Reserve Mining Company is a jointly owned subsidiary of Armco Steel Corporation and Republic Steel Corporation which mines low-grade iron ore, called 'taconite,' near Babbitt, Minnesota. The taconite is shipped by rail to Reserve's 'beneficiating' plant at Silver Bay, Minnesota, on the north shore of Lake Superior, where it is concentrated into 'pellets' containing some 65 percent iron ore. The process involves crushing the taconite into fine granules, separating out the metallic iron with huge magnets, and flushing the residue into Lake Superior. Approximately 67,000 tons of this waste product, known as 'tailings,' are daily discharged into the lake.

The use of Lake Superior for this purpose was originally authorized by the State of Minnesota in 1947, and Reserve commenced operations in 1955. In granting this permit to Reserve, the State of Minnesota accepted Reserve's theory that the weight and velocity of the discharge would insure that the tailings would be deposited at a depth of approximately 900 feet in the 'great trough' area of the lake, located offshore from Reserve's facility. The permit provides that:

Tailings shall not be discharged * * * so as to result in any material adverse effects on fish life or public water supplies or in any other material unlawful pollution of the waters of the lake * * *.

This enforcement litigation was commenced after state and federal pollution control efforts dating from mid-1969 produced an unsuccessful series of administrative conferences and state court proceedings.1 On February 2, 1972, the United States Government-- joined eventually by the States of Minnesota, Wisconsin, and Michigan and by various environmental groups-- filed a complaint alleging that Reserve's discharge of tailings into Lake Superior violated Section 13 of the 1899 Refuse Act (33 U.S.C. 407), Section 10 of the Federal Water Pollution Control Act (33 U.S.C. 1160), and the federal common law of public nuisance.

Until June 8, 1973, the case was essentially a water pollution abatement case, but on that date the focus of the controversy shifted to the public health impact of Reserve's discharge of asbestiform particles into the air and water. Hearings on a motion for preliminary injunction were consolidated with the trial on the merits, and on April 20, 1974, after 139 days of trial extending over a nine month period and after hearing more than 100 witnesses and examining over 1,600 exhibits, Judge Miles Lord of the United States District Court for the District of Minnesota entered an order closing Reserve's Silver Bay facility. In an abbreviated memorandum opinion, Judge Lord held that Reserve's water discharge violated federal water pollution laws and that its air emissions violated state air pollution regulations, and that both were common law nuisances. Most importantly to the question now before this court, Judge Lord concluded in Findings 9 and 10 of his opinion that:

9) The discharge into the air substantially endangers the health of the people of Silver Bay and surrounding communities as far away as the eastern shore of Wisconsin. 10) The discharge into the water substantially endangers the health of the people who procured their drinking water from the western arm of Lake Superior, including the communities of Silver Bay, Beaver Bay, Two Harbors, Cloquet, Duluth (Minnesota), and Superior, Wisconsin.

Defendants Reserve, Armco, and Republic2 noticed their appeal to this court and moved for a stay of the district court's injunction pending the appeal. Judge Lord denied this request and Reserve applied to us for a stay.3

Upon consideration of Judge Lord's opinion of April 20, 1974, the written motion and supporting documents presented by Reserve, and after hearing the arguments of counsel representing appellants and appellees, we entered our order on April 22, 1974, granting a short stay of the temporary injunction until the merits of the motion could be fully heard and decided upon full briefs and extended arguments by both sides. We scheduled a full hearing before the court on May 15, 1974, at the federal courthouse in St. Louis. We have now held that hearing and considered all material pertinent to the motion for stay, including Judge Lord's supplemental memorandum dated May 11, 1974, consisting of 109 typewritten pages of findings of fact and conclusions of law, expanding on his April 20th opinion. The question now before us is whether, considering all facts and circumstances, the injunction order should be stayed panding Reserve's appeal. We grant the stay subject to certain conditions and limitations as stated herein.

