GURFEIN, Circuit Judge:

This appeal involving a regulation of the Food and Drug Administration is not here upon a direct review of agency action. It is an appeal from a judgment of the District Court for the Eastern District of New York (Hon. John J. Dooling, Judge) enjoining the appellants, after a hearing, from processing hot smoked whitefish except in accordance with time-temperature-salinity (T-T-S) regulations contained in 21 C.F.R. Part 122 (1977). The thorough analytical opinion of the District Court is reported at 417 F.Supp. 1364 (Aug. 17, 1976).

The injunction was sought and granted on the ground that smoked whitefish which has been processed in violation of the T-T-S regulation is "adulterated." Food, Drug and Cosmetics Act ("the Act") §§ 302(a) and 301(k), 21 U.S.C. §§ 332(a), 331(k).

Appellant Nova Scotia receives frozen or iced whitefish in interstate commerce which it processes by brining, smoking and cooking. The fish are then sold as smoked whitefish.

The regulations cited above require that hot-process smoked fish be heated by a controlled heat process that provides a monitoring system positioned in as many strategic locations in the oven as necessary to assure a continuous temperature through each fish of not less than 180 ° F. for a minimum of 30 minutes for fish which have been brined to contain 3.5% water phase salt or at 150 ° F. for a minimum of 30 minutes if the salinity was at 5% water phase. Since each fish must meet these requirements, it is necessary to heat an entire batch of fish to even higher temperatures so that the lowest temperature for any fish will meet the minimum requirements.

Government inspection of appellants' plant established without question that the minimum T-T-S requirements were not being met. There is no substantial claim that the plant was processing whitefish under "insanitary conditions" in any other material respect. Appellants, on their part, do not defend on the ground that they were in compliance, but rather that the requirements could not be met if a marketable whitefish was to be produced. They defend upon the grounds that the regulation is invalid (1) because it is beyond the authority delegated by the statute; (2) because the FDA improperly relied upon undisclosed evidence in promulgating the regulation and because it is not supported by the administrative record; and (3) because there was no adequate statement setting forth the basis of the regulation. We reject the contention that the regulation is beyond the authority delegated by the statute, but we find serious inadequacies in the procedure followed in the promulgation of the regulation and hold it to be invalid as applied to the appellants herein.

The hazard which the FDA sought to minimize was the outgrowth and toxin formation of Clostridium botulinum Type E spores of the bacteria which sometimes inhabit fish. There had been an occurrence of several cases of botulism traced to consumption of fish from inland waters in 1960 and 1963 which stimulated considerable bacteriological research. These bacteria can be present in the soil and water of various regions. They can invade fish in their natural habitat and can be further disseminated in the course of evisceration and preparation of the fish for cooking. A failure to destroy such spores through an adequate brining, thermal, and refrigeration process was found to be dangerous to public health.

The Commissioner of Food and Drugs ("Commissioner"), employing informal "notice-and-comment" procedures under 21 U.S.C. § 371(a), issued a proposal for the control of C. botulinum bacteria Type E in fish. 34 F.R. 17,176 (Oct. 23, 1969). For his statutory authority to promulgate the regulations, the Commissioner specifically relied only upon § 342(a)(4) of the Act which provides:

"A food shall be deemed to be adulterated —

"(4) if it has been prepared, packed, or held under insanitary conditions whereby it may have become contaminated with filth, or whereby it may have been rendered injurious to health;"

Similar guidelines for smoking fish had been suggested by the FDA several years earlier, and were generally made known to people in the industry. At that stage, however, they were merely guidelines without substantive effect as law. Responding to the Commissioner's invitation in the notice of proposed rulemaking, members of the industry, including appellants and the intervenor-appellant, submitted comments on the proposed regulation.

The Commissioner thereafter issued the final regulations in which he adopted certain suggestions made in the comments, including a suggestion by the National Fisheries Institute, Inc. ("the Institute"), the intervenor herein. 35 F.R. 17,401 (Nov. 13, 1970). The original proposal provided that the fish would have to be cooked to a temperature of 180 ° F. for at least 30 minutes, if the fish have been brined to contain 3.5% water phase salt, with no alternative. In the final regulation, an alternative suggested by the intervenor "that the parameter of 150 ° F. for 30 minutes and 5% salt in the water phase be established as an alternate procedure to that stated in the proposed regulation for an interim period until specific parameters can be established" was accepted, but as a permanent part of the regulation rather than for an interim period.

