GOLDBERG, Circuit Judge.
This case involves the right of a company to condemn lands in furtherance of its railroading ambitions after it has been denied a certificate of public convenience and necessity by the Interstate Commerce Commission. We must decide whether a federal district court has the [390]*390power under 49 U.S.C.A. § 1(20)1 to issue an injunction against condemnation proceedings in state courts that are said to violate § 1(18) of the Interstate Commerce Act.2 We also find it necessary to consider what defenses, if any, may be asserted in opposition to the alleged violation.
The history of this case begins in July of 1965 when Tampa Phosphate Railroad Company (Tampa), a corporation not yet operating as a railroad, filed an application with the Interstate Commerce Commission (Commission) questing for authority to construct a railroad under Section 1(18)-1(22) of the Interstate Commerce Act. See 49 U.S.C.A. § 1(18)— 1(22). In its application Tampa declared that the proposed railroad was intended primarily to transport phosphate rock and all products thereof to Port Sutton in Hillsborough County, Florida, from a point some thirty miles away in Polk County, Florida. Tampa prayed for the grant of a certificate authorizing construction of this railroad, and in the alternative for dismissal of its application as beyond the Commission’s jurisdiction. Tampa contended in support of this alternative that (a) its proposed railroad would operate only in intrastate commerce, and (b) that the line would constitute a spur or industrial track within the exemption of 49 U.S.C.A. § 1(22).3
[391]*391After hearings, the Interstate Commerce Commission examiner recommended denial of the application. He held that the Commission did have jurisdiction because the proposed railroad would operate in interstate commerce, and he found that the new trackage would not constitute a spur or industrial track within the meaning of § 1(22). He then denied Tampa’s application, ruling that construction and operation had not been shown to be required by the present and future public convenience and necessity.
Subsequent to the completion of the administrative hearing, the Interstate Commerce Commission found in a decision dated August 7, 1967, that “the findings and conclusions of the hearing examiner on matters of fact and law considered and disposed of in his report and recommended order are in all material respects proper and correct.” The order of the examiner denying Tampa’s application was thereupon affirmed and adopted by the Commission. No appeal was taken from this order.
Following the denial of its application, Tampa directed its full energies toward a series of condemnation suits then in progress in the state courts of Florida. The first and second of these suits had begun on October 13, 1965, in the Circuit Court of Hillsborough County, Florida, while Tampa’s application was still pending before the Commission. The third suit began a short time later in the Circuit Court of Polk County, Florida. In all proceedings Tampa sought easements or rights-of-way for its proposed railroad across existing trackage of the Seaboard Coast Line Railroad Company (Seaboard). As authority for these easements, Tampa relied upon certain Florida eminent domain statutes permitting one railroad to cross the tracks of another railroad. See F.S.A. §§ 73.012, 73.021, 73.151, 360.01, 361.01. Seaboard opposed Tampa’s petition in each state court and sought dismissal of the actions on the grounds, inter alia, that the proposed construction would be in interstate commerce and that the Commission had denied Tampa the requisite certificate. The Circuit Court of Polk County, apparently reserving its ruling on Seaboard’s motion to dismiss, scheduled the matter for trial on December 18, 1967.
Meanwhile, Seaboard had initiated this suit in the United States District Court for the Middle District of Florida. Seaboard there sought an injunction to prevent Tampa from continuing condemnation proceedings against Seaboard’s property, and further sought to enjoin Tampa from constructing or operating any railroad facilities without first procuring a certificate from the Commission. Tampa’s actions were alleged to be in violation of § 1(18) of the Interstate Commerce Act, 49 U.S.C.A. § 1(18), and an injunction was sought under § 1(20) of the Act, 49 U.S.C.A. § 1(20), to prevent further violations.
Tampa responded to Seaboard’s complaint with an answer and with a motion to dismiss. Argument on the motion to dismiss was held on December 12, 1967, and on December 14, 1967, the district court denied Tampa’s motion to dismiss and granted a preliminary injunction against any condemnation of Seaboard’s property or construction thereon until a certificate of convenience and necessity was procured from the Commission. In its findings of fact and conclusions of law, the court found that it had jurisdiction under 28 U.S. C.A. § 89(b), 28 U.S.C.A. § 1337, and 49 U.S.C.A. § 1(18)-1(20). It also found that Seaboard was a proper party in interest within the meaning of 49 U.S.C.A. § 1(20). The court overruled Seaboard’s motion for judgment on the pleadings, holding that the order of the Interstate Commerce Commission was not res judicata on the intrastate or interstate character of the proposed railroad. It then reserved this issue of interstate commerce for independent investigation at a later date. The court also determined that the provisions of 28 U.S.C.A. § 2283 “are not applicable or controlling in this case.”
Further argument was heard on the case on March 6, 1968, and on April 22, [392]*3921968, the court overruled Tampa’s motion to dissolve and vacate the preliminary injunction. Again the court reserved ruling on the interstate or intrastate character of the proposed railroad. Tampa appeals to this court from the denial of its motion to vacate. 28 U.S. C.A. § 1292(a) (1).
