted by opposing interests), they acquire an added dimension of respectability and are therefore accorded a binding effect on parties (and their privies) in subsequent proceedings. If there is no adversariness in a proceeding, then there is no assurance that the facts have been truly established; the facts found may well be inaccurate since they have not been tested by an opposing interest. We submit that so long as there is sufficient adversariness in a quasi-judicial, ratemaking proceeding, the necessary guarantee of trustworthiness is present. Accordingly, the facts found therein should be given collateral estoppel effect.
There is little room to quarrel over the adversariness of the Power Commission’s proceedings in So La II. All evidence was developed in an adversarial environment with public and private parties representing sharply opposing interests participating at all stages.152 Witnesses were subjected to rigorous cross-examination, and an opportunity for rebuttal testimony or rebuttal evidence was provided.153 Based on this record, the Power Commission’s staff concluded in its initial brief in So La II that “. the validity and reliability of the reported AGA reserves data [was established] beyond any reasonable doubt.”154 In So La II itself the Power Commission specifically found that the AGA reserve estimates had not been “impeached or substantially contradicted” 155 and that they were “reasonably reliable” for ratemaking purposes.156 This area rate proceeding, bringing together as it did sharply divergent economic interests in the same arena to do battle, provided an excellent context in which to determine finally the accuracy of AGA proved reserve estimates.
Discussing the problem of how and where to draw the line between “judicial” and “nonjudicial” action for purposes of collateral estoppel and res judicata, Professor Davis in his Administrative Law Treatise suggests,
The best approach is to avoid the labels that have been attached to various functions for other purposes and to determine what is judicial or nonjudicial for purposes of res judicata by emphasizing factors which relate to res judicata. For instance, the Supreme Court has held that because the granting of broadcasting licenses is nonjudicial, that function cannot be vested in the Supreme Court. But a function may be nonjudicial for one purpose and judicial for another purpose. The Supreme Court’s holding that a court ought not to grant and deny licenses may be entirely sound, for practical reasons relating to qualifications, and it is natural for the court to say in such a context that the function is nonjudicial. But if an agency having the licensing power conducts a full hearing and adjudicates a controversy about past facts concerning *460the applicant, the determination should ordinarily be res judicata ; for want of a better set of terms, the conclusion may even be announced that the action is deemed judicial. The question is not what is judicial in the abstract or for some other purpose. The question is whether considerations relating to res judicata require that the particular action be regarded as judicial or nonjudicial.157
Like the FCC in Professor Davis’ example, the Power Commission in So La II conducted a full hearing and adjudicated a controversy about past facts concerning appellees. Hence, the Power Commission’s findings of fact with respect to that controversy are entitled to collateral estoppel effect. For purposes of administrative collateral estoppel, the only litmus test is adversariness (i. e., a FULL hearing adjudicating a CONTROVERSY about past facts), not whether the administrative proceeding is labeled “licensing” or “ratemaking”.
As this court has recognized on prior occasions, there is no clear, bright line between adjudicative (judicial) proceedings and rulemaking (legislative or nonjudicial) proceedings. On one side of this coin is our decision in Mobil Oil v. FPC;158 on the other side our decision today. In Mobil Oil we recognized that even in a legislative rulemaking proceeding under APA § 553, the circumstances of the case might call for “some kind of hearing.” 159 Today, however, the majority refuses to recognize that where there is sufficient adversariness in a quasi-judicial, ratemaking proceeding, the principles of collateral estoppel should apply in the same way they do in pure judicial proceedings. In Mobil Oil the requirement of “some kind of hearing” derived from the circumstances of the case; here the application of collateral estoppel should derive from the circumstances of this case — the adversariness of the hearings in So La II.
In a suit for refund of income taxes paid upon a deficiency assessment the Fifth Circuit has given collateral estoppel effect to a prior Internal Revenue Service determination letter revoking a taxpayer’s tax-exempt status.160 Despite the fact that an IRS determination letter bears little resemblance to a “judicial” proceeding, the court held that “. . . the revocation of appellee’s exemption must be considered a judicial act for the purpose of res judicata.” 161 Factors influencing the Fifth Circuit’s decision included, inter alia, finality to litigation, prevention of needless litigation, and avoidance of unnecessary burdens of time and expense; that the agency’s action was directed specifically at the taxpayer and was not a general rule applicable to all organizations of a given class, that the agency’s action had an immediate effect upon the status of the taxpayer; and that the taxpayer was well-informed of the progress of the agency’s proceedings.162
These same factors militate in favor of giving the findings of fact in So La II collateral estoppel effect. There, in a proceeding bearing all the hallmarks of the adversary system, the issue of the accuracy of the AGA proved reserve estimates was fully considered and finally determined.
C. Giving Collateral Estoppel Effect to a Finding of Fact Made for Rate-making Purposes
In So La II the Power Commission specifically found that “. . . the AGA reserve data is reasonably reliable for the purposes used herein.” 163 The Trade Commission interprets this to be a statement intended to limit and qualify the Power Commission’s acceptance of the AGA data. Taken as a whole, the record in So La II *461clearly does not support this inference. In fact, in its initial brief in So La II the FPC staff concluded that “. . . the record established] the validity and reliability of the reported AGA reserves data beyond any reasonable doubt.”164 Furthermore, the statement “for the purposes used herein” can just as easily be taken on the positive side to mean that these figures are good enough even for the FPC’s use in calculating just and reasonable rates, or it can be taken as a relatively meaningless and gratuitous expression intended to have no substantive effect whatsoever. Whatever its meaning, we doubt seriously that the Power Commission intended it to limit the subsequent collateral estoppel effect of the finding it precedes.
Arguing along this same line, the Trade Commission makes much of the fact that the ultimate purpose of its investigation (determining if there has been a violation of Section 5 of the FTCA) differs from the ultimate purpose of So La II (determining the just and reasonable area rate for the Southern Louisiana area). However, since the FTC now seeks to determine whether the producers are underreporting reserves and are thereby violating the FTCA,165 we must agree with appellees that any distinction based on different ultimate purposes is purely illusory. No matter how ingenious or how farfetched it is, any antitrust or unfair trade practice theory must have at its core an illegal activity having some effect (or at least some intended effect) on area rates.166 Hence, the Power Commis*462sion’s finding of reliability, even if for rate-making purposes only, cannot be nonchalantly brushed aside as a limited and irrelevant factual determination. While this was indeed a finding of fact made for ratemaking purposes, it was nevertheless a determination that the only estimates (i. e., AGA proved reserve data) which could possibly affect rates were reliable and accurate. (We repeat: since the only supply estimates relied upon by the Power Commission are AGA proved reserve estimates, only these data can ever affect the rates set by the Commission.)
For basically the same reasons, the majority’s reliance on United States v. RCA167 is misplaced. In RCA the broadcasting company argued that the Federal Communications Commission’s prior approval of its agreement to exchange a television station in Cleveland for one in Philadelphia immunized that agreement from a future antitrust attack by the Justice Department. In other words, the broadcasting company urged the court to give res judicata or collateral estoppel effect to the FCC’s determination that this agreement was “in the public interest” (the ultimate issue before the FCC). Here, appellees are not asking this court to give res judicata or collateral estoppel effect to the FPC’s rate-making determination (the ultimate issue in So La II); nor do they claim any immunity from the antitrust laws or the FTCA. The accuracy of AGA proved reserve data was a factual issue in controversy in So La II, and it is only this one factual issue, not appellees’ antitrust or Section 5 liability, which the FTC is estopped from re-investigating and redetermining. In RCA there was no finding of fact by the FCC on the antitrust question which concerned the Justice Department; nor could there have been since the FCC was not authorized to decide antitrust issues as such.168 Here, in comparison, there was a finding by the FPC on a factual issue concerning the FTC. Additionally, there can be no contention here that the FPC, the expert agency on oil and gas matters, was not empowered to determine the accuracy of the AGA proved reserve estimates.
In its supplemental reply brief on rehearing en banc, the FTC takes one final desperate stab at a “ratemaking purposes” argument, asserting that “factual determinations in ratemaking proceedings are not ‘final’ even among the parties, and hence are not entitled to be the basis for applying collateral estoppel in other contexts.”169 Unfortunately, the FTC overstates its case. As the Trade Commission itself recognizes in the sentence that immediately precedes this statement, the only case in point170 indicates that “. . .an agency is ordinarily free ... to reconsider factual determinations in the light of new evidence. ”171 Thus, on factual issues (e. g., the accuracy of AGA proved reserve data) the FPC can change its mind if new evidence or changed circumstances arise.172 But this observation applies with equal *463force to the factual determinations of a judicial proceeding. Accordingly, it does not preclude application of collateral estoppel in either context. If it did, no factual determination, be it administrative or judicial, could ever be given collateral estoppel effect.