The intervenor suggested that "specific parameters" be established. This referred to particular processing parameters for different species of fish on a "species by species" basis. Such "species by species" determination was proposed not only by the intervenor but also by the Bureau of Commercial Fisheries of the Department of the Interior. That Bureau objected to the general application of the T-T-S requirement proposed by the FDA on the ground that application of the regulation to all species of fish being smoked was not commercially feasible, and that the regulation should therefore specify time-temperature-salinity requirements, as developed by research and study, on a species-by-species basis. The Bureau suggested that "wholesomeness considerations could be more practically and adequately realized by reducing processing temperature and using suitable concentrations of nitrite and salt." The Commissioner took cognizance of the suggestion, but decided, nevertheless, to impose the T-T-S requirement on all species of fish (except chub, which were regulated by 21 C.F.R. 172.177 (1977) [dealing with food additives]).

He did acknowledge, however, in his "basis and purpose" statement required by the Administrative Procedure Act ("APA"), 5 U.S.C. § 553(c), that "adequate times, temperatures and salt concentrations have not been demonstrated for each individual species of fish presently smoked". 35 F.R. 17,401 (Nov. 13, 1970). The Commissioner concluded, nevertheless, that "the processing requirements of the proposed regulations are the safest now known to prevent the outgrowth and toxin formation of C. botulinum Type E". He determined that "the conditions of current good manufacturing practice for this industry should be established without further delay." Id.

The Commissioner did not answer the suggestion by the Bureau of Fisheries that nitrite and salt as additives could safely lower the high temperature otherwise required, a solution which the FDA had accepted in the case of chub. Nor did the Commissioner respond to the claim of Nova Scotia through its trade association, the Association of Smoked Fish Processors, Inc., Technical Center that "[t]he proposed process requirements suggested by the FDA for hot processed smoked fish are neither commercially feasible nor based on sound scientific evidence obtained with the variety of smoked fish products to be included under this regulation." (Exhibit D, Tab A).

Nova Scotia, in its own comment, wrote to the Commissioner that "the heating of certain types of fish to high temperatures will completely destroy the product". It suggested, as an alternative, that "specific processing procedures could be established for each species after adequate work and experimention [sic] has been done — but not before." ( Id.). We have noted above that the response given by the Commissioner was in general terms. He did not specifically aver that the T-T-S requirements as applied to whitefish were, in fact, commercially feasible.

When, after several inspections and warnings, Nova Scotia failed to comply with the regulation, an action by the United States Attorney for injunctive relief was filed on April 7, 1976, six years later, and resulted in the judgment here on appeal. The District Court denied a stay pending appeal, and no application for a stay was made to this court.

I

The argument that the regulation is not supported by statutory authority cannot be dismissed out of hand. See Schilling v. Rogers, 363 U.S. 666, 676-77, 80 S.Ct. 1288, 1295-96, 4 L.Ed.2d 1478 (1960); see Batterton v. Francis, 432 U.S. 416, 97 S.Ct. 2399, 53 L.Ed.2d 448 (1977). The sole statutory authority relied upon is § 342(a)(4), quoted above. As we were instructed in S. E. C. v. Chenery Corp., 318 U.S. 80, 87, 63 S.Ct. 454, 459, 87 L.Ed. 626 (1943) ( Chenery I), "[t]he grounds upon which an administrative order must be judged are those upon which the record discloses that its action was based." Nor is the Commissioner's expressed reliance solely on § 342(a)(4) a technicality which might be removed by a later and wiser reliance on another subsection. For in this case, as the agency recognized, there is no other section or subsection that can pass as statutory authority for the regulation. The categories of "adulteration" prohibited in Section 342 all refer to food as an "adulterated" product rather than to the process of preparing food, except for subsection (a)(4) which alone deals with the processing of food.