We hold that the district court was within its authority in refusing to dissolve the preliminary injunction. We base our decision largely on the view that neither 28 U.S.C.A. § 2283 nor principles of comity present any bar to the issuance of the district court’s injunction. We also find on the record before us that Seaboard “has raised questions going to the merits so serious, substantial, difficult and doubtful, as to make them a fair ground for litigation and thus for more deliberate investigation.” Hamilton Watch Co. v. Benrus Watch Co., 2 Cir. 1953, 206 F.2d 738, 742; Railroad Yardmasters of America v. Pennsylvania Railroad Co., 3 Cir. 1955, 224 F.2d 226, 229. No more definitive examination of the merits is required on appeal from a refusal to vacate a preliminary injunction. See Railroad Yardmasters of America v. Pennsylvania Railroad Co., supra. “* * * [Ojrdinarily the decree of a District Court granting or denying a preliminary injunction will not be disturbed on appeal.” Myers v. Bethlehem Shipbuilding Corp., 1938, 303 U.S. 41, 52, 58 S.Ct. 459, 464, 82 L.Ed. 638. No extraordinary reasons for departure from such rule are here presented.
I.
We deal first with Tampa’s contention that the antiinjunction statute, 28 U.S. C.A. § 2283, bars injunctive process against the further prosecution of condemnation suits in the state courts of Florida. The statute provides:
“A court of the United States may not grant an injunction to stay proceedings in a State court except as expressly authorized by Act of Congress, or where necessary in aid of its jurisdiction, or to protect or effectuate its judgments.”
We note at the outset that § 2283 applies irrespective of whether the federal injunction is directed to the parties or to the state courts. Oklahoma Packing Co. v. Oklahoma Gas & Elec. Co., 1940, 309 U.S. 4, 9, 60 S.Ct. 215, 84 L.Ed. 537. The statute also applies when, as here, the state proceedings are already in progress. Hill v. Martin, 1935, 296 U.S. 393, 403, 56 S.Ct. 278, 80 L.Ed. 293. Under these circumstances the injunction is contrary to the statute unless it falls within one of the exceptions authorized by the statute itself.
The district court found § 2283 inapplicable to the instant case apparently because the injunction could issue “in aid of its jurisdiction.” 4 Argument before the trial court indicates that this exception was invoked on the premise that an injunction against state proceedings is proper when the federal courts have exclusive jurisdiction over the subject matter of the action. Bowles v. Willingham, 1944, 321 U.S. 503, 64 S.Ct. 641, 88 L.Ed. 892; Capital Service v. N.L.R.B., 1954, 347 U.S. 501, 74 S.Ct. 699, 98 L.Ed. 887. Without assaying the merit of this rationale as applied to the instant case, cf. T. Smith & Son, Inc. v. Williams, 5 Cir. 1960, 275 F.2d 397, we hold that the preliminary injunction could issue under the “expressly authorized” exception to § 2283.
Section 2283 permits the issuance of a federal injunction against state court proceedings when “expressly authorized by Act of Congress.” It is well established that an injunctive provision “need not expressly refer to § 2283” in order to come within this exception. Amalgamated Clothing Workers of America v. Richman Brothers Co., 1955, [393]*393348 U.S. 511, 516, 75 S.Ct. 452, 455, 99 L.Ed. 600. It is also clear that the language of the authorizing statute need not “refer to the subject matter of federal stays of state court proceedings.” Dilworth v. Riner, 5 Cir. 1965, 343 F.2d 226, 230. On the other hand, a case does not fall within the exception merely because it involves a statute that permits injunctive relief in general terms. See Baines v. City of Danville, 4 Cir. 1964, 337 F.2d 579, cert. denied, Chase v. McCain, 381 U.S. 939, 85 S.Ct. 1772, 14 L.Ed.2d 702; Sexton v. Barry, 6 Cir. 1956, 233 F.2d 220, cert, denied, 352 U.S. 870, 77 S.Ct. 94, 1 L.Ed.2d 76; Smith v. Village of Lansing, 7 Cir. 1957, 241 F.2d 856; Cameron v. Johnson, S.D. Miss. 1966, 262 F.Supp. 873, 878, aff'd, 390 U.S. 611, 88 S.Ct. 1335, 20 L.Ed.2d 182; Greene v. State of New York, S.D. N.Y.1967, 281 F.Supp. 579, 581. Beyond these general guidelines the rules are less clear. Courts must then consider the nature of the interests protected by the statute authorizing the injunction, Machesky v. Bizzell, 5 Cir. 1969, 414 F.2d 283; Studebaker Corporation v. Gittlin, 2 Cir. 1966, 360 F.2d 692, and the parties authorized to bring suit. See Leiter Minerals, Inc. v. United States, 1957, 352 U.S. 220, 225-226, 77 S.Ct. 287, 291, 1 L.Ed. 267.