A. The Substance of the Controversy. A. The Substance of the Controversy.

Although there is no dispute that significant amounts of waste tailings are discharged into the water and dust is discharged into the air by Reserve, the parties vigorously contest the precise nature of the discharge, its biological effects, and, particularly with respect to the waters of Lake Superior, its ultimate destination. Plaintiffs contend that the mineral cummingtonitegrunerite,4 which Reserve admits to be a major component of its taconite wastes and a member of the mineral family known as amphiboles, is substantially identical in morphology (or shape and form) and similar in chemistry to amosite asbestos, a fibrous mineral which has been found, in certain occupational settings, to be carcinogenic. The plaintiffs further argue that the mineral fibers discharged represent a serious health threat, since they are present in the air of Silver Bay and surrounding communities and, by way of dispersion throughout Lake Superior, in the drinking water of Duluth and other communities drawing water from the lake.

Reserve has maintained throughout this litigation that its cummingtonitegrunerite does not have a fibrous form and is otherwise distinguishable from amosite asbestos. Reserve further maintains the tailings cannot be said to pose any health hazard and, in any event, with respect to its discharge into water, the tailings largely settle to the bottom of the lake in the 'great trough' area within close range of the plant.

The evidence presented on these points was extensive and complex. There was testimony as to the comparisons of mineralogy between Reserve's cummingtonite-grunerite and amosite asbestos, based on electron miscroscope analysis of morphology, x-ray diffraction analysis of crystal structure, and laboratory analysis of chemical composition. As for the dispersion through Lake Superior, there was considerable testimony as to whether Reserve was the sole source of cummingtonite-grunerite in the lake and whether the presence of the mineral could thus be used as a 'tracer' for Reserve's discharge. The district court found, as plaintiffs contended, that Reserve discharged particles identical to and similar to amosite asbestos, and that the particles discharged into Lake Superior were dispersed widely.

The suggestion that particles of the cummingtonite-grunerite in Reserve's discharges are the equivalent of amosite asbestos raised an immediate health issue, since inhalation of amosite asbestos at occupational levels of exposure is a demonstrated health hazard resulting in asbestosis and various forms of cancer. However, the proof of a health hazard requires more than the mere fact of discharge; the discharge of an agent hazardous in one circumstance must be linked to some present or future likelihood of disease under the prevailing circumstances. An extraordinary amount of testimony was received on these issues, and the district court designated several court witnesses to serve as impartial sources of review and evaluation.

B. What the District Court Decided.

The district court's conclusion-- that Reserve's discharge into Lake Superior and its emissions into the air created a common law nuisance under federal and state law, and violated state and federal pollution laws-- made the discharges subject to abatement. In determining the proper remedy, the court concededly balanced the equities and considered the economic and technological feasibility of abatement and the protection of the public health. The course of the proceedings indicate that the rather drastic remedy ordered by the district court-- the immediate closing of the plant-- was a response to the finding of a substantial danger to the public health. This is clear from the court's initial memorandum, where Judge Lord notes:

The Court has no other alternative but to order an immediate halt to the discharge which threatens the lives of thousands. In that defendants have no plan to make the necessary modifications, there is no reason to delay any further the issuance of the injunction.

We come now to our limited review of the district court's injunction ordering Reserve Mining to cease immediately its discharges into the air and water. In considering whether our temporary stay of that injunction should remain in effect, we note the usual formulation of the applicable standards to be met by the party seeking a stay under Fed.R.Civ.P. 62 and Fed.R.App.P. 8: (1) a strong showing that he is likely to succeed on the merits of the appeal; (2) a showing that, unless a stay is granted, he will suffer irreparable injury; (3) a showing that no substantial harm will come to other interested parties; and (4) a showing that a stay will do no harm to the public interest. See, e.g., Long v. Robinson, 432 F.2d 977, 979 (4th Cir. 1970).

