This is a takings case. It involves an almost decade-long dispute between the Government and the property owners over the distinction between a regulatory taking and a taking by permanent physical occupation, and what the scope of proper discovery should be in such a case. The case began with the efforts of the United States, acting through the U.S. Environmental Protection Agency (Government or EPA), to combat ground water pollution from a major hazardous waste site, the Stringfellow Acid Pits in California. Utilizing its authority under CERCLA,1 commonly known as Superfund, the Government, in conjunction with the State of California, undertook a broad-ranging attack on the problem.

As part of this attack, the Government decided to locate ground water wells and associated equipment in the general area of the acid pits to monitor the movement of contaminated ground water. The area of concern included not only the site of the acid pits, but nearby properties as well. Plaintiffs own one of those nearby properties. This case is about the Government's enlistment of plaintiffs' property, without plaintiffs' consent, in the battle against pollution, and plaintiffs' efforts to be recompensed for that use.

In 1983 the Government first undertook activities on plaintiffs' land. In 1984 plaintiffs filed suit in the Claims Court for just compensation for the alleged taking of plaintiffs' property. The suit dragged on for years, with two different trial judges. Extensive discovery was undertaken. Before the first trial judge, the Government moved for summary judgment that there had been no regulatory taking, and that the United States was not responsible for the activities on the property undertaken by the State of California. The court granted both motions. The plaintiffs moved for summary judgment that there had been a physical occupation amounting to a taking. The court denied plaintiffs' motion, and ordered trial. (Hendler v. United States, 11 Cl.Ct. 91 (1986) (Hendler I )).

Then the matter took a sudden and rather unusual turn. Plaintiffs and the Government became embroiled in a heated quarrel over whether plaintiffs' responses to certain interrogatories were sufficiently detailed and forthcoming. At the Government's insistence, the Claims Court ordered the plaintiffs to submit additional and more complete responses to the Government's questions. Plaintiffs responded that they were answering fully in light of the information available to them, and that since certain of the Government's questions focused on finding out more about the activities of the Government itself, plaintiffs were continuing discovery of Government witnesses. The Government, and ultimately the court, were not satisfied, and in December of 1989 the suit, now before a second Claims Court judge, was dismissed under Rule 37--the court granted the Government's motion to sanction plaintiffs by dismissing the case with prejudice and awarding court costs and attorney fees to the Government. Hendler v. United States, 19 Cl.Ct. 27 (1989) (Hendler II ).

Plaintiffs ask that we review the correctness of that dismissal, as well as the correctness of the earlier rulings on the motions for summary judgment. The Government is of the view that the dismissal was fully deserved and should be upheld. And even should we disagree, the Government urges, the Claims Court's decisions on the earlier summary judgment motions did not themselves finally dispose of the case and therefore are not properly before us.

We find the dismissal sanction was improvidently granted--it is reversed. And, for the reasons we shall shortly explain, we conclude that the merits of the rulings in Hendler I are before us, and should be decided. On the merits, we conclude that the partial summary judgment in favor of the Government on the question of the role of the State of California was also in error, and it is reversed. The ruling that no regulatory taking occurred is not disturbed, but, as we shall explain, only in so far as it is limited to the EPA Order per se.

The denial of plaintiffs' motion for summary judgment on a taking by physical occupancy is reversed; judgment is ordered in favor of plaintiffs. The case is remanded to the Claims Court for further proceedings consistent with this opinion.

As a general proposition, when a trial court disposes finally of a case, any interlocutory rulings 'merge' with the final judgment. Thus both the order finally disposing of the case and the interlocutory orders are reviewable on appeal. Some courts have carved out an exception: if the final order is a dismissal resulting from bad faith or dilatory conduct, then the interlocutory orders do not 'merge' and do not become reviewable.2 To hold otherwise would open up a back door route to review of interlocutory orders and would reward bad conduct.

The Government argues that that is the case here. And it is true that the trial judge in Hendler II characterized plaintiffs' responses as indicating bad faith. The record is devoid, however, of any factual underpinning for that characterization beyond the apparent conclusion that in the court's judgment plaintiffs should have had more to tell. But failure to set forth in discovery facts sufficient to establish a cause of action is not proof of bad faith, and is not a basis for a Rule 37 sanction.3 Ingalls Shipbuilding, Inc. v. United States, 857 F.2d 1448, 1451 (Fed.Cir.1988).