Appellants contend that the prohibition against "insanitary conditions" embraces conditions only in the plant itself, but does not include conditions which merely inhibit the growth of organisms already in the food when it enters the plant in its raw state. They distinguish between conditions which are insanitary, which they concede to be within the ambit of § 342(a)(4), and conditions of sterilization required to destroy micro-organisms, which they contend are not.

It is true that on a first reading the language of the subsection appears to cover only "insanitary conditions" " whereby it [the food] may have been rendered injurious to health" (emphasis added). And a plausible argument can, indeed, be made that the references are to insanitary conditions in the plant itself, such as the presence of rodents or insects, e. g., United States v. Park, 421 U.S. 658, 95 S.Ct. 1903, 44 L.Ed.2d 489 (1975); United States v. Cassaro, Inc., 443 F.2d 153 (1st Cir. 1971); United States v. Hammond Milling Co., 413 F.2d 608 (5th Cir. 1969), cert. denied, 396 U.S. 1002, 90 S.Ct. 552, 24 L.Ed.2d 494 (1970).

Yet, when we are dealing with the public health, the language of the Food, Drug and Cosmetic Act should not be read too restrictively, but rather as "consistent with the Act's overriding purpose to protect the public health". United States v. Bacto-Unidisk, 394 U.S. 784, 798, 89 S.Ct. 1410, 1418, 22 L.Ed.2d 726 (1969). As Justice Frankfurter said in United States v. Dotterweich, 320 U.S. 277, 280, 64 S.Ct. 134, 136, 88 L.Ed. 48 (1943):

"The purposes of this legislation thus touch phases of the lives and health of people which, in the circumstances of modern industrialism, are largely beyond self-protection. Regard for these purposes should infuse construction of the legislation if it is to be treated as a working instrument of government and not merely as a collection of English words."

Thus, a provision concerning "food additives" has been held to include even poisonous substances which have not been "added" by human hands. United States v. Ewig Bros. Co., 502 F.2d 715, 721-24 (7th Cir. 1974) (Stevens, J. now Mr. Justice Stevens), cert. denied, 420 U.S. 945, 95 S.Ct. 1325, 43 L.Ed.2d 423 (1975).

Section 371(a), applicable to rulemaking under § 342(a)(4), provides: "The authority to promulgate regulations for the efficient enforcement of this chapter, except as otherwise provided in this section, is vested in the Secretary." We read this grant as analogous to the provision "make . . . such rules and regulations as may be necessary to carry out the provisions of this Act," in which case "the validity of a regulation promulgated thereunder will be sustained so long as it is `reasonably related to the purposes of the enabling legislation.' [citations omitted]" Mourning v. Family Publications Service, Inc., 411 U.S. 356, 369, 93 S.Ct. 1652, 36 L.Ed.2d 318 (1973). When agency rulemaking serves the purposes of the statute, courts should refuse to adopt a narrow construction of the enabling legislation which would undercut the agency's authority to promulgate such rules. United States v. Midwest Video Corp., 406 U.S. 649, 92 S.Ct. 1860, 32 L.Ed.2d 390 (1972) (upholding FCC's authority to promulgate "cablecasting" regulation as "reasonably ancillary to the effective performance of the Commission's various responsibilities for the regulation of television broadcasting"); and United States v. Storer Broadcasting Co., 351 U.S. 192, 203-04, 76 S.Ct. 763, 100 L.Ed.2d 1081 (1956). The court's role should be one of constructive cooperation with the agency in furtherance of the public interest. International Harvester Co. v. Ruckelhaus, 155 U.S.App.D.C. 411, 443, 478 F.2d 615, 647 (1973). As the Supreme Court has said, "We are, in the absence of compelling evidence that such was Congress' intention, unwilling to prohibit administrative action imperative for the achievement of an agency's ultimate purposes." In re Permian Basin Area Rate Cases, 390 U.S. 747, 780, 88 S.Ct. 1344, 20 L.Ed.2d 312 (1968).

Appellant's argument, it should be noted, is not that there has been an unlawful delegation of legislative power, Schechter Poultry Corp. v. United States, 295 U.S. 495, 55 S.Ct. 837, 79 L.Ed. 1570 (1935), or even a delegation of "unfettered discretion." See Amalgamated Meat Cutters Butcher Workers v. Connally, 337 F.Supp. 737, 757 (D.D.C. 1971) (Leventhal, J.). The argument, fairly construed, is that Congress did not mean to go so far as to require sterilization sufficient to kill bacteria that may be in the food itself rather than bacteria which accreted in the factory through the use of insanitary equipment.