The first element goes to the sensible administration of justice: a stay should not ordinarily be granted if the court determines that the injunction will ultimately take effect in any event. The other three elements, while distinguishable in some contexts, dissolve into a single equitable judgment: balancing the health and environmental demands of society at large against the economic well-being of those parties and local communities immediately affected. Of course, foremost consideration must be given to any demonstrable danger to the public health. See United States v. Nutrition Service, Inc., 234 F. Supp. 578, 579 (W.D.Pa.1964), aff'd, 347 F.2d 233 (3d Cir. 1965). Absent such demonstrated health danger, the public interest may arguably be served either way in environmental matters. In considering a stay application where no health hazard is shown, an appellate court must weigh, first, the seriousness of immediate harm to the environment which will result from allowing the alleged pollution to continue and, second, the economic and social dislocation to be suffered by the defendants and by the communities dependent upon them if the injunction immediately goes into effect. See Friends of the Earth v. Armstrong, 360 F. Supp. 165, 195 (D.Utah 1973); Environmental Defense Fund, Inc. v. Froehlke, 348 F. Supp. 338, 366 (W.D.Mo.1972). Chief Justice Burger, sitting as circuit justice, while refusing to grant a stay in an environmental context in Aberdeen & Rockfish R. Co. v. SCRAP, 409 U.S. 1207, 93 S. Ct. 1, 34 L. Ed. 2d 21 (1972), observed:

Our society and its governmental instrumentalities, having been less than alert to the needs of our environment for generations, have now taken protective steps. These developments, however praiseworthy, should not lead courts to exercise equitable powers loosely or casually whenever a claim of 'environmental damage' is asserted. The world must go on and new environmental legislation must be carefully meshed with more traditional patterns of federal regulation. The decisional process for judges is one of balancing and it is often a most difficult task. (409 U.S. at 1217-1218, 93 S. Ct. at 7.)

In relation to the motion for a stay, the parties have directed briefs and arguments almost exclusively to the health hazard issue. Given the concededly enormous economic impact that an immediate plant closure would have upon Reserve, given the personal impact on its approximately 3,000 employees and their families, and given the social and economic impact upon the communities in which the employees live, we think that our preliminary resolution of the health hazard question should control our action as to whether to grant or deny a stay.

A. Testimony Before the District Court.

The plaintiffs argue, and the district court found, that both Reserve's discharge into water and its discharge into air substantially endanger the health of residents in several communities around Lake Superior. We have reviewed the testimony on the health issue, giving careful and particular attention and weight, as we should at this interim stage of review, to the testimony of the impartial court witnesses and that of plaintiffs' chief medical witness, Dr. Irving Selikoff, Director of the Environmental Sciences Laboratory of Mt. Sinai School of Medicine. While not called upon at this stage to reach any final conclusion, our review suggests that this evidence does not support a finding of substantial danger and that, indeed, the testimony indicates that such a finding should not be made. In this regard, we conclude that Reserve appears likely to succeed on the merits of its appeal on the health issue. We proceed now to trace the outlines of the testimony supporting this view.

1. Two Key Unknowns.

The theory by which plaintiffs argue that the discharges present a substantial danger is founded largely upon epidemiological studies of asbestos workers occupationally exposed to and inhaling high levels of asbestos dust. A study by Dr. Selikoff of workers at a New Jersey asbestos manufacturing plant demonstrated that occupational exposure to amosite asbestos poses a hazard of increased incidence of asbestosis and various forms of cancer. Similar studies in other occupational contexts leave no doubt that asbestos, at sufficiently high dosages, is injurious to health.5 However, in order to draw the conclusion that environmental exposure to Reserve's discharges presents a health threat in the instant case, it must be shown either that the circumstances of exposure are at least comparable to those in occupational settings, or, alternatively, that the occupational studies establish certain principles of asbestosdisease pathology which may be applied to predicting the occurrence of such disease in altered circumstances.

Initially, it must be observed that environmental exposure from Reserve's discharges into air and water is simply not comparable to that typical of occupational settings. The occupational studies involve direct exposure to and inhalation of asbestos dust in high concentrations and in confined spaces. This pattern of exposure cannot be equated with the discharge into the outside air of relatively low levels of asbestos fibers. While Dr. Taylor,6 a court-appointed witness, testified that there are statistically significant higher levels of fibers in the air of Silver Bay than in the air of St. Paul, these levels still only amounted to .06267 fibers per cubic centimeter of air, far below the five fibers per cc standard for permissible occupational exposure set by the Secretary of Labor under the Occupational Safety and Health Act. It is clear that the air of Silver Bay, though polluted by a statistically significant level of 'excess' fibers, is not equivalent to the contaminated air found in an asbestos factory.