VII. CONFIDENTIALITY AND PRODUCTION AT SITUS
The District Court attached the following conditions to the disclosure of documents designated as confidential:
(1) The Secretary of the Federal Trade Commission is designated the custodian of the documents;
(2) The documents (and presumably any documents or memoranda derived therefrom) must be kept in a depository with access restricted to the FTC employees assigned to the investigation;
(3) Documents can only be removed from the depository or used for other purposes with the court’s permission; and
(4) Upon the termination of the investigation, the documents (and presumably all copies of documents) must be returned to their owner.173
The Trade Commission does not question the confidential nature of the documents it seeks disclosed. Rather, its position is that the FTCA and the Commission’s Rules of Practice provide appellees with adequate protection. Quite to the contrary, the FTCA and the Rules of Practice merely state that the public disclosure of geophysical data or information and trade secrets is within the discretion of the Commission. The FTC’s rules governing in camera orders, the release of confidential information, and requests for disclosure of records clearly indicate that the Trade Commission will decide ultimately whether records exempt from disclosure under the Freedom of Information Act (as most of these records probably would be) will be disclosed.174
The District Court was not required to rely on the unbounded discretion of the Trade Commission to keep the producers’ estimates confidential.175 In addition, we fail to see how the minor protective procedures fashioned by the court will impose any substantial burden on the FTC’s investigation 176 — especially since the parties have agreed that this portion of the court’s order may be modified by the addition of the following provisos:
Provided that, nothing in this order shall prohibit disclosure of materials produced by respondents to commissioners of the FTC in connection with the performance of said commissioners’ official functions; nothing in this order shall prohibit employees of the Trade Commission from referring to or relying on any of the materials produced by respondents in connection with the presenting of any recommendation to the Trade Commission for or against issuance of a complaint; and nothing in this order shall prohibit the Trade Commission from referring to or relying on any of the materials produced by respondents in connection with any determination for or against issuance of *464any complaint based on such materials. (18(a))
******
In the event the Trade Commission desires to release any confidential material, it shall so notify the Court and each affected respondent, specifying the material it desires to release, and each such respondent may, within ten days of receipt of such notice, file with the Court opposition to such release. The respondents shall bear the burden of proving the material is entitled to confidential treatment upon notice that the Trade Commission intends to release any such material. No such data may, however, be released until this Court shall enter its order permitting such release. (¶ 8(b))
Provided that, in the event the Trade Commission issues a formal complaint, at the conclusion of the investigation, the confidential material produced by respondents may be offered and received in evidence in the complaint proceeding, provided that each affected respondent shall be given opportunity to request appropriate confidential treatment of such material. (¶ 8(c))177
We would adopt these modifications (as the majority opinion does to some extent, e. g., the ten day notice requirement) and hold that the District Court did not abuse its discretion when it attached the conditions listed above to the disclosure of information by all seven appellees.
The Trade Commission also complains about the option permitting the production of documents for inspection where they are stored. The Supreme Court in CAB v. Hermann upheld the enforcement of a subpoena “with appropriate provisions for assuring the minimum interference with the conduct of the business of respondents.178 The Second Circuit has also upheld a similar provision. It noted that “[requiring records to be produced away from the place where they are ordinarily kept may impose an unreasonable and unnecessary hardship which in itself would make the issuance of the subpoena, otherwise proper, arbitrary and capricious.” 179 Since the number of documents to be produced will be quite large, it is not inappropriate to relieve appellees of some of the expense and burden entailed by permitting them the option of producing documents where they were stored180 and requiring the Trade Commission to copy and transport to Washington any documents they consider useful. Adding the one agreed-upon modification, we would affirm the District Court’s in situs restriction.
VIII. THE SUPERIOR ORDER
The Commission concedes that Superior does not report reserve estimates to, or participate in, the work of the AGA. Since the FTC issued identical subpoenas to all producers, it was obvious to the District Court that several specifications, those relating to reporting and participating in the AGA, were not relevant to Superior. As a result, the District Court simply refused to enforce specifications G through J. In all other respects, the court required Superior to provide the Commission with the same documents and information that were being required of the other six producers, and subject to the same confidentiality protec*465tions previously discussed. We do not understand, and the majority opinion does not bother to explain, how this logical factual determination can be viewed as clearly erroneous.
IX. CONCLUSION
The preceding pages set forth in detail our differences with the majority opinion. In our view, the majority is not only wrong in its evaluation of the details of this case and in its misconception of the economic facts of life in the gas industry, but, more importantly, the majority has abandoned any standard of appellate review of the factual determinations of a trial court in proceedings to enforce an administrative subpoena.
It cannot be denied that questions of relevance and burdensomeness are peculiarly within the ken of the trial court, and the recognized standard is that the district court’s findings are to be upset only if they represent clear error or an abuse of discretion. The majority apparently feels this standard would be too restrictive of what it is resolved to do. So, instead, it cavalierly labels the trial court’s findings of fact as questions of law and proceeds to substitute its judgment for that of the trial court.
Our concern over the majority’s approach thus derives, not from a disagreement over the application of the accepted standard of appellate review to the particular factual setting in this case, but rather from a more basic apprehension over the consequences of the undefined standard of review which was employed.
The majority has engaged in a standard-less, directionless review in this case, and no euphemism can disguise this embarrassing fact. The majority opinion demonstrates this assertion by failing to even define the purpose of the FTC investigation which is being subjected to de novo review, although the trial court had elucidated this quite well from FTC counsel, taking the original “We’re Going to Investigate the Gas Industry” resolution of the FTC as a beginning. The precedents clearly establish this definition of purpose as the starting point for analysis; if the majority chooses to disregard these precedents, the dictates of logic should serve as an adequate substitute. The failure to focus on the FTC’s purpose in turn causes the majority to roam into those areas committed by precedent to, and more appropriate for, the district court. A particularly striking example of this is to be found in the majority’s decision to ignore the scientific and economic realities of the gas industry which the District Court correctly took into account.
The enlarged role which the majority has assigned to this court in this case distorts the proper relationship between the federal agencies, the federal trial courts, and the federal appellate courts. In so doing, the majority sets a pernicious precedent for future trials de novo which would leave the Court of Appeals as the primary determinant of factual matters more properly suited for the District Court.
The strength of this sloppy precedent is, of course, weakened by the composition of the en banc court in this case; here the majority consists of only four of our colleagues. It is our hope that the approach adopted by this diminished majority of the court will not be carried over into future cases. If it is attempted to be so applied in the future, it will be a divergence from accepted practice of such magnitude that a close examination by our full court will be warranted, if the errors of our four colleagues have not already received their just reward from an even higher authority.
. Throughout this opinion we will use the term “proved reserves.” The following is the definition of proved reserves adopted by the AGA in its annual publication, “Reserves of Crude Oil, Natural Gas Liquids, and Natural Gas in the United States and Canada and United States Productive Capacity,” Volume 28, June 1974. The first two paragraphs of the following definition appear on page 103 of this publication, the third paragraph is derived from page 99 and the last paragraph is derived from pages 96 and 97 (emphasis added):
Proved Reserves are the estimated quantity of natural gas which analysis of geologic and engineering data demonstrate with reasonable certainty to be recoverable in the future from known oil and gas reservoirs under existing economic and operating conditions. Reservoirs are considered proved that have demonstrated the ability to produce by either actual production or conclusive formation test.
The area of a reservoir considered proved is that portion delineated by drilling and defined by gas-oil, gas-water contacts or limited by the structural deformation or lenticularity of the reservoir. In the absence of •fluid contacts, the lowest known structural occurrency of hydrocarbons controls the proved limits of the reservoir. The proved area of a reservoir may also include the adjoining portions not delineated by drilling but which can be evaluated as economically productive on the basis of geological and engineering data available at the time the estimate is made. Therefore, the reserves reported should include total proved reserves which may be in either the drilled or uridrilled portions of the field or reservoir.
Natural gas reserves take into account the shrinkage of the reservoir gas volume resulting from the removal of the liquefiable portions of the hydrocarbon gases and the reduction of volume due to the exclusion of non-hydrocarbon gases where they occur in sufficient quantity to render the gas unmarketable.
The proved reserves estimated are to include all gas reserves regardless of size, availability of market, ultimate disposition or use.
See also note 91 infra.
. These statistics were made available to the FTC subject to an agreement restricting access to and disclosure of the data. The agreement provided, inter alia, that:
(2) Representatives of your Bureau [i. e., the Commission’s Bureau of Competition] will make use of such reports only in connection with its current investigation into the reporting of natural gas reserves by the American Gas Association, and for the purpose of verification of natural gas reserves estimates reported by the A.G.A. Committee on Natural Gas Reserves; and shall not release, disclose, disseminate or publicize in any manner, to any person, any statistic or data contained in such reports, unless otherwise available in published records or documents, without twenty days prior notice, and opportunity to seek appropriate legal protection or relief, to A.G.A. and to each member of the South Louisiana Subcommittee whose statistics are to be disclosed by the Commission.
Supplemental Brief for Appellee Mobil Oil Corp. (filed 4 April 1975), Exhibit A at 2. (Emphasis supplied). The agreement also provided that custody of the documents be maintained by a neutral third party, Price, Waterhouse & Co.
The producers have alleged in supplemental filings before this court that the FTC breached this agreement by releasing to a Congressman certain staff and working papers containing excerpts from the AGA reserve statistics after less than 72 hours’ notice. In a reply the Commission conceded that it had released the information to the Congressman, who by a phone call “required that the documents be immediately released” to him. The Commission argues, however, that the above paragraph was intended to prohibit disclosure of the data to the public or to competitors, and was not meant to cover the case where a member of Congress or congressional committee might immediately require use of the data. In addition, the Trade Commission argues that it lacks the statutory power to keep this data confidential when a Member of Congress demands disclosure.