There are arguments which can indeed be mustered to support such a broad-based attack under 5 U.S.C. § 706.

First, the Act deals with standards of identity and various categories that can render food harmful to health. Yet, so far as the category of harmful micro-organisms is concerned, there is only a single provision, 21 U.S.C. § 344, which directly deals with "micro-organisms." That provision is limited to emergency permit controls dealing with any class of food which the Secretary finds, after investigation, "may, by reason of contamination with micro-organisms during the manufacture, processing or packing thereof in any locality, be injurious to health, and that such injurious nature cannot be adequately determined after such articles have entered interstate commerce, [in which event] he then, and in such case only, shall promulgate regulations providing for the issuance . . . of permits . . .." (Emphasis added.) It may be argued that the failure to mention "micro-organisms" in the "adulteration" section of the Act, which includes § 342(a)(4), means that Congress intended to delegate no further authority to control micro-organisms than is expressed in the "emergency" control of Section 344.

On the other hand, as Judge Dooling held, the manner of processing can surely give rise to the survival, with attendant toxic effects on humans, of spores which would not have survived under stricter "sanitary" conditions. In that sense, treating "insanitary conditions" in relation to the hazard, the interpretation of the District Court which described the word "sanitary" as merely "inelegant" is a fair reading, emphasizing that the food does not have to be actually contaminated during processing and packing but simply that "it may have been rendered injurious to health," § 342(a)(4), by inadequate sanitary conditions of prevention.

The other argument of some force would involve the difference in agency rulemaking procedure which results from treating the regulations in question as supported by § 342(a)(4). The Act was enacted in substantially its present form in 1938, preceding the APA by about eight years. In 1938 Congress, groping for standards for rulemaking and the scope of judicial review (as it still is) created a bifurcated structure for rulemaking and, hence, for judicial review.

Congress decided to allow informal rulemaking to the FDA generally, § 371(a), but it also provided for formal rulemaking which, upon request, required "a public hearing for the purpose of receiving evidence relevant and material to the issues raised by such objections." § 371(e). It further provided that in such cases the "order shall be based only on substantial evidence of record at such hearing and shall set forth, as part of the order, detailed findings of fact on which the order is based."

The formal rulemaking with its concomitant standard of "substantial evidence of record," see United States v. Florida East Coast Railway Co., 410 U.S. 224, 93 S.Ct. 810, 35 L.Ed.2d 223 (1973), is limited, however, to rulemaking under specific enumerated sections of the Act. Section 342, the "adulteration" section upon which the Commissioner relied for his statutory authority is not one of these. The section dealing with temporary permits for micro-organisms, § 344, is. See § 371(e)(1). Thus, a temporary suspension because of the presence of micro-organisms in food merits a formal procedure while permanent regulation of micro-organisms is achievable by informal "notice-and-comment" procedure. Even though we read the statute § 342(a)(4) broadly in terms of the authority delegated to the agency, we must, nevertheless, view with some strictness the minimal requirements for the informal "notice and comment" procedure that follows as of course — a matter we shall discuss below.

We do not discount the logical arguments in support of a restrictive reading of § 342(a)(4), but we perceive a larger general purpose on the part of Congress in protecting the public health.

We come to this conclusion, aside from the general rules of construction noted above, for several reasons: First, until this enforcement proceeding was begun, no lawyer at the knowledgeable Food and Drug bar ever raised the question of lack of statutory delegation or even hinted at such a question. Second, the body of data gathered by the experts, including those of the Technical Laboratory of the Bureau of Fisheries manifested a concern about the hazards of botulism. Third, analogously, the Meat Inspection Act of 1907 (now codified as amended at 21 U.S.C. § 608), which hardly provided a clearer standard than does the "insanitary conditions" provision in the Food and Drug Act, has regulations under it concerning mandatory temperatures for processing pork muscle tissue to eliminate the hazard of trichonosis. The statute permits the Secretary "to prescribe the rules and regulations of sanitation under which such establishments shall be maintained". The current regulation, 9 C.F.R. § 318.10 (1977), provides: "All parts of the pork muscle tissue shall be heated to a temperature not lower than 137 ° F., and the method used shall be one known to insure such a result." 9 C.F.R. § 318.10(c)(1) (1977). The same regulation was codified as early as 1949 as 9 C.F.R. § 18.10(c)(1) (1949). These regulations have been assumed for years to have been properly promulgated by the Secretary of Agriculture under the statutory authority given to him.