Nor can the occupational pattern of exposure be equated with the exposure resulting from the ingestion of fibers via the Duluth drinking water. This fact was confirmed by a tissue study, discussed in detail later in this opinion, in which the tissues of recently deceased Duluth residents were examined for asbestos fibers. The virtual absence of any fibers moved Dr. Brown, the principal court-appointed expert, to testify:

Q. (Dr. Pooley stated in his summary of the tissue analysis that:) 'There was no indication of an occupational exposure to asbestos dust.'

A. I agree with that.

Thus, it cannot be said that either the discharge into the water or the discharge into the air results in circumstances of exposure comparable to those in an occupational context.

If this is true, no conclusions about health hazards in occupational settings may be utilized in the present situation except on the ground that certain principles of asbestos-disease pathology may be extrapolated from relevant medical knowledge and applied in altered circumstances. The principal altered circumstance is the lower level of exposure. In order to make a prediction, based on the occupational studies, as to the likelihood of disease at lower levels of exposure, at least two key findings must be made. First, an attempt must be made to determine, with some precision, what that lower level of exposure is. Second, that lower level of exposure must be applied to the known pathology of asbestos-induced disease, i.e., it must be determined whether the level of exposure is safe or unsafe.

Unfortunately, the testimony of Dr. Arnold Brown8 indicates that neither of these key determinations can be made. Dr. Brown testified that, with respect to both air and water, the level of fibers is not readily susceptible of measurement. This results from the relatively imprecise state of counting techniques and the wide margins of error which necessarily result, and is reflected in the widely divergent sample counts received by the court. We cite some of the relevant testimony by Dr. Brown:

It is reasonable to assume an error in the counts of fibers in both water and air of at least nine times on the high side to one-ninth on the low side.

Well, in my thinking, Your Honor, as far as using the information that is supplied to the Court on counts, in view of the incredibly large errors associated with this procedure, I would have to assume virtually only a qualitative base for what is being reported. In other words, Your Honor, I do not recall having been exposed to a procedure with an error this large, and which people have seriously proposed a number based on this very poor procedure. This (poor counting procedure) has somewhat wider implications, Your Honor, than just to the question of the presence of fibers in air and water.

It has become, at least in my thinking, the kind of soft underbelly, if you will, of all of the research, much of the research, I should say, of the work that has been reported on asbestos in patients and in experimental animals.

Well, I interpret the information that was obtained in the water study as indicating that there are amphibole fibers, certainly, in the water of Lake Superior and in the water supply in Duluth. But I have little confidence in the estimate of the numbers. Which is another way of saying, Your Honor, that I consider the counts that I have been exposed to as qualitative, not quantitative in their implications. The same to be true in air.

Well, analytical tests are divided into two broad categories, one a qualitative test which is used merely to indicate the presence (or) absence of something. The quanitative kind of tests are used to quantify or measure with some degree of precision how much of that substance is present.

This testimony indicates that little more can be said about the level of fibers present in air or water other than that some fibers are present.

Even assuming that one could avoid imprecision and uncertainty in measuring the number of fibers at low levels, there remains vast uncertainty as to the medical consequences of low levels of exposure to asbestos fibers. In order to predict the likelihood and magnitude of disease resulting from exposure, one must have some idea of the relevant threshold value and dose-response relationships.9 Although there seems to be agreement that threshhold values and dose-response relationships are observable with respect to cancer generally, the particular values and relationships associated with asbestos-induced cancer appear to be unknown. Regarding this, Dr. Brown testified:

Q. What is your view on the doseresponse relationship between asbestos and cancer?

A. I would like to approach that from the general point of view first. It is my view that in virtually-- well, I will strike the word 'virtually'-- in every carcinogen that I know in which it has been possible to study whether or not a dose-response relationship exists, there has, indeed, been such a dose-response relationship. That is the general statement.

Proceeding from the general to the specific, I would have to say that there is no reason for me to believe that asbestos (is) other than like any other agent, chemical, that causes cancer. So in a hypothetical way I would say that, yes, there is a dose-response relationship between asbestos and the development of cancer. I haven't the foggiest idea, Your Honor, as to what that level might be, either in air or in water.