In the circumstances of this case we look to Section 1(20) of the Interstate Commerce Act, 49 U.S.C.A. § 1(20), for injunctive authority. That statute reads:
“ * * * Any construction, operation, or abandonment contrary to the provisions of this paragraph or of paragraph (18) or (19) of this section may be enjoined by any court of competent jurisdiction -at the suit of the United States, the Commission, any commission or regulating body of the State or States affected, or any party in interest. * * * ”
There appears to be no serious question that an attempt to condemn lands for the purpose of constructing new trackage thereon without first acquiring a certificate from the Interstate Commerce Commission constitutes “construction” within the meaning of this statute. See Gilmore v. Sandersville Railroad Company, M.D.Ga.1955, 149 F. Supp. 725. Cf. Bingham v. Taylor, 5 Cir. 1926, 12 F.2d 15. Any uncertainty in this regard should be resolved in favor of coverage in view of the national interests to be protected. See this opinion, p. 394, infra. The undecided question is whether these same interests, and their function in the overall regulatory scheme, justify treating § 1(20) as an express exception to Section 2283.
We begin by noting that Section 1(20) authorizes suit by both public and private parties, a strong indication that the Act contemplates supplementing governmental regulation with private action. One court in confronting a similar statutory scheme in § 222(b) (2) of the Interstate Commerce Act has compared such supplemental private actions to private suits under § 15 of the Clayton Act. Baggett Transportation Co. v. Hughes Transportation, Inc., 8 Cir. 1968, 393 F. 2d 710, 718, cert, denied, 393 U.S. 936, 89 S.Ct. 297, 21 L.Ed.2d 272. The court there quoted extensively from Justice Clark’s opinion in United States v. Borden Co., 1954, 347 U.S. 514, 518-519, 74 S.Ct. 703, 98 L.Ed. 903, deriving therefrom the inference that such private suits also serve public interests.
The public interest concept and its relation to § 2283 have been extensively discussed by this court in the recent case of Machesky v. Bizzell, supra. We there pointed out that “One area in which the command of § 2283 has been held to give way is where the United States is a party.” See Leiter Minerals v. United States, supra. We found in the Leiter case authority for the proposition that the rule of comity embodied in § 2283 must yield to “national” or “superior federal interests.” We then cited with approval the case of Studebaker Corp. v. Gittlin, 2 Cir. 1966, 360 F.2d 692, where, in the opinion by Judge Friendly, the court found that § 2283 was not a bar to a private action under the Securities and [394]*394Exchange Act. The court in Gittlin observed that § 2288 would not have prevented an enforcement action by the Securities and Exchange Commission. It therefore reasoned that a different result should not obtain merely because public interests were being asserted by a private party:
“If the policy of the anti-injunction statute is superseded by the need for immediate and effective enforcement of federal securities regulations and statutes, the fact that enforcement here is by a private party rather than the agency should not be controlling.” 360 F.2d at 698.
The public interest concept and the need for effective enforcement of a regulatory scheme are no less applicable to the regulation of railroads than to the regulation of securities. Both are of national concern; both are of national dimensions. This is particularly true of situations like that before us in which a carrier proposes to build new trackage. In such circumstances the Supreme Court has found the strongest of public interests:
“Paragraphs 18 to 22, supra, were considered here in Texas & P. R. Co. v. Gulf, C. & S. F. R. Co., 270 U.S. 266, 46 S.Ct. 263, 70 L.Ed. 578, and were declared to be part of the general plan by which Congress intended to promote development and maintenance of adequate railroad facilities. It was there said, page 277 of 270 U.S., 46 S.Ct. 263, 266: ‘It [Congress] recognized that preservation of the earning capacity, and conservation of the financial resources, of individual carriers, is a matter of national concern; ■ that the property employed must be permitted to earn a reasonable return; that the building of unnecessary lines involves a waste of resources, and that the burden of this waste may fall upon the public; that competition between carriers may result in harm to the public, as well as in benefit; and that, when a railroad inflicts injury upon its rival, it may be the public which ultimately bears the loss. [Cases cited.]
“The Act sought, among other things, to avert such losses.’ ” Western Pac. Cal. R. Co. v. Southern Pacific Co., 1931, 284 U.S. 47, 50-51, 52 S.Ct. 56, 57, 76 L.Ed. 160, 162.
Based on the foregoing, we conclude that Seaboard in bringing the present suit was engaged in protecting and enforcing public as well as private rights, and that the same suit could have been brought by the Commission itself. Under such circumstances the principle of comity embodied in § 2283 must yield to paramount federal interests. Machesky v. Bizzell, supra; Studebaker Corp. v. Gittlin, supra. See also Sheridan & Townley v. Garrison,. 5 Cir. 1969, 415 F.2d 699.
II.
Appellant’s second objection to the preliminary injunction also relates to comity, but to comity as a principle apart from Section 2283. Tampa argues that if the condemnation proceedings in the state courts were in personam, then both federal and state courts have concurrent jurisdiction over the cause of action and comity requires that each suit proceed to judgment undisturbed by injunctive process. See Donovan v. Dallas, 1964, 377 U.S. 408, 84 S.Ct. 1579, 12 L.Ed.2d 409; Kline v. Burke Construction Co., 1922, 260 U.S.. 226, 43 S.Ct. 79, 67 L.Ed. 226. - On the other hand, if the condemnation proceeding were in rem, then the state courts, having first acquired jurisdiction over the “res,” are the only courts entitled to jurisdiction. Princess Lida of Thurn and Taxis v. Thompson, 1939, 305 U.S. 456, 59 S.Ct. 275, 83 L.Ed. 285. This argument is wide of the mark.