Lastly, a holding that the regulation of smoked fish against the hazards of botulism is invalid for lack of authority would probably invalidate, to the extent that our ruling would be followed, the regulations concerning the purity of raw materials before their entry into the manufacturing process in 21 C.F.R. Part 113 (1977) (inspection of incoming raw materials for microbiological contamination before thermal processing of low-acid foods packed in hermetically sealed containers), in 21 C.F.R. Part 118 (1977) (pasteurization of milk and egg products to destroy Salmonella micro-organisms before use of the products in cacao products and confectionery), and 21 C.F.R. Part 129 (1977) (product water supply for processing and bottling of bottled drinking water must be of a safe, sanitary quality when it enters the process).

The public interest will not permit invalidation simply on the basis of a lack of delegated statutory authority in this case. A gap in public health protection should not be created in the absence of a compelling reading based upon the utter absence of any statutory authority, even read expansively. Here we find no congressional history on the specific issue involved, and hence no impediment to the broader reading based on general purpose. We believe, nevertheless, that it would be in the public interest for Congress to consider in the light of existing knowledge, a legislative scheme for administrative regulation of the processing of food where hazard from micro-organisms in food in its natural state may require affirmative procedures of sterilization. This would entail, as well, a decision on the type of rulemaking procedure Congress thinks fit to impose.

II

Appellants contend that there is an inadequate administrative record upon which to predicate judicial review, and that the failure to disclose to interested persons the factual material upon which the agency was relying vitiates the element of fairness which is essential to any kind of administrative action. Moreover, they argue that the "concise general statement of . . . basis and purpose" by the Commissioner was inadequate. 5 U.S.C. § 553.

The question of what is an adequate "record" in informal rulemaking has engaged the attention of commentators for several years. The extent of the administrative record required for judicial review of informal rulemaking is largely a function of the scope of judicial review. Even when the standard of review is whether the promulgation of the rule was "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law," as specified in 5 U.S.C. § 706(2)(A), judicial review must nevertheless, be based on the "whole record" ( id.). Adequate review of a determination requires an adequate record, if the review is to be meaningful. Davis, Administrative Law in the Seventies, supra, at 669-71. What will constitute an adequate record for meaningful review may vary with the nature of the administrative action to be reviewed. Friendly, "Some Kind of Hearing," 123 U.Pa.L.Rev. 1267, 1291-92 (1975). Review must be based on the whole record even when the judgment is one of policy, except that findings of fact such as would be required in an adjudicatory proceeding or in a formal "on the record" hearing for rulemaking need not be made. Overton Park, supra, 401 U.S. at 416-18, 91 S.Ct. at 823-25 (1971). Though the action was informal, without an evidentiary record, the review must be "thorough, probing, [and] in depth". Id., 401 U.S. at 415, 91 S.Ct. 814. See Scalia Goodman, Procedural Aspects of the Consumer Product Safety Act, 20 U.C.L.A.L.Rev. 899, 934-35 (1973).

This raises several questions regarding the informal rulemaking procedure followed here: (1) What record does a reviewing court look to? (2) How much of what the agency relied on should have been disclosed to interested persons? (3) To what extent must the agency respond to criticism that is material?

A

With respect to the content of the administrative "record," the Supreme Court has told us that in informal rulemaking, "the focal point for judicial review should be the administrative record already in existence, not some new record made initially in the reviewing court." See Camp v. Pitts, 411 U.S. 138, 142, 93 S.Ct. 1241, 1244, 36 L.Ed.2d 106 (1973).