In commenting on the statement, 'This suggests that there are levels of asbestos exposure that will not be associated with any detectable risk,' Dr. Brown stated:

As a generalization, yes, I agree to that. But I must reiterate my view that I do not know what that level is.

2. The Tissue Study.

In the face of these two key unknowns in evaluating the risk of disease, we must now look to additional important evidence, a tissue study conducted at the request of the court and designed to measure the hazard to Duluth residents of ingesting fiber-contaminated water. This study was prompted by an almost complete lack of knowledge with respect to the human ingestion of asbestos fibers, since previous experiments had dealt largely with the effects of fiber inhalation, where interaction by asbestos with the respiratory tract was established. Any theory attempting to deal with the effects of ingestion of asbestos in liquid had to bridge the gap between the ingestion of fibers and the interaction by those fibers with the body tissues. If the fibers do not interact with the tissues but simply are eliminated by the body as wastes, presumably no disease will result. Accordingly, the court-appointed experts formulated a 'protocol' or study plan designed to test whether people who drink Lake Superior waters accumulate asbestos-like fibers in body tissues from taconite.

This protocol involved analysis by electron microscope of the tissues of recently deceased Duluth residents who had ingested Duluth water for at least 15 years, that is, since the beginning of operations by Reserve. As a 'control' check on results, samples were taken from the residents of Houston, Texas, where the water is free of asbestos fibers. Although this study was necessarily expedited, plaintiffs' principal medical witness, Dr. Selikoff, accepted its likely value:

Q. And you regard that as a sound protocol which will provide significant information?

A. I think so, sir.

* * * I think that their study has great value and we certainly would like to see the results of it.

Those results, as explained to the court by one of its own experts, Dr. Pooley,10 indicated that the tissues of Duluth residents were virtually free of any fibers which could be attributed to the Reserve discharge.11 Dr. Brown summarized the results as follows:

It's my conclusion, from the tissue study, that residents of Duluth have not been found to have asbestiform fibers in their tissues when compared with Houston.

The significance which may be attributed to these results is a matter of dispute. Before the results were in, Dr. Selikoff had no hesitancy in forecasting that negative results would possess substantial significance, and stated:

Now, our feeling was that no matter what air samples show or water samples show or anything else, unless it is found that asbestos is in the tissues of people who have drunk this water * * * if we do not find it in the tissues in appreciable quantities, then I would risk a professional opinion that there is no danger, at least up to this point, to the population no matter what our samples show or water samples.

Moreover, Dr. Brown, while suggesting that the tissue study did not exonerate Reserve's operations as a hazard, did state:

It does tell me that it is not an emergency situation, and that's about as far as I can go.

Plaintiffs, however, sought subsequently to discount the significance of these tissue studies. First, the argument was made, and was accepted by the district court, that the specimens of tissue from body organs surveyed were too minute, and thus fibers that were present may have been overlooked. However, this judgment must be balanced by the following exchange between the court and Dr. Selikoff, which occurred prior to the conclusion of the study:

THE COURT: And if we examine those tissues and there are not fibers there, this operation is home-free as far as the health hazard goes?

THE WITNESS (Dr. Selikoff): I would think we should find some fibers there. We're looking for needles in a haystack, but that's all right, we should find needles in the haystack with all the difficulties of the study, the technical difficulties, if we examine sufficiently large numbers of samples in some instances we should find some fibers there.

As noted previously, Dr. Selikoff had stated his belief that the design of the protocol was sound and would yield significant information. The weight to be given the tissue study has also been challenged on collateral grounds, most notably by testimony as to the increased rate of gastro-intestinal cancer among workers occupationally exposed to asbestos. This increased rate of gastro-intestinal cancer, it was theorized, is the result of asbestos workers' ingesting or swallowing particles during periods of exposure. However, a theory of ingestion is by no means the only basis for explaining the increased rates of gastro-intestinal cancer, and there is no scientific or medical certainty regarding the mechanism actually involved.