The cases cited by appellant all pertain to the situation in which two courts, one federal and the other state, assert jurisdiction over the same cause of action, or, in the case of suits in rem, over the same property. In such circumstances the Supreme Court has found that concurrent [395]*395jurisdiction exists in both courts if the suits are in personam, Donovan v. Dallas, supra; Kline v. Burke Construction Co., supra, but only in the first court if the suit is in rem. Princess Lida of Thurn and Taxis v. Thompson, supra.
The case before us involves neither the same cause of action in both state and federal courts, nor a common res over which the district court had to assume jurisdiction. The state proceedings in the Florida courts were merely condemnation suits brought under local statutes. The federal proceeding was a suit in personam under § 1(20) of the Interstate Commerce Act. Under these circumstances it must be plain that the district court’s injunction did not issue to prevent repetitive suits. Rather, it issued to prevent state proceedings in complete derogation of federal law. Porter v. Lee, 1946, 328 U.S. 246, 250, 66 S.Ct. 1096, 90 L.Ed. 1199; Studebaker Corporation v. Gittlin, supra, at 360 F.2d 698. Cf. Red Rock Cola Co. v. Red Rock Bottlers, 5 Cir. 1952, 195 F.2d 406 at 409.
A situation analogous to the one before us was presented to the Supreme Court in Porter v. Lee, supra. In that case a suit was brought in a federal district court by the Price Administrator pursuant to the Emergency Price Control Act. The Administrator sought to enjoin a landlord from continuing a rent eviction proceeding in the state Justice of the Peace Court since the rents were violative of the Act. The landlord argued: “Since the Justice of the Peace Court action * * * was commenced prior to the Administrator’s injunction proceeding in the federal court, the Justice of the Peace Court had acquired sole power to decide the crucial issue and the federal District Court therefore lacked jurisdiction.” 328 U.S. at 250, 66 S.Ct. at 1099. In rejecting this argument, the Supreme Court indicated that there was no obstacle to injunctive process against the state court proceeding where the jurisdiction of the two courts was over separate subject matter, and where the state proceeding was itself a violation of the Act:
“ * * * Here the landlord’s eviction proceeding in the Justice of the Peace Court clearly was not an enforcement proceeding authorized by the Act. It was, rather, if the allegations of the Administrator proved to be true, a violation of the Act. The state court’s jurisdiction was based on state law and not on § 205 of the Price Control Act. It was therefore not part of the ‘concurrent’ jurisdiction contemplated by § 205. Over the enforcement proceedings contemplated by that section not only did the District Court acquire jurisdiction first, but the state court never acquired any jurisdiction at all. It was consequently within the power of the federal District Court to grant the injunction provided the Government succeeded in proving the merits of its case.” 328 U.S. at 250, 66 S.Ct. at 1099.
We detect no controlling difference between the federal enforcement proceeding in Porter and the federal enforcement proceeding in the case before us. Porter involved a state eviction suit; here we deal with a state condemnation suit. In Porter there was governmental enforcement of public rights; here there is private enforcement of public rights. In Porter there was no concurrent jurisdiction between state and federal courts; hence no conflicts of jurisdiction or problems of comity. Here there is likewise no concurrent jurisdiction. In light of these similarities we find no substance to the argument that comity or jurisdictional barriers prevent the issuance of a federal injunction. We note that comity was no obstacle to an injunction in Porter where the anti-injunction statute was held inapplicable.5 It should therefore [396]*396be no obstacle here. As a general rule principles of comity will not bar a federal injunction where “there are circumstances which bring the case within one of the exceptions to § 2283. * * * ” 1A, W. M. Moore, Federal Practice, ff 0.208 [4] (1965). See Studebaker Corp. v. Gittlin, supra.
III.
We turn finally to the defense of intrastate commerce which Tampa raised in the court below. While strictly speaking this issue goes to the merits of the controversy, not to the propriety of a preliminary injunction, and therefore lies at the periphery of this appeal, see Doeskin Products, Inc. v. United Paper Co., 7 Cir. 1952, 195 F.2d 356, 361; Chicago Great Western Ry. Co. v. Chicago, B. & Q. R. Co., 8 Cir. 1952, 193 F.2d 975, 978, it has nonetheless been fully briefed and argued by the parties in this court, and an extensive brief on the question has been filed by the government.6 We have also before us a ruling by the trial judge that leaves no doubt of his intention to fully try the matter of interstate commerce on remand. Under these circumstances we think the interests in conserving judicial time and the expense of litigation compel us to suggest that the district court’s decision rejecting principles of collateral estoppel may have been premature.