. Preliminarily, it is necessary to recognize the central importance of specification G of the FTC’s subpoena. This specification, if left unmodified by the District Court’s order of any of the subsequently agreed upon stipulations, calls for the production of
Documents either received (from whatever source) or written by the Company, in whole or in part, at any time between January 1, 1962 to December 31, 1970, which contain estimates or evaluations of the volume of natural gas present or recoverable or ultimately recoverable (1) throughout all of South Louisiana (2) throughout all of Offshore South Louisiana and/or (3) in specific fields, portions of fields, leaseholds and/or portions of leaseholds located in Offshore South Louisiana.
*425Excluded from this specification are any documents previously made available to the Commission by the American Gas Association and presently in the custody of Price, Waterhouse & Company, 1801 K Street, Washington, D. C. Included in this specification by way of illustration but not limitation are documents containing estimates or evaluations including re-estimates or reevaluations made in connection with or in preparation for or as the result of the following: (1) bidding on or nominating leases (2) deciding whether to erect permanent platforms (3) compiling or inventorying total company reserves or supply (4) negotiating or contracting for the sale of natural gas, or for the joint or common exploration, development, production, purchase or sale of acreage, or for obtaining bank loans (5) filing depreciation expense schedules with Internal Revenue Service or (6) submitting field-by-field estimates to subcommittees or committees of the American Gas Association or the American Petroleum Institute.
FTC Subpoena, App. I at 54a. This is the most significant specification of the subpoena not only because of the enormous breadth of its coverage but also because the next two specifications (H and 1(1)) define their breadth by referring back to the documents produced pursuant to specification G. Id. at 55a-56a.
. 338 U.S. 632, 70 S.Ct. 357, 94 L.Ed. 401 (1950).
. FTC v. Loaning, 176 U.S.App.D.C. 200, 208 n.14, 539 F.2d 202, 210 n.14 (1976). See text, infra, at note 39.
. Opinion and Order Determining Just and Reasonable Rates for Natural Gas Produced in the Southern Louisiana Area (So La II), 46 F.P.C. 86, 115 (16 July 1971), aff’d sub nom., Placid Oil Co. v. FPC, 483 F.2d 880 (5th Cir. 1973), aff’d sub nom., Mobil Oil Corp. v. FPC, 417 U.S. 283, 94 S.Ct. 2328, 41 L.Ed.2d 72 (1974).
. App. IV at 807a (¶ 2.d.).
. Id. at 806a-07a (¶ 2.a.-c.) (emphasis added).
. Id. at 807a-08a (¶ 3) (emphasis added).
. Id. at 808a (¶ 7).
. Id. at 809a-10a (¶ 8).
. Id. at 808a (fl 6).
. 338 U.S. 632, 70 S.Ct. 357, 94 L.Ed. 401 (1950).
. App. Ill at 469a-71a.
. Brief for Appellant (filed 30 Aug. 1974) at 17-18.
. United States v. Morton Salt Co., 338 U.S. at 652, 70 S.Ct. 357.
. In addition to So La II, on 23 February 1971 the Power Commission commenced work on the National Gas Reserves Study (NGRS), a massive audit of all United States gas reserves. In its final report, issued May 1973, the Commission concluded that AGA proved reserve estimates slightly overstated total reserves. FPC STAFF REPORT ON NATIONAL GAS RESERVE STUDY (May 1973) at 3, App. VI at 1048a. See infra at p. 430 of 180 U.S.App.D.C., at p. 902 of 555 F.2d. Counting NGRS and So La II, the investigation contemplated by the Trade Commission’s subpoenas would have been the third plenary investigation into the same gas reserve data, i. e., proved reserve estimates in South Louisiana as of 31 December 1970. Since its NGRS report, the Power Commission has continued to monitor closely the accuracy of the AGA reports.
. As we shall see, the two most important modifications made by the District Court — (1) limiting the production of documents to those containing or underlying proved reserves and (2) restricting the use of this information to an investigation of an alleged conspiracy in the reporting of proved reserves — can and should each be affirmed on two of these three grounds.
. Rule 81(a)(3), FED.R.CIV.P., provides that the Federal Rules of Civil Procedure “. . apply to proceedings to compel the giving of testimony or production of documents in accordance with a subpoena issued by an officer or agency of the United States . . ..” We are thus bound by Rule 52(a), FED.R.CIV.P., which states that “[f]indings of fact shall not be set aside unless clearly erroneous . . ..’’
. App. IV at 805a.
.Southern Louisiana Area Rate Cases, 428 F.2d at 421.
This second proceeding was initially denominated the “Offshore Louisiana Area” proceeding to distinguish it from So La I (the “Louisiana Area” proceeding). The impetus for this proceeding was contained in the record of So La I where the FTC said:
. the pleadings do lead us to the conclusion that the vital importance of future additional gas supply from the offshore areas in Southern Louisiana warrants the immediate commencement of the further proceeding we had contemplated in our original Opinion herein looking towards a possible revision of the area price ceilings for such gas.
41 F.P.C. at 307. The Court of Appeals for the Fifth Circuit expressly endorsed the FPC’s new initiative commenting on “new evidence of a possible impending gas supply shortage” and leaving the proceeding open to explore this development. See In re Southern Louisiana Area Rate Cases, 444 F.2d 125, 126 (5th Cir. 1970). This new proceeding initially focused strictly on the offshore area, 41 F.P.C. at 307-08, but was later enlarged to include onshore
Southern Louisiana, 42 F.P.C. 1110. The order officially beginning the record proceedings was entered on 20 March 1969. 41 F.P.C. 378. The FPC subsequently stayed the effectiveness of its So La I orders at the request of the Court of Appeals for the Fifth Circuit. 41 F.P.C. 675--76. This stay was continued on 15 December 1969. 42 F.P.C. 1110, 1111. When the Fifth Circuit issued its opinion in June 1970, it expressly left the So La I record open for “retrospective as well as prospective adjustments.” 444 F.2d at 126-127. Since the FPC had never made its orders fully effective, it was legally and conceptually logical for it to consolidate So La I with the new proceeding, and it did so on 24 December 1970, 44 F.P.C. 1638, in order to assess supply factors which it had ignored in So La I.
. Order Enlarging Investigation and Proposed Rulemaking Area Rate Proceeding (Southern Louisiana Area), 42 F.P.C. 1110, 1112 (15 December 1969).
. See MDG’s Initial Brief in So La II at 15.
.As part of its examination of the false shortage allegations, the FPC had conducted a spot audit of natural gas reserves requiring producers to report uncommitted reserves in the Southern Louisiana area. A composite of this data was admitted into evidence in So La II after the FPC had closely audited both the composite and the underlying individual responses of the producers. The data underlying the producers’ uncommitted reserve estimates, (e. g., electric and other technical logs, core analyses, formation tests, shut-in and flowing pressure tests, structure maps, isopachus maps, directional surveys, daily drilling records, etc.) was also thoroughly scrutinized by the FPC auditing team. See 43 F.P.C. 444-48 (1970). According to appellees, “In most of the analyses, the audit team derived independent factors for estimating reserve volumes and made its own reserve estimate based upon these independent factors.” Supplemental Memorandum for Appellees Texaco Inc., Standard Oil Co. (Indiana), Shell Oil Co., Exxon Corp., and Mobil Oil Corp. (filed 13 April 1976) at 6. The FPC staff members in charge of the So La II audit concluded that APA reserve data was accurate and established “. . . beyond any doubt that ... a serious gas supply shortage does in fact exist throughout the nation’s gas supply areas, and in Southern Louisiana in particular.” Hearings on Concentration by Competing Raw Fuel Industries in the Energy Market and its Impact on Small Business Before the Subcomm. on Special Small Business Problems of the House Select Comm, on Small Business (hereinafter 1971 House Concentration Hearings), 92d Cong., 1st Sess. A43 (1971). Of the AGA data, the FPC staff also stated,
Certain parties have questioned the reliability of the AGA data in these proceedings . However, the record establishes the validity and reliability of the reported AGA reserves data beyond any reasonable doubt.
Id. at A56. See also id. at A72.
. Opinion and Order Determining Just and Reasonable Rates for Natural Gas Produced in the Southern Louisiana Area, 46 F.P.C. 86, 113 (16 July 1971), aff’d sub nom., Placid Oil Co. v. FPC, 483 F.2d 880 (5th Cir. 1973), aff’d sub nom., Mobil Oil Corp. v. FPC, 417 U.S. 283, 94 S.Ct. 2328, 41 L.Ed.2d 72 (1974). In affirming, the Supreme Court expressly approved the FPC’s reliance on the non-cost factor of supply and noted that both the Commission and the Fifth Circuit were working “. . . against the background of a serious and growing domestic gas shortage . . 417 U.S. at 320, 94 S.Ct. at 2352.
. Id. at 116.
. See 46 F.P.C. at 113.
. Placid Oil Co. v. FPC, 483 F.2d 880, 894 n. 13 (5th Cir. 1973), aff’d sub nom., Mobil Oil Corp. v. FPC, 417 U.S. 283, 94 S.Ct. 2328, 41 L.Ed.2d 72 (1974).
. FPC STAFF REPORT ON NATIONAL GAS RESERVE STUDY (May 1973) at 3, App. VI at 1048a.