Although, based on our review of the testimony, we agree with Dr. Brown's statement that the ingestion of asbestos fibers cannot be exonerated as a hazard, we feel that, on any fair reading of the circumstances of the protocol, the results of the tissue study must weigh heavily against the assessment of any demonstrated hazard to health. We think it is clear that the tissue study raises a major obstacle to the proof that ingestion of Duluth water is hazardous. The study can only be overcome either by questioning the initial assessment, joined in by plaintiffs' chief medical witness, that the protocol was sound and would yield significant results, or by accepting the proposition that theoretical medical opinions alone may overcome direct evidence. Neither alone nor together can these two bases for attacking the study amount to affirmative evidence that ingestion of the water is harmful. At most, they only provide a theoretical but unsubstantiated basis for leaving the results of the study in doubt, and leaving the effects of ingestion unresolved.

B. Evaluation of Testimony.

A fair review of this impartial testimony by the court's own witnesses-- to which we necessarily must give great weight at this interim stage of review-- clearly suggests that the discharges by Reserve can be characterized only as presenting an unquantifiable risk, i.e., a health risk which either may be negligible or may be significant, but with any significance as yet based on unknowns. This conclusion is simply a logical deduction from the following facts: (1) that fiber levels are not at occupational levels; (2) that the low levels present cannot be expresed or measured as a health risk; and (3) that, in any event, threshold values and dose-response relationships are undetermined. In other words, it is not known what the level of fiber exposure is, other than that it is relatively low, and it is not known what level of exposure is safe or unsafe. Finally, no basis exists, save a theoretical one, for assuming that drinking water, otherwise pure but containing asbestoslike particles, is dangerous to health.

We are confirmed in our evaluation by the testimony of Dr. Brown:

I suspect I have stated my opinion on this before, and I don't recall now the precise words that I used, but I suspect I could sum it up by saying that in my review of the scientific literature that I would be unable to predict from that information that there will be an adverse effect that will occur in the citizens of Duluth or Silver Bay based on the presence of fibers in their environment.

Moreover, we believe there is essential agreement with this point by Dr. Selikoff, plaintiffs' chief medical witness. Although perhaps more expressive than Dr. Brown of the medical concerns which are posed by the unknown risks of Reserve's discharges, Dr. Selikoff seems to recognize that the likelihood of increased disease attributable to the discharges cannot be estimated:

We have no quantitative information on the relationship between the burden in tissues and the possible health effects.

Q. You don't have evidence that fibers, if they were ingested with the drinking water of Duluth would, in fact, occasion a public health hazard?

A. No * * *.

In the final analysis, Dr. Selikoff's vigorously expressed judgment that the discharges represent a danger reflects his underlying opinion that:

I advocate that where, to the extent feasible, that carcinogenic substances should not be added to the environmental burden in which we live.

Because asbestos is a known human carcinogen we have to know a lot more about it than we now know in order to establish safe limits for its presence. I have no expectations that our environment will ever be entirely free of asbestos. But until we know what safe limits are, as a physician, who would rather see well people than sick people, I have some sort of compulsion to protect ourselves against known agents that produce cancer until we know what the safe levels are.

That is my entire base.

We think Doctors Brown and Selikoff share a common medical concern, but are essentially in agreement that the discharges here simply have not been provien to be a demonstrable hazard.

Interestingly enough, although Judge Lord does on occasion speak of a substantial health danger, based on the of asbestiform fibers in the environment, a careful reading of his initial and supplemental memoranda reveals an acknowledgement, at least in certain instances, that any ill effects are simply beyond proof. This is indicated in the 'negative' findings which Judge Lord sometimes makes. For example, Judge Lord responds to the small number of excess deaths from rectal cancer in the Duluth area by finding:

We cannot say that the increase seen, although small in number at this time, is not due to ingestion by these persons of asbestos from Reserve's taconite waste.

Of course, neither can the opposite be said.12 Perhaps the most revealing statement of all is contained in Judge Lord's initial memorandum:

The state of the scientific and medical knowledge available in this area is in its early stages and there is insufficient knowledge upon which to base an opinion as to the magnitude of the risks associated with this exposure.