Tampa argued before the district court that investigation of the interstate or intrastate character of its proposed railroad would not constitute a collateral attack upon the order of the Interstate Commerce Commission. Relying on several somewhat dated Supreme Court cases, see Texas & Pacific R. Co. v. Gulf, C. & S. F. R. Co., 1926, 270 U.S. 266, 46 5. Ct. 263, 70 L.Ed. 578; Piedmont & Northern R. Co. v. United States, 1930, 280 U.S. 469, 50 S.Ct. 192, 74 L.Ed. 551; United States v. Idaho, 1936, 298 U.S. 105, 56 S.Ct. 690, 80 L.Ed. 1070; Powell v. United States, 1937, 300 U.S. 276, 57 S.Ct. 470, 81 L.Ed. 643, Tampa insisted that the matter of interstate commerce, since it affected the Commission’s jurisdiction, was open for independent investigation by a federal district court in a proceeding under § 1(20) of the Act even though there was outstanding an order of the Commission denying Tampa a certificate. We must disagree.
Tampa’s cases were all decided during what might be termed the “era of negativity.” During this period the Supreme Court followed the negative order rule which denied courts the power to review agency decisions that imposed no affirmative burdens or obligations upon the applicant. The rule was regarded as an aspect of the doctrine of administrative finality under which courts declined to review interim steps in an administrative proceeding. In Piedmont & Northern R. Co. v. United States, supra, the Supreme Court held that a statutory three judge court did not have jurisdiction under the Urgent Deficiencies Act, 28 U.S.C.A. § 47, to review the denial of a certificate of public convenience and necessity. The court observed that the order denying the certificate “is entirely negative. It is not susceptible of violation and cannot call for enforcement. It does not finally adjudicate the railway’s standing; nor does it enjoin it to do or refrain from doing anything.” 280 U.S. at 477, 50 S.Ct. at 194. Taking the view that the negative order doctrine barred review by a three judge court, the direct appeal to the Supreme Court was dismissed for want of jurisdiction.
[397]*397The court in Piedmont did not discuss the impact of its holding on enforcement proceedings under § 1(20) of the Act which had been at issue some four years earlier in Texas & P. R. Co. v. Gulf, C. & S. F. R. Co., supra. Nonetheless the court made clear in Powell v. United States, supra, that review of the validity of a negative order was available in an enforcement proceeding under § 1(20) of the Act. The Court stated:
“The purpose of section 1 subds. (18) to (22) of the Act was to empower the Commission in proceedings instituted by a carrier proposing to engage in transportation over or by means of an additional or extended line authoritatively to decide whether it would be in the public interest. Unless the project is one covered by section 1(18), the Commission is not authorized by the act to consider whether it is in the public interest and, for lack of jurisdiction to determine that question, it must deny the application. Upon presentation by the carrier of application for a certificate, the Commission, for the purpose of determining whether it is authorized "by the act to consider the merits, may pass incidentally upon the question whether the project is one covered by section 1(18). But the decision of that question is for the court in either a suit to set aside an order granting a certificate or in a suit under section 1(20) to enjoin a violation of section 1(18). The function of the court is to construe that paragraph; that of the commission is to determine whether the project, if it is one covered by the paragraph, is in the public interest.” 300 U.S. at 287, 57 S.Ct. at 476.
The rule that a court in a § 1(20) proceeding has the obligation to determine whether the Commission had the jurisdiction to deny the certificate on its merits is based on the view that the Commission cannot be the final arbiter of" its own power. That decision lies with the courts. However, the problem on this appeal is not, as it was in Powell', whether a court shall pass on the Commission’s power to deny a certificate to Tampa, but which court is the proper forum for such a determination. The court in Powell noted that such a determination was proper in “either a suit to set aside an order granting a certificate or in a suit under section 1(20) to enjoin a violation of section 1(18).” However, neither the court in Powell nor the court in Texas & P. R. Co. v Gulf, C. & S. F. R. Co., faced or passed upon the possibility of concurrent jurisdiction in both the three judge court and the § 1(20) court. Statements that “either” a § 1(20) court or a three judge court could inquire anew into jurisdictional issues such as the interstate character of the railroad or the nature of its trackage must be read in light of the fact that jurisdiction in both courts was not contemplated. If the issue was the denial of a certificate of public convenience and necessity, as it was in Piedmont and Texas P. R. Co. v. Gulf, then the negative order rule precluded review in a three judge court. On the other hand, if the plaintiff sought to set aside, an order of the Commission that was not negative in character, the exclusive forum for review was a three judge court. Powell v. United States, 300 U.S. at 288-289, 57 S.Ct. 470; United States v. Idaho, supra. In either case a court would review de novo the jurisdictional basis for the Commission’s actions. Nowhere was it stated that both courts were expected to reexamine matters relating to the Commission’s jurisdiction.
Subsequent to Texas & P. R. Co. v. Gulf, C. & S. F. R. Co., Piedmont & N. R. Co. v. United States, United States v. Idaho, and Powell v. United States, the United States Supreme Court repudiated the negative order rule. Rochester Telephone Corp. v. United States, 1939, 307 U.S. 125, 59 S.Ct. 754, 83 L.Ed. 1147. In so doing the Court described the very type of situation that had prevented review in Piedmont, and disapproved its earlier decisions insofar as they were [398]*398affected by the artificial distinction between negative and affirmative orders:
“An order of the Commission dismissing a complaint on the merits and maintaining the status quo is an exercise of administrative function, no more and no less, than an order directing some change in status. The nature of the issues foreclosed by the Commission’s action and the nature of the issues left open, so far as the reviewing power of courts is concerned, are the same. * * * We conclude, therefore, that any distinction, as such, between ‘negative’ and ‘affirmative’ orders, as a touchstone of jurisdiction to review the Commission’s orders, serves no useful purpose, and insofar as earlier decisions have been controlled by this distinction, they can no longer be guiding.” 307 U.S. at 142-143, 59 S.Ct. at 763-764.