No contemporaneous record was made or certified. When, during the enforcement action, the basis for the regulation was sought through pretrial discovery, the record was created by searching the files of the FDA and the memories of those who participated in the process of rulemaking. This resulted in what became Exhibit D at the trial of the injunction action. Exhibit D consists of (1) Tab A containing the comments received from outside parties during the administrative "notice-and-comment" proceeding and (2) Tabs B through L consisting of scientific data and the like upon which the Commissioner now says he relied but which were not made known to the interested parties.

Appellants object to the exclusion of evidence in the District Court "aimed directly at showing that the scientific evidence relied upon by the FDA was inaccurate and not based upon a realistic appraisal of the true facts. Appellants attempted to introduce scientific evidence to demonstrate that in fixing the processing parameters FDA relied upon tests in which ground fish were injected with many millions of botulism [sic] spores and then tested for outgrowth at various processing levels whereas the spore levels in nature are far less and outgrowth would have been prevented by far less stringent processing parameters." (Br. p. 33). The District Court properly excluded the evidence.

In an enforcement action, we must rely exclusively on the record made before the agency to determine the validity of the regulation. The exception to the exclusivity of that record is that "there may be independent judicial fact-finding when issues that were not before the agency are raised in a proceeding to enforce non-adjudicatory agency action." Overton Park, supra, 401 U.S. at 415, 91 S.Ct. at 823 (1971). (Emphasis added.)

Though this is an enforcement proceeding and the question is close, we think that the "issues" were fairly before the agency and hence that de novo evidence was properly excluded by Judge Dooling. Camp v. Pitts, supra. Our concern is, rather, with the manner in which the agency treated the issues tendered.

B

The key issues were (1) whether, in the light of the rather scant history of botulism in whitefish, that species should have been considered separately rather than included in a general regulation which failed to distinguish species from species; (2) whether the application of the proposed T-T-S requirements to smoked whitefish made the whitefish commercially unsaleable; and (3) whether the agency recognized that prospect, but nevertheless decided that the public health needs should prevail even if that meant commercial death for the whitefish industry. The procedural issues were whether, in the light of these key questions, the agency procedure was inadequate because (i) it failed to disclose to interested parties the scientific data and the methodology upon which it relied; and (ii) because it failed utterly to address itself to the pertinent question of commercial feasibility.

1. [41] The History of Botulism in Whitefish

The history of botulism occurrence in whitefish, as established in the trial record, which we must assume was available to the FDA in 1970, is as follows. Between 1899 and 1964 there were only eight cases of botulism reported as attributable to hot-smoked whitefish. In all eight instances, vacuum-packed whitefish was involved. All of the eight cases occurred in 1960 and 1963. The industry has abandoned vacuumpacking, and there has not been a single case of botulism associated with commercially prepared whitefish since 1963, though 2,750,000 pounds of whitefish are processed annually. Thus, in the seven-year period from 1964 through 1970, 17.25 million pounds of whitefish have been commercially processed in the United States without a single reported case of botulism. The evidence also disclosed that defendant Nova Scotia has been in business some 56 years, and that there has never been a case of botulism illness from the whitefish processed by it.

2. [43] The Scientific Data

Interested parties were not informed of the scientific data, or at least of a selection of such data deemed important by the agency, so that comments could be addressed to the data. Appellants argue that unless the scientific data relied upon by the agency are spread upon the public records, criticism of the methodology used or the meaning to be inferred from the data is rendered impossible.

We agree with appellants in this case, for although we recognize that an agency may resort to its own expertise outside the record in an informal rulemaking procedure, we do not believe that when the pertinent research material is readily available and the agency has no special expertise on the precise parameters involved, there is any reason to conceal the scientific data relied upon from the interested parties. As Judge Leventhal said in Portland Cement Ass'n v. Ruckelhaus, 158 U.S.App.D.C. 308, 326, 486 F.2d 375, 393 (1973): "It is not consonant with the purpose of a rulemaking proceeding to promulgate rules on the basis of inadequate data, or on data that [in] critical degree, is known only to the agency." (Emphasis added.) This is not a case where the agency methodology was based on material supplied by the interested parties themselves. Cf. International Harvester Co. v. Ruckelhaus, 155 U.S.App.D.C. 411, 428, 478 F.2d 615, 632 (1973). Here all the scientific research was collected by the agency, and none of it was disclosed to interested parties as the material upon which the proposed rule would be fashioned. Nor was an articulate effort made to connect the scientific requirements to available technology that would make commercial survival possible, though the burden of proof was on the agency. This required it to "bear a burden of adducing a reasoned presentation supporting the reliability of its methodology." International Harvester, supra, 155 U.S.App.D.C. at 439, 478 F.2d at 643.