Considering all of the above, we think one conclusion is evident: although Reserve's discharges represent a possible medical danger, they have not in this case been proven to amount to a health hazard. The discharges may or may not result in detrimental health effects, but, for the present, that is simply unknown. The relevant legal question is thus, what manner of judicial cognizance may be taken of the unknown.

We do not think that a bare risk of the unknown can amount to proof in this case. Plaintiffs have failed to prove that a demonstrable health hazard exists. This failure, we hasten to add, is not reflective of any weakness which it is within their power to cure, but rather, given the current state of medical and scientific knowledge, plaintiffs' case is based only on medical hypothesis and is simply beyond proof.

We believe that Judge Lord carried his analysis one step beyond the evidence. Since testimony clearly established that an assessment of the risk was made impossible by the absence of medical knowledge, Judge Lord apparently took the position that all uncertainties should be resolved in favor of health safety. Since the appropriate threshold level for safe toleration of fibers was unknown, the district court tipped the balance in favor of attempting to protect against the unknown and simply assumed that Reserve's discharge presents a health hazard. In doing so, he disregarded the tissue studies of his own experts which provided direct evidence to the contrary. If we are correct in our conclusion that evidence does not exist in the record on which to find Reserve's discharges to be unsafe, the district court's determination to resolve all doubts in favor of health safety represents a legislative policy judgment, not a judicial one. See Industrial Union Department, AFL-CIO, et al. v. Hodgson, No. 72-1713, 499 F.2d 467, at 474 (D.C.Cir., filed April 15, 1974). As Judge McGowan stated in Industrial Union Department, AFL-CIO, with regard to legislative resolution of a similar issue:

Some of the questions involved in the promulgation of these standards are on the frontiers of scientific knowledge, and consequently as to them insufficient data is presently available to make a fully informed factual determination. Decision making must in that circumstance depend to a greater extent upon policy judgments and less upon purely factual analysis. (Id. at 474)

We emphasize that our evaluation rests not on any view that the discharge exposes North Shore residents to no risk, but rather on the view that, given the evidence, no substantial danger has been, or could be proven. It cannot be said, other than as a matter of conjecture, that the discharges will result in any higher incidence of disease than that experienced by a general public not similarly exposed. Although we are sympathetic to the uncertainties facing the residents of the North Shore, we are a court of law, governed by rules of proof, and unknowns may not be substituted for proof of a demonstrable hazard to the public health.

After our examination of the relevant portions of the lengthy record in this case, we come to these conclusions: (1) it is unlikely that Reserve will prevail on the merits of the pollution issue and overcome the trial court's determination that pollution of Lake Superior must be abated; (2) it is also unlikely that Reserve will overcome the trial court's determination that the air emission must be controlled; but (3) Reserve, as we have demonstrated, may well prevail in its contention that its emissions into the air and water have not proven to be a substantial health hazard. If and when our court is called upon, on the appeal on the merits, to affirm or modify the terms of the injunction based upon the existence or nonexistence of a substantial health hazard, we can forecast, on the showing presently made, that Reserve will succeed on the health hazard issue. However, the district court's findings on the pollution aspects of this case appear to be based on sufficient evidence, apart from mere speculation, to withstand our initial scrutiny. Therefore, in the posture of this motion, we cannot forecast any success for Reserve in upsetting these findings relating to pollution, and are satisfied that the circumstances call for a stay conditioned upon assurances that there will be a speedy termination of Reserve's discharges into Lake Superior and control of its emission into the air.

The controversy between Reserve and governmental agencies over alleged pollution of Lake Superior has existed for more than five years. In retrospect, it must now be painfully clear to all who participated in the original decision to permit the discharge of tailings into Lake Superior, that such a decision amounted to a monumental environmental mistake. The actors in that decision, 25 years ago, included leading citizens and governmental officials of Minnesota as well as officials of Reserve, Armco, and Republic Steel. That decision obviously was made in good faith to create jobs, to provide other economic opportunities in an economically depressed area of northern Minnesota, and to utilize the almost unlimited supply of hitherto unusable, low-grade, taconite ore found in that area. To us there are neither heroes nor villains among the present participants in this lawsuit, nor among their predecessors in government, business, and society who were once allies in encouraging and creating a taconite industry in northern Minnesota. Nevertheless, the pollution of Lake Superior must cease as quickly as feasible under the circumstances.