See also 307 U.S. at 133, n. 11, 59 S.Ct. 754, where Piedmont was repudiated.
On the same day as the decision in Rochester, the Supreme Court reaffirmed its rejection of the negative order rule in United States v. Maher, 1939, 307 U.S. 148, 59 S.Ct. 768, 83 L.Ed. 1162, permitting review by a three judge court of the denial by the Interstate Commerce Commission of a certificate of public convenience and necessity. Some ten years later the court once again announced that the negative order doctrine had no further vitality. In United States v. Interstate Commerce Commission, 1948, 337 U.S. 426, 436, 69 S.Ct. 1410, 93 L.Ed. 1451, 1460, the court informs us that the doctrine of- negativity was “wholly abandoned in Rochester Teleph. Corp. v. United States, 307 U.S. 125, 59 S.Ct. 754, 83 L.Ed. 1147.” Under these circumstances no further burial ritual would seem required. Today it is clear, contrary to the teachings of Piedmont, that a carrier may receive review in a three judge district court of the denial by the Interstate Commerce Commission of his petition for a certificate of public convenience and necessity. 28 U.S.C.A. §§ 2321-2325; United States v. Maher, supra; Northwestern Pacific R. Co. v. United States, N.D.Calif., 1964, 228 F.Supp. 690, aff’d per curiam, 379 U.S. 132, 85 S.Ct. 274, 13 L.Ed.2d 333; Chicago, Milwaukee, St. P. & P. R. Co. v. United States, E.D.Wis., 1963, 214 F. Supp. 244, reversed on other grounds, 380 U.S. 448, 85 S.Ct. 1102, 14 L.Ed.2d 151; State of Georgia v. United States, N.D.Ga., 1957, 156 F.Supp. 711, aff’d per curiam, 356 U.S. 273, 78 S.Ct. 771, 2 L.Ed.2d 760.
With the way open for direct review of the validity of the Commission’s order in a statutory three judge court, the question is raised whether such court is the exclusive forum for the adjudication of jurisdictional questions, or whether they may also be considered in a § 1(20) proceeding. We believe that all issues essential to the judgment of the Commission, including matters relating to its jurisdiction, should be challenged only in the statutory three judge court. Consideration of such issues in a § 1(20) proceeding amounts to a collateral attack upon the validity of the Commission’s order.
Our conclusion that the statutory three judge court provided for in 28 U.S.C.A. §§ 2321-2325 is the sole forum for review of issues decided by the Commission finds support in United States v. Southern Railway Company, 5 Cir. 1966, 364 F.2d 86, cert. denied, 386 U.S. 1031, 87 S. Ct. 1479, 18 L.Ed.2d 592.7 In that case we held that suits by the United States for civil penalties resulting from the violation by a carrier of the Commission’s order could not be defended on grounds that were reviewable before a statutory three judge court. In holding that the district court was in error in considering such defenses, our court observed:
“ * * * The requirement of the statute that Administrative and Judicial Review be applicable to all cases of suits to enforce, enjoin, suspend or set aside orders of the Commission, [399]*399has a direct reference to 28 U.S.C. § 2321, in which the Congress directed that the procedure to be employed in such cases required the creation of a three-judge court. Explicit in the language of the legislation is the requirement that judicial relief, othér than for the payment of money or the collection of fines, penalties or forfeitures, must be obtained under §§ 2321-2325 of Title 28. A long line of Supreme Court cases establishes the principle that where Congress has provided administrative and judicial review procedures which are designed to permit agency expertise to be brought to bear upon particular problems, the prescribed procedures are exclusive, and this is true even though Congress may not have expressly provided for the exclusiveness of the statutory procedure.” 364 F.2d at 91.