Further, the trial judge in Hendler II necessarily viewed plaintiffs' answers to the interrogatories in the light of the issues remaining before him. Since we are of the view that those issues had been skewed by erroneous rulings in Hendler I, this case is not one in which plaintiffs acted wrongly, but rather one in which errors of law led to an erroneous dismissal. No Rule 37 sanction is appropriate when a litigant's failure is not the result of bad faith or misconduct. Societe Int'l v. Rogers, 357 U.S. 197, 212, 78 S. Ct. 1087, 1095, 2 L. Ed. 2d 1255 (1958).

When errors of law in the underlying orders sent the trial court down the wrong path, courts have reviewed the merits of the underlying decisions, as well as the final dismissal sanction. For example, in Familias Unidas v. Briscoe, 544 F.2d 182 (5th Cir.1976), the trial court dismissed the plaintiffs after they refused to answer interrogatories seeking to establish the identities of members of an unincorporated body organized to address deficiencies in the local public school system. Prior to the Rule 37 dismissal, the trial court had denied plaintiffs' requests for a temporary restraining order, a three-judge court, and for declaratory judgment.

The action was originally filed as a class action, but was amended to name only one party in interest after the trial court denied plaintiffs' motion to strike these interrogatories. The Fifth Circuit found the dismissal to be an abuse of discretion, particularly in light of the amended complaint which dropped the class action allegations. 544 F.2d at 192. This abuse of discretion had as its source an erroneous interpretation of a Supreme Court decision, N.A.A.C.P. v. Alabama, 357 U.S. 449, 78 S. Ct. 1163, 2 L. Ed. 2d 1488 (1958), which the Fifth Circuit corrected. The Fifth Circuit then went on to review the merits of the other rulings. It affirmed the trial court's decisions regarding the denial of injunctive relief and a three-judge court, but reversed and remanded with regard to the declaratory judgment issue.

In Allied Air Freight v. Pan American World Airways, Inc., 393 F.2d 441 (2d Cir.), cert. denied 393 U.S. 846, 89 S. Ct. 131, 21 L. Ed. 2d 117 (1968), the Second Circuit reversed the calendar judge's dismissal for failure to prosecute, vacated the trial court's stay pending exhaustion of administrative remedies, and remanded to the trial court for further proceedings. Here, the legal error lay in the initial stay. The Second Circuit explicitly rejected Pan Am's contention that the appellate court should limit itself to reversing the dismissal:

[I]n the interests of efficient judicial administration, we conclude that we should review the interlocutory stay order in this appeal from a final order dismissing the action....

Id. at 443.

The ... error that we see in the issuance of the 90 day order and the dismissal under its terms was that requiring exhaustion of available administrative remedies as a condition precedent to further action in the district court, in accordance with the terms of the stay, was an erroneous application of the doctrine of primary jurisdiction which imposed an undue burden upon Allied. Therefore we consider the merits of appellant's argument that the district court erred in issuing the interlocutory stay order.

Id. at 445.

Our facts are directly analogous to the situation in Allied Air Freight. The underlying rulings were the basis for the conclusions that led to the dismissal. Following the Government's proposed course of action would result in reversing the Rule 37 dismissal and simply remanding for further disposition premised on the earlier rulings, the course of action explicitly rejected by the Allied court. Presumably, Hendler I would continue to be followed as law of the case. If so, Hendler I would compel the trial judge to make further rulings applying erroneous law. When the ensuing trial on the merits was complete, a court on appellate review would again reverse, and yet another trial ensue. This is a clear waste of judicial resources. Further, since the matter here turns on errors of law rather than on disagreement over the facts of the particular case, there is no question of transgression upon "the appropriate relationship between the [appellate and trial] courts...." Coopers & Lybrand v. Livesay, 437 U.S. 463, 476, 98 S. Ct. 2454, 2462, 57 L. Ed. 2d 351 (1978). We believe the approach taken in Familias and Allied Air Freight is sound. Thus we must review and decide both the issue of the dismissal and any underlying erroneous rulings.