Though a reviewing court will not match submission against counter-submission to decide whether the agency was correct in its conclusion on scientific matters (unless that conclusion is arbitrary), it will consider whether the agency has taken account of all "relevant factors and whether there has been a clear error of judgment." Overton Park, supra, 401 U.S. at 415-16, 91 S.Ct. at 823-24; Appalachian Power Co. v. Environmental Protection Agency, 477 F.2d 495, 507 (4th Cir. 1973). In this circuit we have said that "it is `arbitrary or capricious' for an agency not to take into account all relevant factors in making its determination." Hanly v. Mitchell, 460 F.2d 640, 648 (2d Cir.), cert. denied, 409 U.S. 990, 93 S.Ct. 313, 34 L.Ed.2d 256 (1972) (an enforcement action under NEPA).

If the failure to notify interested persons of the scientific research upon which the agency was relying actually prevented the presentation of relevant comment, the agency may be held not to have considered all "the relevant factors." We can think of no sound reasons for secrecy or reluctance to expose to public view (with an exception for trade secrets or national security) the ingredients of the deliberative process. Cf. Mobil Oil Corp. v. FPC, 157 U.S.App.D.C. 235, 256-58, 483 F.2d 1238, 1259-61 (1973). Indeed, the FDA's own regulations now specifically require that every notice of proposed rulemaking contain "references to all data and information on which the Commissioner relies for the proposal (copies or a full list of which shall be a part of the administrative file on the matter . . .)." 21 C.F.R. § 10.40(b)(1) (1977). And this is, undoubtedly, the trend. See, e.g., National Nutritional Foods v. Weinberger, 512 F.2d 688 (2d Cir.), cert. denied, 423 U.S. 827, 96 S.Ct. 44, 46 L.Ed.2d 44 (1975).

We think that the scientific data should have been disclosed to focus on the proper interpretation of "insanitary conditions." When the basis for a proposed rule is a scientific decision, the scientific material which is believed to support the rule should be exposed to the view of interested parties for their comment. One cannot ask for comment on a scientific paper without allowing the participants to read the paper. Scientific research is sometimes rejected for diverse inadequacies of methodology; and statistical results are sometimes rebutted because of a lack of adequate gathering technique or of supportable extrapolation. Such is the stuff of scientific debate. To suppress meaningful comment by failure to disclose the basic data relied upon is akin to rejecting comment altogether. For unless there is common ground, the comments are unlikely to be of a quality that might impress a careful agency. The inadequacy of comment in turn leads in the direction of arbitrary decision-making. We do not speak of findings of fact, for such are not technically required in the informal rulemaking procedures. We speak rather of what the agency should make known so as to elicit comments that probe the fundamentals. Informal rulemaking does not lend itself to a rigid pattern. Especially, in the circumstance of our broad reading of statutory authority in support of the agency, we conclude that the failure to disclose to interested persons the scientific data upon which the FDA relied was procedurally erroneous. Moreover, the burden was upon the agency to articulate rationally why the rule should apply to a large and diverse class, with the same T-T-S parameters made applicable to all species. Cf. Associated Industries of N.Y.S., Inc. v. U.S. Dept. of Labor, 487 F.2d 342, 352-53 (2d Cir. 1973). And cf. Industrial Union Dept. AFL-CIO v. Hodgson, 162 U.S. App.D.C. 331, 499 F.2d 467 (1974).