. Chapman v. Maren Elwood College, 225 F.2d 230, 234 (9th Cir. 1955); accord, NLRB v. Northern Trust Co., 148 F.2d 24, 29 (7th Cir. 1945).
. Supplemental Brief for Appellant on Rehearing En Banc (filed 31 Mar. 1976) at 5 (emphasis added).
. See the last paragraph of both District Court Orders, App. Ill at 471a & IV at 810a.
. Oklahoma Press Publishing Co. v. Walling, 327 U.S. 186, 208-09, 66 S.Ct. 494, 505-506, 90 L.Ed. 614 (1946).
. See v. City of Seattle, 387 U.S. 541, 544, 87 S.Ct. 1737, 18 L.Ed.2d 943 (1967).
. United States v. Morton Salt Co., 338 U.S. 632, 652, 70 S.Ct. 357, 369, 94 L.Ed. 401 (1950) (emphasis added). The majority correctly decides that this is the applicable standard. Court’s opinion note 23 & accompanying text. Previously, however, in its “II. Court Enforcement of Administrative Subpoenas — The Applicable Legal Principles,” the majority also relied upon Endicott Johnson Corp. v. Perkins, 317 U.S. 501, 63 S.Ct. 339, 87 L.Ed. 424 (1943), and Oklahoma Press Publishing Co. v. Walling, 327 U.S. 186, 66 S.Ct. 494, 90 L.Ed. 614 (1946). As we demonstrate in Part VI-A, infra, pp. 455-456 of 180 U.S.App.D.C., pp. 927-928 of 555 F.2d these two cases deal with the agency’s jurisdiction under their basic statutes, on which judicial review may logically be postponed until the administrative proceeding is completed, while here we deal with issues of relevance, burdensomeness, and collateral estoppel, which if not decided at the outset may become, in effect, moot by the time the companies have gone to the time and expense of baring their confidential files. See Safir v. Gibson, note 131, infra.
. Oklahoma Press Publishing Co. v. Walling, 327 U.S. at 217 n. 57, 66 S.Ct. at 509 n. 57.
. See United States v. Nixon, 418 U.S. 683, 702, 94 S.Ct. 3090, 3104, 41 L.Ed.2d 1039 (1974):
Enforcement of a pretrial subpoena duces tecum must necessarily be committed to the sound discretion of the trial court since the necessity for the subpoena most often turns upon a determination of factual issues.
. FTC v. Lonning, 176 U.S.App.D.C. 200, 208 n. 14, 539 F.2d 202, 210 n. 14 (emphasis supplied).
. 148 F.2d 24, 29 (7th Cir.), cert. denied, 326 U.S. 731, 66 S.Ct. 38, 90 L.Ed. 435 (1945) (emphasis added); accord, Lynn v. Biderman, 536 F.2d 820, 824 & 826 (9th Cir. 1976).
. The fact that a subpoena is issued in an investigative setting does not diminish this requirement; respondents to any administrative subpoena may object on grounds of relevance. See United States v. Associated Merchandising Corp., 261 F.Supp. 553, 560 (S.D.N.Y.1966) (“It is clear, even in the investigative cases, that respondents may object to the subpoena on the ground of relevance and upon the ground of oppressiveness, i. e., the undue burden of compliance.”)
. 5 U.S.C. § 554(b) (1970).
. Montship Lines, Ltd. v. Federal Maritime Bd., 111 U.S.App.D.C. 160, 167-168, 295 F.2d 147, 154-55 (1961); Hellenic Lines, Ltd. v. Federal Maritime Bd., 111 U.S.App.D.C. 151, 153, 295 F.2d 138, 140 (1961); FTC v. Green, 252 F.Supp. 153 (S.D.N.Y.1966).
.During the hearing of 13 December 1973 the following exchange took place between Judge Hart and FTC counsel, Gerald Harwood:
THE COURT: Wait a minute. You are trying to prove that they conspired together to make certain returns with regard to reserves that were incorrect, aren’t you? Is that what you are trying to say?
MR. HARWOOD: That was one of the subjects we are investigating, yes.
THE COURT: What else are you investigating?
MR. HARWOOD: Your Honor, the resolution states that we are engaged—
THE COURT: Don’t tell me what the resolution says. Tell me what you are trying to do. I don’t understand the resolution.
App. II at 399a-400a.
.When, as here, the District Court decides not to refuse enforcement because of an insufficient statement of purpose, the court must determine for itself the agency’s actual purpose. In making this essentially factual determination the District Court may look to the agency’s statement of purpose (here the FTC’s authorizing resolution quoted in majority opinion at p. 396 of 180 U.S.App.D.C., at p. 868 of 555 F.2d) and any other relevant evidence, but neither the District Court nor this court may rely on its own or counsel’s post hoc rationalizations. As the Supreme Court explained in Burlington Truck Lines, Inc. v. United States,
The courts may not accept appellate counsel’s post hoc rationalizations for agency action; [SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947)] requires that an agency’s discretionary order be upheld, if at all, on the same basis articulated in the order by the agency itself: “[A] simple but fundamental rule of administrative law . is . that a reviewing court, in dealing with a determination or judgment which an administrative agency alone is authorized to make, must judge the propriety of such action solely by the grounds invoked by the agency. If those grounds are inadequate or improper, the court is powerless to affirm the administrative action . . . .” Ibid.
For the courts to substitute their or counsel’s discretion for that of the Commission is incompatible with the orderly functioning of the process of judicial review. This is not to deprecate, but to vindicate (see Phelps Dodge Corp. v. Labor Board, 313 U.S. 177, 197, 61 S.Ct. 845, 853, 854, 85 L.Ed. 1271), the administrative process, for the purpose of the rule is to avoid “propel[ling] the court into the domain which Congress has set aside exclusively for the administrative agency.” 332 U.S., at 196, 67 S.Ct. 1575.
371 U.S. 156, 168-69, 83 S.Ct. 239, 246, 9 L.Ed.2d 207 (1962).
. Transcript of the 13 Dec. 1973 Hearing, App. II at 352a & 360a.
. Response to Senator Hart — FTC Investigation of Reporting of Natural Gas Reserves, June 1973, App. IX at 1690a-91a (emphasis added). See also Statement of .James T. Helverson, Director, Bureau of Competition, FTC, Before the Subcomm. on Antitrust and Monopoly of the Senate Judiciary Comm., 27 June 1973, App. IX at 1738a-39a.
. See note 171 infra.
. As recently as 28 July 1975, in a letter to Senator Hart from then-Chairman Engman, the FTC has reconfirmed that the purpose of its investigation is “to determine whether proved gas reserves have been underreported, either by collusion or by individual action.” Memorandum in Support of Supplemental Opposition by Appellant, FTC, to Motion by Appellee, Superior Oil Co., for Permission to File Supplemental Memorandum (filed 1 Aug. 1975) (Attachment).
One regrettable obfuscation of what the FTC is seeking and what is relevant to the investigation’s purpose is repeated in the majority opinion. It refers to “all reserve estimates made by the producers, both for various internal business purposes and for reports to the AGA.” Maj.Op. at 401 of 180 U.S.App.D.C., at 873 of 555 F.2d. The same false distinction is made at p. 396, at p. 868 of 555 F.2d, comparison of the various estimates used by producers in their internal procedures and business operations with those reported as proved estimates to the AGA”; and again at p. 403, at p. 875 of 555 F.2d, “ . . .to compare estimates prepared for various business purposes with those reported to the AGA.” The implication appears that the producers keep two sets of books, for their own use and for reporting to AGA. The proved reserve figures are prepared for “internal business purposes” just as much as any other calculations; more so, in fact, for without them no gas company could accurately manage its business, and proved reserve calculations were made long before AGA existed. These proved reserve figures are now reported to AGA, and they are the only figures ever so reported. The other calculations are various preliminary estimated reserves, derived in different companies from different assemblages of scientific data, and not commonly defined. These, too, are prepared for “internal business purposes,” just as are proved reserve calculations, but at a much earlier stage in the gas exploration or prospecting process. These are never reported to AGA, because they are abandoned for internal business purposes as soon as “proved reserve” figures become available for each individual tract.
The lack of understanding of the FTC as to the total non relationship of these preliminary estimates to proved reserves, which alone are reported, and the resulting total irrelevancy of these preliminary estimates to any purpose of the FTC in investigating false or collusive reporting, are discussed in Part B. infra. The immediate point here is to make clear that all producer reserve data is prepared for internal business purposes, though only proved reserve data is reported to AGA.
The subpoenas, as modified by the District Court and the stipulations, cover every document which relates to proved reserve data. As is apparent from the purpose of the FTC investigation, discussed above, the modified subpoenas will cause the production by the companies of every document, prepared for any purpose, relating to proved reserves. The subpoenas will not cause the production of documents concerning data unrelated to proved reserves and therefore never reported to the AGA, as such material is irrelevant to the FTC inquiry purpose.
.Maj.Op. 404, n. 29, 876 of 555 F.2d, n. 29.
. Note 44, supra.
. Maj.Op. 404, n. 29, 876 of 555 F.2d, n. 29 (emphasis added).
. Id. at 413, at 885 of 555 F.2d (emphasis added). See also id. at 406, at 878 of 555 F.2d.
. Transcript of 13 Dec. 1973 Hearing, App. II at 431a (emphasis added).
. Maj.Op. 402 & 406, 874 & 878 of 555 F.2d.