Plaintiffs own property in Riverside County, California near the hazardous waste disposal site known as the Stringfellow Acid Pits (Stringfellow). EPA became aware that a plume of ground water contaminated with toxic substances from Stringfellow was threatening to enter a nearby source of drinking and agricultural water. EPA requested access to plaintiffs' property to install wells for monitoring and extracting these migrating hazardous substances. Plaintiffs refused.

In September 1983, EPA issued an administrative order granting itself and the State of California access to plaintiffs' property for, inter alia, "locating, constructing, operating, maintaining, and repairing monitor/extraction wells." Henry Hendler, Order (EPA Sept. 20, 1983); see infra note 10. Shortly after, EPA went upon plaintiffs' property and began the installation of a series of wells; five were installed by contractors for the EPA, and (by the time of the first hearing) at least another thirteen by the State of California.

Plaintiffs filed their initial complaint on September 5, 1984.4 Plaintiffs alleged that the EPA's actions constituted a taking of their property; they sought $4.5 million as compensation. The Government filed its answer on November 1, 1984. After an initial court conference at which the parties agreed that no material facts were in dispute, the court approved a briefing schedule that set May 29, 1985 as the final deadline for the filing of responses to discovery requests. A few weeks later, the parties filed Joint Fact Stipulations which set out the events leading up to the law suit; the document detailed the five wells placed on the property by the EPA and the thirteen installed by the State of California pursuant to its role under a cooperative agreement with the EPA. Hendler I at 98.

On June 24, 1985, plaintiffs moved for an order permitting additional discovery of activities occurring on the property after May 29, 1985. The motion was granted. Subsequently, plaintiffs filed a motion for summary judgment and the Government filed a cross-motion for summary judgment. Hendler I at 93.

In Hendler I, the trial judge, in his memorandum order of October 24, 1986, determined that

(1) the mere issuance of the EPA Order of September 1983 did not constitute a regulatory taking of any of plaintiffs' property (the court did not address the question of whether the Order as subsequently applied might be deemed such a taking);

(2) the record afforded insufficient evidence upon which to base a decision whether there had been a taking by physical occupation--the court wanted to know more about the Government's long-range intentions; and

(3) the activities of the State of California were not attributable to the Government for purposes of takings law.

The Government's motion for summary judgment was granted on points one and three. Point two was left for trial--the plaintiffs' motion for summary judgment on the issue of taking by physical occupation was denied.

Both before and after the trial court's disposition of the summary judgment motions in Hendler I, the discovery skirmishes continued. Both parties continued to engage in discovery throughout the briefing period prior to Hendler I. On December 5, 1985, the Claims Court granted the Government's first motion to compel answers to its interrogatories. The Government believed plaintiffs' responses to be inadequate and on March 25, 1986 again moved to compel answers to its interrogatories.

On January 14, 1987, the court again granted a Government motion to compel answers to interrogatories. On March 16, 1987, the Government moved to dismiss plaintiffs' suit for failure to comply with the January 14 order. Plaintiffs submitted their responses on March 26. The Government found the responses inadequate and again sought dismissal. On April 30, the court issued a third order to compel a full response to the Government's interrogatories. Plaintiffs responded on May 14 and again the Government deemed the responses inadequate. On May 26, the Government again filed a motion to dismiss.

Discovery closed August 14, 1987. On October 21, 1987, now before a new judge, the parties argued the motion to dismiss. On September 16, 1988, plaintiffs moved to suspend the proceedings on the basis of new information indicating that the Government intended to install extraction wells on the property. The Government objected and moved for sanctions.

In December 1989, over two years after the motion was argued, the judge granted the Government's motion to dismiss pursuant to Rule 37, bringing this acrimonious procedural fray to an abrupt halt. Hendler II.B.

A man's home may be his castle, but that does not keep the Government from taking it. As an incident to its sovereignty, the Government has the authority to take private property for a public purpose. In England, when the King did so, payment for what was taken was, at least before Magna Carta, a sometime thing, and largely at the discretion of the sovereign. See 1 P. Nichols, The Law of Eminent Domain § 2 (1917). In the United States, the Constitution does not leave that issue to the sovereign's good will. The ringing phrases of the Fifth Amendment conclude with the simple statement: "nor shall private property be taken for public use, without just compensation." U.S. Const. amend. V.