The rule relied upon in United States v. Southern Railway Company, supra, is an application of the doctrine of exhaustion of administrative remedies, see Whitney National Bank v. Bank of New Orleans, 1965, 379 U.S. 411, 422, 85 S.Ct. 551, 13 L.Ed.2d 386, 395, and as such is not limited only to matters within the agency’s expertise. One of its purposes is to discourage the piecemeal litigation which necessarily would result if parties could begin a suit before an agency, then disengage from combat, and challenge the agency’s ruling in a collateral enforcement proceeding at some later date. See United States v. Ruzicka, 1946, 329 U.S. 287, 67 S.Ct. 207, 91 L.Ed. 290; Callanan Road Improvement Co. v. United States, 1953, 345 U.S. 507, 73 S.Ct. 803, 97 L.Ed. 1206. The rule also avoids unnecessary duplication and conflicting litigation likely to result where two courts have concurrent jurisdiction to resolve the same issues. Whitney National Bank v. Bank of New Orleans, supra. Both considerations have relevance whether the issues sought to be raised in a collateral proceeding are committed to the expertise of the agency, see United States v. Southern Railway, supra, or whether they are jurisdictional and go to the power of the administrative tribunal. See Myers v. Bethlehem Shipbuilding Corp., 1938, 303 U.S. 41, 49-51, 58 S.Ct. 459, 82 L.Ed. 638, 643-644; Callanan Road Improvement Co. v. United States,. supra; Interstate Commerce Commission v. Cowan, D.C.S..D. 1955, 127 F.Supp. 247, 249. In either case the policy of exhausting administrative remedies in the reviewing court specifically designated for such purpose is served. To insure fealty to such policy, and to preserve the statutory mode of review, a court in a collateral proceeding may invoke principles of collateral estoppel. See Meyers v. Famous Realty, Inc., 2 Cir. 1959, 271 F.2d 811, 817, cert, denied, 362 U.S. 910, 80 S.Ct. 681, 4 L.Ed.2d 619; Southwest Freight Lines v. Interstate Commerce Commission, 8 Cir. 1950, 184 F.2d 149, 151; Schwartz v. Bowman, S.D.N.Y. 1965, 244 F.Supp. 51, 67, cert, denied, 385 U.S. 921, 87 S.Ct. 230, 17 L.Ed.2d 145; Interstate Commerce Commission v. Cowan, supra; United States v. Railway Exp. Agency, D.Del., 1951, 101 F.Supp. 1008, 1012; Interstate Commerce Commission v. G. & M. Motor Transfer Co., Inc., W.D.N.C.1945, 64 F.Supp. 302; Interstate Commerce Commission v. Consolidated Freightways, S.D.N.D.1941, 41 F.Supp. 651, 655. Such principles are properly applied to protect the findings of an administrative tribunal, Pacific Seafarers, Inc. v. Pacific Far East Line, Inc., D.C.Cir.1968, 404 F.2d 804, 809, cert, denied, 393 U.S. 1093, 89 S.Ct. 872, 21 L.Ed.2d 784, as well as the findings of a court, and may be invoked to prevent re-litigation of jurisdictional findings, Stoll v. Gottlieb, 1938, 305 U.S. 165, 59 S.Ct. 134, 83 L.Ed. 104; Pacific Seafarers, Inc. v. Pacific Far East Line, Inc., supra; Seatrain Lines v. Pennsylvania R. Co., 3 Cir. 1953, 207 F.2d 255, 257; Interstate Commerce Commission v. Cowan, supra, no less than findings on the merits. Under these circumstances we think the trial court should not have acquiesced, at least on the basis of appellant’s cases, to the request [400]*400that the issue of interstate commerce be tried anew.
For some reason, not apparent on the record, the court refused to reexamine the Commission’s ruling that Tampa’s proposed railroad was not a spur or industrial track within the exemption of § 1(22). Yet this issue is no less jurisdictional than the allegation that the defendant is not an interstate carrier. See Texas & P. R. Co. v. Gulf, C. & S. F. R. Co., 270 U.S. at 274, 46 S.Ct. 263; United States v. Idaho, supra. Both questions should have been foreclosed in this § 1(20) proceeding. Only questions not essential to the Commission’s ruling were open for review. See Restatement of Judgments, §§ 68 (o) and 68 (p); Hyman v. Regenstein, 5 Cir. 1958, 258 F.2d 502, 510, cert. denied, 359 U.S. 913, 79 S.Ct. 589, 3 L.Ed.2d 575.
The result that we have reached is not affected by the fact that a § 1(20) proceeding is not strictly speaking an action to enforce an order of the Commission, but an action to enjoin a violation of § 1(18), § 1(19), or § 1(20) of the Act. The form the issues may take in a subsequent proceeding is not material to the application of collateral estoppel so long as the issues were in fact determined in the original suit. “In such cases, it is * * * immaterial that the two actions are based on different grounds, or tried on different theories, or instituted for different purposes, and seek different relief. * * * ” American Jurisprudence, Judgments, Section 371; Hyman v. Regenstein, supra, at 510. Cf. Meyers v. Famous Realty, Inc., supra-, I.C.C. v. Cowan, supra. Complete identity of the parties or mutuality of estoppel is also not required where the party against whom the estoppel is applied has already had his day in court. Seguros Tepeyac, S.A. Compania Mexicana de Seguros Generales v. Jernigan, 5 Cir. 1969, 410 F.2d 718; United States v. Webber, 3 Cir. 1968, 396 F.2d 381, 385.
IV.
Tampa’s final argument against the application of collateral estoppel involves an allegation of changed circumstances. Tampa makes much of changes that were made in its proposed railroad subsequent to the Commission’s ruling and denial of a certificate. Tampa argues that these changes “remove any possible contention that the Tampa Railroad is seeking to obtain interstate authority by a collateral attack upon an order of the Commission denying it interstate operating rights.”
Whether or not Tampa has so modified its proposed railroad as to take itself out from under the doctrine of collateral estoppel cannot be determined on the record before us. Such a determination must abide factual inquiries by the district court on the extent and nature of Tampa’s modifications. However, a few words on the standard which Tampa must meet are appropriate.