. Maj.Op. 402, 874 of 555 F.2d.
. Id
. Maj.Op. 405, 877 of 555 F.2d
. Maj.Op. 402, 874 of 555 F.2d (emphasis in original).
. Id at 404 n. 29, at 876 of 555 n. 29 (emphasis in original).
. Id
. Id. at 405, 877 of 555 F.2d.
. Id. at 394, at 866 of 555 F.2d. See also id. at 399, at 871 of 551 F.2d.
. Six-Company Order, App. IV at 860(a) (11 2.(a)).
. Six-Company Stipulation at 5. Even though the FTC and Superior were “unable to agree on any modifications of the district court’s order of March 22, 1974, concerning Superior which would eliminate or narrow the issues remaining for decision by this Court . . . ” (Superior Stipulation at 1), the FTC did offer to Superior the following modification, among others, as a basis for settling their dispute: “Specification G may be modified to require Superior to produce only documents containing proved natural gas reserve estimates.” FTC’s Statement of Issues and Proposed Modifications re Superior (filed 19 May 1976) at 3 (emphasis added). In light of this offer (which is less demanding than *439the District Court’s Six-Company Order in that it would not require Superior to produce data underlying its proved reserve estimates), it also seems quite safe to assume (as the majority apparently does) (see Court’s opinion note 64 and accompanying text) that the Trade Commission is no longer interested in Superior’s raw field data.
One of the most puzzling and totally unexplained aspects of this case is why the FTC would make this eminently sensible and logical modification offer to Superior, but not to the other six appellees. Perhaps it is because the FTC and the other producers agreed not only that the District Court’s “. . Order may be affirmed insofar as it excludes raw field data . ,” but also because they agreed that “[t]he Order may be modified to expressly exclude maps or other documents relating to the suspected location of natural gas in currently unleased acreage.” Depending on what is and what is not included under the term “raw field data,” these two stipulations, when read together with the proviso agreed to by all the parties for paragraph 2.(a) of the Six-Company Order, could be interpreted to require appellees to produce only those documents containing proved natura! gas reserve estimates, at least for “currently unleased acreage.” This would at least partially explain why the FTC did not bother to make the same “proved reserve” modification offer to the other six appellees as it did to Superior.
Under specification G the FTC seeks, inter alia,
documents containing estimates or evaluations including re-estimates or reevaluations made in connection with or in preparation for or as the result of the following: (1) bidding on or nominating leases (2) deciding whether to erect permanent platforms (3) compiling or inventorying total company reserves or supply (4) negotiating or contracting for the sale of natural gas, or for the joint or common exploration, development, production, purchase or sale of acreage, or for obtaining bank loans (5) filing depreciation expense schedules with Internal Revenue Service or (6) submitting field-by-field estimates to subcommittees or committees of the American Gas Association or the American Petroleum Institute.
FTC Subpoena, App. I at 54a. The stipulations filed by the parties do not indicate which or what portion of these categories of documents can be characterized as “raw field data.” Paragraph 2.(a) of the Six-Company Order reads as follows:
Provided that, to the extent otherwise called for, any and ail proved reserve data contained in bid calculation files and relating to tracts covered by subparagraph 2(c) [/. e., fields in which the respondent had an ownership interest as of the date reports were submitted by the Southern Louisiana Subcommittee of the AGA Committee on Natural Gas Reserves (FTC Subpoena, fl 2(c)) or fields in which an employee of such respondent had reporting responsibility to this Subcommittee (Six-Company Stipulation at 3)] shall be produced.
Six-Company Stipulation at 2 (emphasis supplied).
. Superior Stipulation at 1.
. Bid files are merely collections of documents developed for, and used in, nominating and bidding for the right to lease tracts for oil and gas exploration (here limited to Offshore Louisiana). They contain only data assembled pri- or to ownership of a tract. Such files consist of:
(1) Speculative reserve estimates (based not on drilling, but on highly speculative tests such as measurement of deviations in the earth’s magnetic and gravitational fields);
(2) Backup data for such estimates, consisting of raw geological and geophysical data and interpretations thereof; and
(3) Documents reflecting the methodology employed by Appellees in developing bids.
Supplemental Memorandum for Appellees Texaco, Inc., Standard Oil Co. (Indiana), Shell Oil Co., Exxon Corp., and Mobil Oil Corp. (filed 13 April 1976) at 35. Accord, Affidavit of H. R. Hirsch at 4-7, appended to Brief of Appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
. Mobil Oil Co. refers in its brief to “the all too frequent cases in which such information has been stolen and sold for large sums of money. See, e. g., Tlapek v. Chevron Oil Co., 407 F.2d 1129 (8th Cir. 1969); Abbott v. United States, 239 F.2d 310 (5th Cir. 1956); Hunter v. Shell Oil Co., 198 F.2d 485 (5th Cir. 1952); Ohio Oil Co. v. Sharp, 135 F.2d 303 (10th Cir. 1943).” Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974) at 27.
. See Affidavit of H. R. Hirsch at 9-11, appended to Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
. Six-Company Stipulation at 2.
. Reply Brief for Appellant (filed 23 Dec. 1974) at 9-10 (footnote omitted).
. Transcript of 13 Dec. 1973 Hearing, App. II at 431a. The “one exception” which the court provided for is discussed in more detail infra at note 76 and accompanying text, and pp. 409-410 of 180 U.S.App.D.C., pp. 881-882 of 555 F.2d.
. 1974-1 Trade Cases ¶ 74,964 (D.D.C.1974).
. Stipulation, FTC v. Continental Oil Co., No. 74-1552 (D.C.Cir., filed 18 June 1974).
. Supplemental Memorandum for Appellees Texaco Inc., Standard Oil Co. (Indiana), Shell Oil Co., Exxon Corp. and Mobil Oil Corp. (filed 12 April 1976) at 34.
. FTC Statement of Issues and Proposed Modifications re Superior (filed 19 May 1976) at 3 (emphasis added).
. Supplemental Brief for Appellant on Rehearing En Banc (filed 31 March 1976) at 29-30.
. Affidavit of Theodore L. Lytle, Jr., App. IV at 633a.
. See our discussion of the FTC’s delay theories infra at note 83 and accompanying text, and pp. 415-417 of 180 U.S.App.D.C., pp. 887-889 of 555 F.2d.
. In fact, the only expert testimony before the court indicates that “[o]nce a lease is purchased, . , [d]rilling commences promptly . . . Affidavit of H. R. Hirsch at 13, appended to Brief of Appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
. App. Ill at 471a and App. IV at 810a.
. Transcript of 13 Dec. 1973 Hearing, App. II at 431a (quoted in its entirety in text accompanying note 72 supra).
. The FTC’s subpoena as modified by the District Court’s order will still yield:
(1) documents indicating the date drilling commenced in specific fields, portions of fields, leaseholds, and/or portions of leaseholds which underlie proved natural gas reserve estimates (specification H.5.d);
(2) documents which refer, analyze, compare, comment on, set forth, and/or relate to any lease nominations or bids which underlie proved natural gas reserves (specification 1.3); and
(3)documents which refer, analyze, compare, comment on, set forth, and/or relate to any failures or delays, for whatever reason, in reporting proved reserves of natural gas to the AGA, including any failures or delays by personnel of the AGA to identify to subcommittee members all fields containing proved reserves (specification J.I.).
The District Court did order that “only the immediate data used to make the [proved reserve] estimate need be submitted” (Six-Company Order 1i 2.a.), but we would not read this limitation to exclude any of the above documents; nor would we characterize any of the above information as “raw field data.”
. Supplemental Brief for Appellant on Rehearing En Banc (filed 31 Mar. 1976) at 31-32.
. Id. at 32.
. Id. at 32-33.
.In response to this argument made by the FTC in the District Court, Mobil undertook a search for any proved reserve data in its bid files and found none. Brief for Appellee Mobil Oil Corp. (filed 8 Nov. 1974) at 17-18; Transcript of 13 Dec. 1973 Hearing, App. II at 422a. At the 13 Dec. 1973 hearing counsel for Mobil also persuasively discounted the significance of any proved reserve data it or any other company might have for a tract leased by a competitor:
[I]f somebody buys a tract next to a tract that we have developed, that means that they are going to develop that tract as quickly as they can because if there is an overlap in the reservoir between the two tracts and they don’t immediately drill wells on it, then Mobil’s developed tract will drain the contiguous tract. So in every situation where that may occur our reserve estimates would be immediately superseded [and] completely obsolete because whoever purchased the adjoining tract would drill wells on it themselves.
Transcript of 13 Dec. 1973 Hearing, App. II at 422-23a.
. Six-Company Stipulation at 2.
. Supplemental Memorandum for Appellees Texaco Inc., Standard Oil Co. (Indiana), Shell Oil Co., Exxon Corp. and Mobil Oil Corp. (filed 13 April 1976) at 39.