Accordingly, our stay of the injunction will be conditioned upon Reserve taking prompt steps to abate its discharges into air and water. We invited Reserve to advise this court concerning plans for the on-land disposal of its tailings and the significant control of its air emissions. Reserve's counsel stated that the company envisioned a three and one-half year to five year 'turn-around' time, but added that investigation continues in an effort to reduce further the time for achieving abatement.

Our stay of the injunction rests upon the good faith preparation and implementation of an acceptable plan.13 Therefore, we grant a 70-day stay upon these conditions:

1) Reserve's plans shall be promptly submitted to plaintiff-states and to the United States for review and recommendations by appropriate agencies concerned with environmental and health protection. Such plan shall be filed with the district court and submitted to all plaintiffs in no event later than 25 days from the filing of this order.

2) Plaintiffs shall then have an additional 20 days within which to file their comments on such plan.

3) The district court shall consider Reserve's plan and any recommendations made by the United States and plaintiff-states and make a recommendation, within 15 days following submission of plaintiffs' comments, whether or not a stay of the injunction should be continued pending the appeal.14

4) Based on these plans, comments, and recommendations, this court will then review the status of its stay order within the time remaining.

We remand this aspect of the case to the district court for its recommendation, such recommendation and record to be delivered to us no later than 60 days from the filing of this order.

We add a comment. At oral argument counsel for all parties indicated that the controversy ought now to be settled. Counsel for Reserve suggested a Master or Referee ought to be appointed who could 'in some way establish some mechanism that could focus on whether or not there is a reasonable basis for resolution of this controversy.' Counsel for the United States indicated amenability to settlement, stating 'I think this case ultimately has got to be settled.' Counsel for the State of Minnesota at argument advised the court that Judge Lord during trial contemplated keeping Reserve in operation conditioned upon future settlement.

We think settlement of this kind of case represents a necessary and desirable goal, and the matter should be pursued in the district court. One possible approach is the use of mediation techniques in resolving disputes on technical matters by utilizing experts from business and science who can assist the parties in reaching a settlement or can advise the district court on technical matters in his consideration of Reserve's plan. Judge Lord has demonstrated innovative techniques in trying this case by bringing into court some of the world's leading scientists, physicians, and other experts in seeking resolution of the controversy before him. Although we do not find ourselves, at least at this point of the lawsuit, in full agreement with Judge Lord's views on the health issue, this disagreements in no way detracts from our confidence in his ability to assist the parties to achieve a meaningful and appropriate settlement.

Finally, we emphasize that our stay is also conditioned upon prompt presentation of the merits of this appeal. The court authorizes the preparation of a deferred appendix pursuant to Fed.R.App.P. 30(c), and consents to briefs not to exceed 100 pages, subject to the further order of this court. The briefing schedule may be expedited at the request of Reserve or of any of the government parties.15

1) On appeal, Reserve is likely to prevail on the merits of whether the evidence supports the district court's finding that Reserve's emissions into the air and its discharges into the water substantially endanger the public health.

2) The plaintiffs (state and federal governments and environmental groups) are likely to prevail on the merits of whether the evidence establishes that Reserve's present operations otherwise pollute the air and water.

3) It is likely that abatement will ultimately be required upon reasonable terms that take into consideration the multiple environmental, social, and economic interests involved.

4) We grant a stay of injunction for 70 days upon conditions stated in item 5 herein, and as amplified in the opinion.

5) This stay is conditioned upon a showing by Reserve that it is taking prompt steps to prepare and implement an appropriate plan for abatement, which it has represented to the court that it is now ready to do.

6) We remand this case to the district court for the purpose of receiving its recommendation for continuance of the stay order, after it has considered Reserve's plan and plaintiffs' comments on that plan, and for the further purpose of assisting the parties in attempting to reach agreeable settlement of the controversy, such settlement to be consistent with the public interest.

It is so ordered.