C

Appellants additionally attack the "concise general statement" required by APA, 5 U.S.C. § 553, as inadequate. We think that, in the circumstances, it was less than adequate. It is not in keeping with the rational process to leave vital questions, raised by comments which are of cogent materiality, completely unanswered. The agencies certainly have a good deal of discretion in expressing the basis of a rule, but the agencies do not have quite the prerogative of obscurantism reserved to legislatures. "Congress did not purport to transfer its legislative power to the unbounded discretion of the regulatory body." F.C.C. v. RCA Communications, Inc., 346 U.S. 86, 90, 73 S.Ct. 998, 1002, 97 L.Ed. 1470 (1953) (Frankfurter, J.). As was said in Environmental Defense Fund, Inc. v. EPA, 150 U.S.App.D.C. 348, 371, 465 F.2d 528, 540-51 (1972): "We cannot discharge our role adequately unless we hold EPA to a high standard of articulation. Kennecott Copper Corp. v. EPA, . . . 149 U.S.App.D.C. 231, 462 F.2d 846 (1972)."

The test of adequacy of the "concise general statement" was expressed by Judge McGowan in the following terms:

"We do not expect the agency to discuss every item of fact or opinion included in the submissions made to it in informal rulemaking. We do expect that, if the judicial review which Congress has thought it important to provide is to be meaningful, the `concise general statement of . . . basis and purpose' mandated by Section 4 will enable us to see what major issues of policy were ventilated by the informal proceedings and why the agency reacted to them as it did." Automotive Parts Accessories Ass'n v. Boyd, 132 U.S.App.D.C. 200, 208, 407 F.2d 330, 338 (1968).

And Judge Friendly has noted that "[i]n a case where a proposed standard under OSHA [Occupational Safety and Health Act] has been opposed on grounds as substantial as those presented here, the Department has the burden of offering some reasoned explanation." Associated Industries of New York State, Inc. v. U.S. Department of Labor, supra, 487 F.2d at 352 (emphasis in original).

The Secretary was squarely faced with the question whether it was necessary to formulate a rule with specific parameters that applied to all species of fish, and particularly whether lower temperatures with the addition of nitrite and salt would not be sufficient. Though this alternative was suggested by an agency of the federal government, its suggestion, though acknowledged, was never answered.

Moreover, the comment that to apply the proposed T-T-S requirements to whitefish would destroy the commercial product was neither discussed nor answered. We think that to sanction silence in the face of such vital questions would be to make the statutory requirement of a "concise general statement" less than an adequate safeguard against arbitrary decision-making.

We cannot improve on the statement of the District of Columbia Circuit in Industrial Union Dep't, AFL-CIO v. Hodgson, 162 U.S.App.D.C. 331, 339, 499 F.2d 467, 475 (1974).

"What we are entitled to at all events is a careful identification by the Secretary, when his proposed standards are challenged, of the reasons why he chooses to follow one course rather than another. Where that choice purports to be based on the existence of certain determinable facts, the Secretary must, in form as well as in substance, find those facts from evidence in the record. By the same token, when the Secretary is obliged to make policy judgments where no factual certainties exist or where facts alone do not provide the answer, he should so state and go on to identify the considerations he found to be persuasive."

One may recognize that even commercial infeasibility cannot stand in the way of an overwhelming public interest. Yet the administrative process should disclose, at least, whether the proposed regulation is considered to be commercially feasible, or whether other considerations prevail even if commercial infeasibility is acknowledged. This kind of forthright disclosure and basic statement was lacking in the formulation of the T-T-S standard made applicable to whitefish. It is easy enough for an administrator to ban everything. In the regulation of food processing, the worldwide need for food also must be taken into account in formulating measures taken for the protection of health. In the light of the history of smoked whitefish to which we have referred, we find no articulate balancing here sufficient to make the procedure followed less than arbitrary.

After seven years of relative inaction, the FDA has apparently not reviewed the T-T-S regulations in the light of present scientific knowledge and experience. In the absence of a new statutory directive by Congress regarding control of micro-organisms, which we hope will be worthy of its consideration, we think that the T-T-S standards should be reviewed again by the FDA.

We cannot, on this appeal, remand to the agency to allow further comments by interested parties, addressed to the scientific data now disclosed at the trial below. We hold in this enforcement proceeding, therefore, that the regulation, as it affects nonvacuum-packed hot-smoked whitefish, was promulgated in an arbitrary manner and is invalid.

When the District Court held the regulation to be valid, it properly exercised its discretion to grant the injunction. In view of our conclusion to the contrary, we must reverse the grant of the injunction and direct that the complaint be dismissed.