. Affidavit of H. R. Hirsch at 11, appended to Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
. See definition of “proved reserves at note 2 supra. We recognize that “[t]he proved area of a reservoir may also include the adjoining portions not delineated by drilling but which can be evaluated as economically productive on the basis of geological and engineering data available at the time the estimate is made.” (This, is, of course, the source of the FTC’s adjacent tract argument discussed in the immediately preceding section of this dissent.) This, however, does not mean that drilling is not an essential ingredient in the definition of “proved reserves.” As both the FTC and appellees agree, it is possible that the drilling of one portion of a field or reservoir will yield sufficient “geological and engineering data” to permit the calculation of proved reserves for an adjoining portion. It is nevertheless, impossible to come up with a proved reserve estimate without some drilling — at least on an adjoining portion of the field or reservoir. It should also be noted that the AGA’s proved reserve definition (accepted by all parties) requires the reporting of such adjacent-tract proved reserves: “[T]he reserves reported should include total proved reserves which may be in either the drilled or undrilled portions of the field or reservoir.” (Emphasis added.) The important point, for our purposes, is that there always is a drilled portion. .There may or may not be undrilled portions yielding proved reserve estimates.
. There is no consistently used or accepted definition for reserves other than “proved reserves.” Various companies describe their data, at different stages, as “probable,” “possible,” “recoverable,” and “ultimately recoverable.”
. The FTC’s and the majority’s reasoning presumes that there is some correlation between a producer’s speculative estimates and its actual production history. No such correlation could, or does in fact, exist. On the contrary, the uncontradicted evidence on the record demonstrates that speculative reserve estimates have no “relationship with or bearing on proved reserve estimates made on the basis of drilled well data.” Affidavit of H. R. Hirsch at 10, appended to Brief for appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
By definition “proved" reserves are those reserves whose existence has been established through well logs, core samples, and other concrete data obtained by actual drilling, whereas “speculative” reserve data (the only type of data found in most bid files) are based upon seismic studies, measurements of the changes in the earth’s gravitational and magnetic fields, and other inferential geological techniques. Even after all inferential calculations are completed there is no way, apart from drilling, to determine whether hydrocarbons in any form are present and, if so, whether they exist in producible quantities. Nor is it possible to determine accurately, without drilling, the depth of the reservoir, its thickness, extent, and porosity — factors which bear significantly upon whether, and to what extent reserves are present. Thus, even the use of the term “speculative reserves” or “unproved reserves” is itself “a misnomer because [by definition] no hydrocarbons have yet been found.” See id. at 2, 5 and 11.
. The total irrelevance of speculative preliminary reserve estimates to proved reserves is vividly demonstrated by Exhibits F and G to the Hirsch affidavit. Those exhibits show that, for the two tracts with which they deal, the original speculative estimate was, in one case, less than one-tenth of and, in the other case, more than double the eventual proved estimate. Another example is provided by Mobil’s experience in the 1968 Texas Offshore lease sale. On the basis of what Mobil considered reliable inferential geological information with respect to so-called speculative reserves, it spent $32 million for certain offshore leases. Eighteen wells were drilled without any producible hydrocarbons being found, and Mobil subsequently relinquished all ownership rights in these tracts. See id. at 10 and Exhibits F and G.
. As demonstrated above, two of the premises underlying the FTC’s erroneous theory are also mistaken. Appellees do not “regard their un*445tested estimates to be of a comparable degree of certainty” with their proved reserve estimates, as the FTC asserts. Supplemental Brief for Appellant on Rehearing En Banc at 33 (filed 31 Mar. 1976). Nor do “the same or substantially similar underlying data” support both proved and unproved reserve estimates. Id. at 34. The inferential seismic and geophysical data underlying speculative reserve estimates are not comparable and bear no resemblance to the drilling results upon which proved reserve estimates are based.
.According to the Hirsch affidavit, “Once a lease is purchased, the bid files are of no further use whatsoever. Drilling commences promptly and a whole new set of hard data is obtained from the actual drilling.” Affidavit of H. R. Hirsch at 13, appended to Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974). This comports with common economic sense, and there is certainly no evidence on the record to the contrary.
. Transcript of 13 Dec. 1973 Hearing, App. II at 431a (quoted at p. 440, at p. 912 of 555 F.2d supra).
. This same provision was included in the order entered by Judge Hart in FTC v. Continental Oil Co., 1974-1 Trade Cases ¶ 74,964 (D.D.C.1974), appeal dismissal, No. 74-1552 (D.C.Cir., 2 July 1974). Apparently, the FTC found the provision to be satisfactory since it dismissed its appeal. Stipulation, FTC v. Continental Oil Co., No. 74-1552 (D.C.Cir., filed 18 June 1974). See text accompanying note 65 supra.
. See supra at note 83 and accompanying text.
. Maj.Op. at 404 of 180 U.S.App.D.C., at 876 of 555 F.2d (emphasis added).
. Id. at 405 n. 32, at 877 of 555 F.2d n. 32 (emphasis in original).
. See note 3 supra.
. 338 U.S. 632, 70 S.Ct. 357, 94 S.Ct. 401 (1950).
. We noted earlier that the District Court may not have reached the issue of relevance with regard to all documents that do not relate to proved reserves. Although the bid files certainly fall within this larger category, throughout these proceedings they have been treated and discussed as if they were a discrete class of documents. The record clearly indicates that the District Court gave separate consideration to the bid files and ultimately based those portions of its •final order concerned with bid files on a finding that these files lacked relevance. See Transcript at 13 Dec. 1973 Hearing, App. II at 431a.
. A fortiori, the FTC has not met the higher standard of relevance which must be established when such extremely sensitive documents as bid files are subpoenaed. See, e. g., Hartley Pen Co. v. United States District Court, 287 F.2d 324, 330 (9th Cir. 1961); E. B. Mueller & Co. v. FTC, 142 F.2d 511, 520 (6th Cir. 1944); Zenith Radio Corp. v. RCA, 106 F.Supp. 561, 565 n. 6 (D.Del. 1952); British Oxygen Co., No. 8955 (FTC, 23 April 1974) (Slip op. at 8), aff’d, 29 May 1974; Seeburg Corp., 70 F.T.C. 1809, 1813 (1966); Puget Sound Salmon Canners, Inc., 53 F.T.C. 342, 353 (1956); Maico Co., 51 F.T.C. 1197, 1203 (1955). See generally Henke, Federal Trade Commission Hearings: Rights of a Non-Party to Protect Its Witnesses and Documents, 21 Am.L.Rev. 130, 143 (1971).
Access to the documents reflecting appellees bid methodology and preliminary speculative reserve estimates is strictly limited to executive personnel. In fact, this information is kept in safe deposit boxes or vaults and is generally maintained in handwritten form to limit exposure to clerical personnel. See Affidavit of H. R. Hirsch at 11-12, appended to Brief for Appellee, Mobil Oil Corp. (filed 8 Nov. 1974).
. United States v. Nixon, note 38, supra; FTC v. Lonning, note 39, supra.
. See v. City of Seattle, 387 U.S. 541, 544, 87 S.Ct. 1737, 18 L.Ed.2d 943 (1967) (quoted at p. 431 of 180 U.S.App.D.C., at p. 903 of 555 F.2d supra).
. SEC v. Brigadoon Scotch Distrib. Co., 480 F.2d 1047, 1056 (2d Cir. 1973), cert. denied, 415 U.S. 915, 94 S.Ct. 1410, 39 L.Ed.2d 469 (1974); Adams v. FTC, 296 F.2d 861, 866-67 (8th Cir. 1961), cert. denied, 369 U.S. 864, 82 S.Ct. 1029, 8 L.Ed.2d 83 (1962); Hunt Foods & Indus., Inc. v. FTC, 286 F.2d 803, 811 (9th Cir. 1960), cert. denied, 365 U.S. 877, 81 S.Ct. 1027, 6 L.Ed.2d 190 (1961); Chapman v. Maren Elwood College, 225 F.2d 230 (9th Cir. 1955); Comet Electronics, Inc. v. United States, 381 F.Supp. 1233, 1242 (W.D.Mo.1974), aff’d, 420 U.S. 999, 95 S.Ct. 1439, 43 L.Ed.2d 758 (1975); Genuine Parts Co. v. FTC, 313 F.Supp. 855, 857 (N.D. Ga.1970), aff’d, 445 F.2d 1382 (5th Cir. 1971); In re United Shoe Mach. Corp., 6 F.R.D. 347 (D.Mass. 1947). See also See v. City of Seattle, 387 U.S. 541, 544, 87 S.Ct. 1737, 18 L.Ed.2d 943 (1967).
. FTC v. Lonning, 176 U.S.App.D.C. 200, 209, 539 F.2d 202, 211 (1976); NLRB v. Northern Trust Co., 148 F.2d 24, 29 (7th Cir.), cert. denied, 326 U.S. 731, 66 S.Ct. 38, 90 L.Ed. 435 (1945); FCC v. Cohn, 154 F.Supp. 899, 912 (S.D.N.Y.1957). See also Penfield Co. v. SEC, 330 U.S. 585, 67 S.Ct. 918, 91 L.Ed. 1117 (1947) where all members of the Supreme Court (six in the majority, one concurring, and two dissenting) agreed that an abuse of discretion standard was the appropriate standard for appellate review of a district court’s refusal to grant the coercive enforcement relief sought by the Agency. Justice Douglas, writing for the majority, held that in the absence of a finding by the district court (1) that the Commission’s request had become moot, (2) that the documents produced satisfied the Agency’s legitimate needs, or (3) that the additional documents sought were irrelevant to the SEC’s statutory functions, the district court had committed an “abuse of discretion” in failing to grant the full remedial relief which the Securities Act of 1933 placed behind the Commission’s orders. Id. at 592, 67 S.Ct. 918. Concurring, Justice Rutledge would have “. . remanded to the District Court with directions to exercise its discretion in framing the relief adequate and appropriate to make effective the Commission’s right to disclosure.” Id. at 603, 67 S.Ct. at 928 (emphasis added). Dissenting for himself and Justice Jackson, Justice Frankfurter wrote,
Bearing in mind that the District Court was not an automation which must unquestionably compel obedience to a subpoena simply because the Commission had issued it, we must consider whether the District Court had abused the fair limits of judicial discretion.