In Amerada Petroleum Corp. v. Federal Power Comm., 8 Cir. 1964, 334 F.2d 404, reversed, 379 U.S. 687, 85 S.Ct. 632, 13 L.Ed.2d 605, the court had before it a jurisdictional issue strikingly similar to the one here presented. The court noted that in an earlier suit involving some of the same parties, North Dakota v. Federal Power Commission, 8 Cir. 1957, 247 F.2d 173, it had sustained a decision by the Federal Power Commission in which the Commission had declined jurisdiction over certain natural gas contracts on the ground that the gas had been produced, sold and transported only in intrastate commerce. Subsequent to the Commission’s ruling and after this first decision by the Eighth Circuit, the gas producers had entered into new contracts with their distributors, and the Commission had asserted jurisdiction over the natural gas covered by the new contracts. On appeal, the Eighth Circuit set aside the order of the Commission, holding that under the doctrine of collateral estoppel the question of interstate commerce was [401]*401foreclosed by the court’s earlier decision. In so deciding, however, the court noted that “Unlike the facts of the former case, this appeal from the present administrative ruling concerns the Commission’s passage upon new contracts providing for future sales by Petitioners, as well as additional producers, of different quantities of gas destined to travel over expanded pipelines in the MDU system.” 334 F.2d at 411. The changes that had occurred subsequent to the Commission’s ruling were therefore quite substantial.
On certiorari, the Supreme Court reversed the Eighth Circuit on its application of collateral estoppel. Federal Power Commission v. Amerada Petroleum Corp., 1965, 379 U.S. 687, 85 S.Ct. 632, 13 L.Ed.2d 605. The rationale for the court’s decision is briefly stated as follows:
“The Court of Appeals thought that its decision in North Dakota v. Federal Power Comm., 8 Cir., 247 F.2d 173, brought collateral estoppel into play in the present case. 334 F.2d 404, 411-412. But that rule has no place here for no judgment governing past events is in jeopardy, only the scope of future regulations that involves different events and transactions.” 379 U.S. at 690, 85 S.Ct. at 634.
The test established by the Supreme Court in Amerada, for the avoidance of collateral estoppel on the grounds of changed circumstances would seem to require that the change between the old and the new facts be so substantial that the new facts constitute in effect “different events and transactions.” In Amerada that test was easily met since the changed circumstances involved new contracts, new producers, and several hundred miles of new pipeline. It must also be noted that Amerada involved the attempted use of collateral estoppel to foreclose governmental regulation in the public interest. In such circumstances there is a natural reluctance to apply collateral estoppel in all its vigor lest a prior adjudication, involving only intrastate activities, become a method and a license for unbridled interstate expansion.
The instant case involves policies almost the reverse of those present in Amerada. Here the Commission has asserted its jurisdiction, and the carrier involved, disappointed with the results of its encounter, has ignored not only the Commission’s ruling, but the statutorily prescribed method of review. Under these circumstances collateral estoppel must be rigorously applied lest all carriers frustrate the rulings of the Commission merely by alleging some slight modification in their proposed plans. No plan is so constituted that it will not admit of modification if the aim of such sacrifice is a chance to relitigate issues already considered and resolved by the Commission.
In the case before us we cannot subscribe to a “one inch” rule whereby the removal of a few railroad ties becomes a significant event. Nor can we admit that all changes are significant. At the very least, modifications in Tampa’s proposed railroad must bear directly upon the issue sought to be foreclosed, here interstate commerce. Absent such a relationship, the change is of no avail in suspending the effects of collateral estoppel. Finally the proposed modifications must constitute such substantial changes that the court can say in good conscience that the ruling of the Commission is not “in jeopardy,” Federal Power Commission v. Amerada Petroleum Corp., swpra, and that the case before it involves in effect “different events and transactions.” If these things may be said, then the court is justified in re-examining issues that are similar to those presented to the agency. Events and circumstances do change and therefore the opportunity for a second look is sometimes justified. But the need to avoid giving a judicial blessing to the circumvention of the Commission’s order, and the need to prevent the disregard of a prescribed mode of review, are equally valid considera[402]*402tions. Tampa must not be permitted to subvert the role which Congress has assigned to the Commission, and by means of a unilateral decision to modify its proposed railroad, to render the Commission’s order a nullity. If, in fact, the district court is faced with new matter and new circumstances which justify treating Tampa’s case as if there never had been an order of the Commission, then the burden of such a showing must be heavily on Tampa.
We note in conclusion that Tampa’s afterthoughts and postscripts on the modifications to its proposed railroad must now be examined in light of the proposals previously presented to the Commission. In the course of such examination there must be no sanction to the proposition that every minute deviation from a carrier’s application to the Commission furnishes a detour or escape route into the state courts. Tampa’s gambit should be analyzed with the principle in mind that forum shopping and litigation fragmentation are anathema to sound judicial administration. Accordingly, on remand the trial court must determine whether Tampa has modified its original proposal to the extent necessary to avoid the application of collateral estoppel.
The judgment of the district court granting the preliminary injunction is affirmed and the cause is remanded for further proceedings not inconsistent with this opinion.