On the record before us, Judge Hall exercised allowable discretion in finding that the *449subpoena had spent its force, and in concluding not to compel obedience to it.
Id. at 607-08, 67 S.Ct. at 930 (emphasis added).
. FTC v. Loaning, 176 U.S.App.D.C. 200, 204, 539 F.2d 202, 206 (1976).
. Id. at 209, 539 F.2d at 211 (footnotes and citations omitted) (emphasis added).
. 148 F.2d at 29.
. See our discussion of these FPC proceedings supra at pp. 427-430 of 180 U.S.App. D.C., at pp. 899-902 of 555 F.2d.
. See Westinghouse Elec. Corp. v. City of Burlington, 122 U.S.App.D.C. 65, 70, 351 F.2d 762, 767 (1965); Application of Consumers’ Union of the United States, Inc., 27 F.R.D. 251, 254 (S.D.N.Y.1961).
. Cf. S & E Contractors, Inc. v. United States, 406 U.S. 1, 92 S.Ct. 1411, 31 L.Ed.2d 658 (1972); Safir v. Gibson, 432 F.2d 137 (2d Cir. 1970), cert. denied, 400 U.S. 942, 91 S.Ct. 241, 27 L.Ed.2d 246 (1970).
. Supplemental Brief for Appellee Standard Oil Co. of California on Rehearing En Banc (filed 15 April 1976) at 20.
. Affidavit of L. A. Swanson, App. IX at 1761a-62a. See also Affidavit of John T. Marshall (an employee of Mobil Oil Corp.), App. XI at 2035a-39a.
. See our discussion of “purpose” in section IV-A supra.
. As we demonstrated in Part II supra, the District Court’s factual determination that, for the years covered by the FTC’s subpoena, the Federal Power Commission had already (1) determined natural gas reserves and (2) considered and ruled upon the validity and accuracy of such reserves, was soundly based on evidence in the record.
. See Maj.Op. at 409-410 of 180 U.S.App. D.C. at 881-882 of 555 F.2d.
. Maj.Op. at 410, at 882 of 555 F.2d.
. See our discussion under Part I.B. supra, pp. 425-426, pp. 897-898 of 555 F.2d.
.Part III supra, pp. 431-133, pp. 903-905 of 555 F.2d.
. Joint Brief for Appellees Texaco, Inc., Standard Oil Co. (Indiana), Shell Oil Co., and Exxon Corp. (filed 5 Nov. 1974), at 3 (emphasis in original).
. See note 18 supra.
. United States v. Utah Construction Co., 384 U.S. 394, 421-422, 86 S.Ct. 1545, 16 L.Ed.2d 642 (1966); Sunshine Anthracite Coal Co. v. Adkins, 310 U.S. 381, 60 S.Ct. 907, 84 L.Ed. 1263 (1940); Pacific Seafarers, Inc. v. Pacific Far E. Lines, Inc., 131 U.S.App.D.C. 226, 231, 404 F.2d 804, 809 (1968), cert. denied, 393 U.S. 1093, 89 S.Ct. 872, 21 L.Ed.2d 784 (1969) (“Principles of collateral estoppel may properly be applied in administrative cases.”) accord, Fairmont Aluminum Co. v. Commissioner, 222 F.2d 622, 627 (4th Cir.), cert. denied, 350 U.S. 833, 76 S.Ct. 76, 100 L.Ed. 748 (1955); Tampa Phosphate R.R. v. Seaboard Coast Line R.R., 418 F.2d 387, 399 (5th Cir. 1969), cert. denied, 397 U.S. 910, 90 S.Ct. 907, 25 L.Ed.2d 90 (1970); International Wire v. Local 38, Int’l Bhd. of Elec. Workers, 357 F.Supp. 1018 (N.D.Ohio 1972), aff’d, 475 F.2d 1078 (6th Cir.), cert. denied, 414 U.S. 867, 94 S.Ct. 63, 38 L.Ed.2d 86 (1973).
.384 U.S. at 421-22, 86 S.Ct. at 1559-1560.
.Sunshine Anthracite Coal Co. v. Adkins, supra, 310 U.S. at 402-03, 60 S.Ct. at 916-17. See also French v. Rishell, 40 Cal.2d 477, 254 P.2d 26 (1953) (en banc).
Another requirement for the application of collateral estoppel is that “[t]he determination made of the issue in the prior action must have been necessary and essential to the resulting judgment.” IB J. Moore, Federal Practice ¶ 0.443[1] (2d ed. 1974). Apparently acknowledging that a determination of the validity and accuracy of AGA proved reserve estimates was necessary and essential to the FPC’s ratemaking task, the FTC does not argue this issue in this appeal.
. 113 F.2d 583 (8th Cir. 1940).
. 141 F.2d 141 (7th Cir. 1944).
. 432 F.2d 137 (2d Cir. 1970), cert. denied, 400 U.S. 942, 91 S.Ct. 241, 27 L.Ed.2d 246 (1970).
. Id. at 143 (citations omitted).
. Six-Company Order, App. IV at 807a (II 2.(d)). For a similar “use” restriction formulated in another investigative subpoena enforcement proceeding, see Lynn v. Biderman, 536 F.2d 820 (9th Cir. 1976).
. Supplemental Brief for Appellant on Rehearing En Banc (filed 31 March 1976) at 10.
. A. Duda & Sons Cooperative Ass’n v. United States, 495 F.2d 193, 197 (5th Cir. 1974), quoting from Painters Dist. Council No. 38, Bhd. of Painters, Decorators and Paperhangers v. Edgewood Contracting Co., 416 F.2d 1081 (5th Cir. 1969).
. 317 U.S. 501, 63 S.Ct. 339, 87 L.Ed. 424 (1943).
. 327 U.S. 186, 66 S.Ct. 494, 90 L.Ed. 614 (1946).
. Such as FMC v. Port of Seattle, 521 F.2d 431 (9th Cir. 1975); SEC v. Savage, 513 F.2d 188 (7th Cir. 1975); SEC v. Brigadoon Distrib. Co., 480 F.2d 1047 (2d Cir. 1973), cert. denied, 415 U.S. 915, 94 S.Ct. 1410, 39 L.Ed.2d 469 (1974); FTC v. Gibson, 460 F.2d 605 (5th Cir.).
.This, as the Supreme Court recognized in Oklahoma Press, was the gravamen in both Endicott Johnson and Oklahoma Press:
[I]n the Endicott Johnson case [we] hold that under the Walsh-Healy Act . the district court was not authorized to decide the question of coverage or, on the basis of its adverse decision, to deny enforcement to the Secretary’s subpoena seeking relevant evidence on that question, because Congress had committed its initial determination to [the Secretary]
We think . . . that Congress has authorized the Administrator, rather than the district courts in the first instance, to determine the question of coverage in the preliminary investigation of possibly existing violations .
327 U.S. at 211 & 214, 66 S.Ct. at 507 & 508.
. Endicott Johnson Corp. v. Perkins, 317 U.S. at 503, 63 S.Ct. at 340 (quoting from WalshHealy Act § 5).
. Id. at 507, 63 S.Ct. at 342.
. Id.
. But even if the determination of the gas supply issue was not primarily the duty of the FPC, the Trade Commission should still be es-topped from re-investigating and relitigating (i. e., collaterally attacking) the prior findings of fact of an equally (if not more) competent sister agency. On the facts of this case, application of collateral estoppel need not rest upon a primary jurisdiction-like policy. Indeed, before we could hold that the reserves question fell within the primary jurisdiction of the FPC (and correspondingly, outside the subject matter jurisdiction of the FTC), we would probably need to await an initial determination of this jurisdictional question by the FTC. See Safir v. Gibson, 432 F.2d at 144; cf. Endicott Johnson Corp. v. Perkins, 317 U.S. 501, 63 S.Ct. 339, 87 L.Ed. 424 (1943).
.432 F.2d at 143-44 (citations omitted) (emphasis added).
. Dollar Vitamin Plan, Inc., 69 F.T.C. 933, 971 n.2 (1966).
. 532 F.2d 1092 (7th Cir. 1976).
. 532 F.2d 541 (6th Cir. 1976).
. Regrettably, the majority opinion, at 407 of 180 U.S.App.D.C. n. 41, at 879 of 555 F.2d n. 41 and accompanying text, thoroughly confuses what it means by “these jurisdictional questions” and tries to lump in statutory coverage (jurisdiction) and collateral estoppel with the only “jurisdiction” argument the appellees made, i. e., “primary jurisdiction.” Endicott Johnson and that entire line of cases relied on by the majority deal with “jurisdiction” as referring to a statutory coverage issue, which is properly held to be decided by the responsible *458agency and not subject to judicial review until after the administrative proceeding is completed. Utah Construction, Safir, and other decisions relied on in this dissent deal with collateral estoppel by a previous factual issue determination, do not deal with statutory coverage, and hold that administrative collateral estoppel must be considered at the outset of the proceeding, else the sole purpose and benefit of the doctrine will be lost. The issue of “primary jurisdiction” relates to none of the above, has been abandoned by appellees on this appeal, and all reference to it in the majority opinion is most unfortunately misleading and confusing.
.Compare Endicott Johnson Corp. v. Perkins, 317 U.S. at 508 and 509, 63 S.Ct. at 343 where the Court found that
. the District Court refused to order . . . production, tried the issue of coverage itself, and decided it against the Secretary.
To perform her function [the Secretary of Labor] must draw inferences and make findings from the same conflicting materials that the District Court considered in anticipating and foreclosing her conclusions, (emphasis added)
. The Trade Commission itself has relied on something closely akin to the doctrine of administrative collateral estoppel when it works to its advantage. See National Ass’n of Women & Children’s Apparel Salesmen, Inc. v. FTC, 479 F.2d 139 (5th Cir.), cert. denied, 414 U.S. 1004, 94 S.Ct. 360, 38 L.Ed.2d 240 (1973), where, in a collateral proceeding, the FTC relied heavily on a previous NLRB holding.
. 384 U.S. at 422, 86 S.Ct. at 1560 (emphasis added).
. As mentioned earlier, a group of municipal distributor intervenors played a very active role in the proceedings. See So La I, 428 F.2d at 414 and n. 3.
. See So La II, 46 F.P.C. at 113 & 115. Examination of the transcript in So La II (admitted into the record of these proceedings by stipulation, App. Ill at 620a-21a) indicates that witnesses were vigorously cross-examined on the reliability of AGA data and the procedures used in the FPC’s independent NGRS audit. See Transcript of So La II at 4126-61, 4170-93, 4303-34 and 4652-4753. The FPC staff members in charge of the NGRS audit appeared, testified in detail, and were cross-examined on the methods and procedures followed in performing the audit. See Transcript of So La II at 5190-5207A & 5421-72. Other oil and gas experts on the FPC staff testified and were cross-examined on the reliability of AGA reserve data. See Transcript of So La II at 4003, 4055, 4074-75, 4132, 4137 and 4166.
. Brief Reprinted in Hearings on Concentration by Competing Raw Fuel Industries in the Energy Market & Its Impact on Small Business Before the Subcomm. on Special Small Business Problems of the House Select Comm, on Small Business, 92d Cong., 1st Sess. at A72 (1971) (hereinafter 1971 House Concentration Hearings).
. 46 F.P.C. at 113.
. Id. at 116.
. K. Davis, Administrative Law Treatise, § 18.08 at 598 (1958) (footnote omitted and emphasis added).
. 157 U.S.App.D.C. 235, 483 F.2d 1238 (1973).
. See Friendly, Some Kind of Hearing, 123 U.Pa.L.Rev. 1267 (1975).
. A. Duda & Sons Cooperative Ass’n v. United States, 495 F.2d 193 (5th Cir. 1974).
. Id. at 197 (emphasis added).
. Id.
. 46 F.P.C. at 116 (emphasis added).
. See 1971 House Concentration Hearings at A56. See also id. at A72.
. In its supplemental reply brief the FTC states,
This case touches upon two distinct questions which the companies repeatedly confuse or fail to distinguish: is there a natural gas shortage, and have natural gas companies taken action having the purpose or effect of misstating the extent of any such shortage? The questions are related but independent. While the first question is of primary concern to the FPC, the second is the one of particular concern to the FTC.
Reply Supplemental Brief for Appellant on Rehearing En Banc (filed 13 April 1976) at 2. We agree with the FTC; the purpose of their investigation is to determine whether appellees have taken action having the purpose or effect of misstating [i. e., exaggerating through underreporting] the extent of any . shortage.” See also id. at 14 where the FTC explains, “If a complaint were at issue, for example, the FTC (more particularly its staff) would be the proponent of a claim that some of the companies had engaged in concerted activities to underreport their proved reserves."
Thus, if the FTC investigation is on reporting or misstating, then the FTC’s only concern is with proved reserves, because these are the only reserves reported or about which statements have ever been made. (See our discussions of Relevance, Part IV, supra).
. In its supplemental brief on rehearing en banc, the FTC virtually concedes this point with the following admission:
We do not dispute that one of the causes for the FTC’s concern about the possibility of unlawful conduct concerning the reporting of reserves is that such conduct might result in artificial understatement of reserves which, because of the FPC’s reliance upon such data, could result in higher rates.
Supplemental Brief for Appellant on Rehearing En Banc (filed 31 March 1976) at 28-29. The Trade Commission then goes on to claim another “cause for concern”:
[I]t is quite possible that one or more of the companies has engaged in conduct concerning their development and reporting of reserves which has not yet sufficiently ripened so that it would have caused the AGA data to be substantially inaccurate. . . Similarly, one or more of the companies may have attempted to engage in such unlawful conduct without success .
Id. at 29 (footnote omitted). This second alleged “cause for concern” is totally fallacious for at least three reasons. First, only proved reserve estimates are ever reported and the FTC does at least concede that its investigation is aimed in part at “. . the possibility of unlawful conduct concerning the reporting of reserves . ..” (Emphasis added.) Second, since the only reserve data relied upon by the FPC are AGA proved reserve estimates, only these figures can ever have any effect on FPC-determined rates. Third, as we have painstakingly demonstrated in section IV supra, there is no correlation whatsoever between proved reserve estimates and the earlier, speculative, non proved reserve estimates. After drilling (and the immediately following determination whether any proved reserves are present), all former non proved reserve estimates become superseded, irrelevant, and probably misleading. Without some nexus between AGA proved reserve estimates and other non proved reserve data, the FTC’s allegation *462that the producers may be engaged in “. conduct that may eventually affect, but has not yet affected, proved reserves data” makes utterly no sense.
. 358 U.S. 334, 79 S.Ct. 457, 3 L.Ed.2d 354 (1959).
. In RCA the Supreme Court concluded,
[T]he legislative history of the [Federal Communications] Act reveals that the Commission was not given the power to decide antitrust issues as such, and that Commission action was not intended to prevent enforcement of the antitrust laws in federal courts.
Id. at 346, 79 S.Ct. at 464.
. Reply Supplemental Brief for Appellant on Rehearing En Banc (filed 13 April 1976) at 8.
. Tagg Bros. & Moorhead v. United States, 280 U.S. 420, 50 S.Ct. 220, 74 L.Ed. 524 (1930).
. Reply Supplemental Brief for Appellant on Rehearing En Banc (filed 13 April 1976) at 8 (emphasis added).
. Of course, on policy matters affecting the ultimate issue in its ratemaking proceedings (the just and reasonable area rate for gas), the FPC is always free to change its mind according to its current perception of “the public interest.” This, however, does not preclude application of collateral estoppel to underlying findings of fact essential to a determination of the ultimate issue.
. Six-Company Order, App. IV at 809-10a (¶ 8); Superior Order, App. Ill at 469a-71a.
. 16 C.F.R. §§ 3.45, 4.10 and 4.11 (1976). See also 15 U.S.C. § 46(f) (1970).
. See note 3 supra. The action of the Trade Commission, in acceding to the telephoned demand of a Congressman for immediate access to the documents in so short a period of time as to preclude resort to a court, underlines the wisdom and reasonableness of the District Court’s protective order on confidentiality. The Federal Rules on Civil Procedure currently provide for motions to quash duly authorized subpoenas, but not telephone calls.
.In order to protect confidential commercial information and trade secrets, District Courts frequently have required appropriate protection as a precondition to enforcement of FTC investigative subpoenas. See, e. g., FTC v. St. Regis Paper Co., 304 F.2d 731, 732 n. 1 (7th Cir. 1962); Graber Mfg. Co. v. Dixon, 223 F.Supp. 1020 (D.D.C.1963); FTC v. Bowman, 149 F.Supp. 624 (N.D.Ill.) aff'd, 248 F.2d 456 (7th Cir. 1957); FTC v. Menzies, 145 F.Supp. 164, 171 (D.Md.1956), aff’d, 242 F.2d 81, 84 (4th Cir.), cert. denied, 353 U.S. 957, 77 S.Ct. 863, 1 L.Ed.2d 908 (1957). Accord, FTC v. Lonning, 176 U.S.App.D.C. 200, 539 F.2d 202 (1976).
. Six-Company Stipulation (filed 19 May 1976) at 3-4.
. 353 U.S. 322, 323, 77 S.Ct. 804, 805, 1 L.Ed.2d 852 (1957).
. Walling v. American Rolbal Corp., 135 F.2d 1003, 1005 (2d Cir. 1943).
. The Trade Commission’s and the majority’s arguments notwithstanding, it is common for a district court to require the administrative agency to inspect subpoenaed records or documents at the place where they are stored. NLRB v. Friedman, 352 F.2d 545 (3d Cir. 1965); Hunt Foods & Indus., Inc. v. FTC, 286 F.2d at 812; FTC v. Bowman, 149 F.Supp. at 630; FTC v. Menzies, 145 F.Supp. at 171. In addition, the parties have agreed that
each company shall designate no more than one corporate office in each state in which documents located within that state shall be produced for inspection and copying.
Six-Company Stipulation (filed 19 May 1976) at 3.