At the time of the writing of the Constitution and for many years thereafter a government taking meant exactly that--the Government would physically occupy the land. If the Government needed a place for a military base, or for a building to house a government office or activity, that was a public purpose for which a government taking was authorized upon payment of just compensation. Much of the law of eminent domain--both statutory and case--developed for the purpose of providing the procedural structure for government takings; the main issue in the cases was what compensation was just. See J. Sackman, Nichols' The Law of Eminent Domain § 8 (1991); see generally L. Orgel, Valuation under the Law of Eminent Domain (2d ed. 1953).

As government activities expanded, situations arose in which government action resulted in an invasion of an owner's private property, but the government had not undertaken the procedural steps called for by statute to acquire the affected property interests. For example, the government's roadbuilding activity on A's land, the taking of which was authorized and paid for by the government, might cause permanent flooding on the nearby land of B. The suit by B, to require the government to pay just compensation for the taking of B's property as well, acquired the name of inverse condemnation.

Since the suit was based upon the constitutional provision protecting property rights, and the provision was considered to be self-executing with respect to compensation, it escaped the problems of sovereign immunity. See Jacobs v. United States, 290 U.S. 13, 54 S. Ct. 26, 78 L. Ed. 142 (1933); 3 J. Sackman, supra, at §§ 8.01 & 8.01[a].5 Whether by planned acquisition through the exercise of eminent domain, or a payment after the fact through inverse condemnation, there was routinely present in these cases the fact of physical entry upon and occupation of private property by the government.

Occasionally an issue arose as to whether the government's activity was so short lived as to be more like the tort of trespass than a taking of property. The distinction between the government vehicle parked one day on O's land while the driver eats lunch, on the one hand, and the entry on O's land by the government for the purpose of establishing a long term storage lot for vehicles and equipment, on the other, is clear enough. The fact that sovereign immunity might insulate the government from liability in the tort but not in the taking makes for interesting line-drawing in the close cases, and provides employment for lawyers. See 1 P. Nichols, supra., at § 113.

Traditional takings doctrine, based as it is on the indicia of physical occupation of land, does not fit easily into the issues that arose with the emergence of the regulatory state. In an early case Justice Holmes declared this basic proposition: "the general rule at least is, that while property may be regulated to a certain extent, if regulation goes too far it will be recognized as a taking." Pennsylvania Coal Co. v. Mahon, 60 U.S. 393, 415, 43 S. Ct. 158, 160, 67 L. Ed. 322 (1922).

A few years later the Supreme Court was confronted with what had emerged as the government's most pervasive mechanism for regulating land use: zoning. While upholding zoning as a constitutionally permissible activity in general, the Court reserved in a passing nod to Justice Holmes the thought that there could be specific instances in which a zoning regulation would be so intrusive as to violate the constitutional norm. Village of Euclid v. Ambler Realty Co., 272 U.S. 365, 395, 47 S. Ct. 114, 121, 71 L. Ed. 303 (1926). Two years later, in Nectow v. City of Cambridge, 277 U.S. 183, 48 S. Ct. 447, 72 L. Ed. 842 (1928), that passing thought became law, and a specific zoning regulation was held constitutionally invalid as applied to plaintiff's property.

For the next fifty years the Supreme Court left it largely to the state courts to sort out the validity of zoning ordinances and other regulatory enactments affecting land use.6 Those that were found too intrusive--had gone 'too far'--were struck down by the courts as violative of the Constitution.

Curiously, these court opinions often did not bother to mention exactly what provision of the Constitution was violated. Perhaps that explains why for many years a landowner's sole relief in a constitutional challenge based on a land use regulation was invalidation of the regulation. See, e.g., Agins v. City of Tiburon, 24 Cal. 3d 266, 157 Cal. Rptr. 372, 375, 598 P.2d 25, 28 (1979), aff'd, 447 U.S. 255, 100 S. Ct. 2138, 65 L. Ed. 2d 106 (1980). The constitutional alternative provided by the Fifth Amendment of upholding an overly-intrusive regulation by treating it as a lawful taking requiring just compensation was not an issue that attracted judicial attention.

Beginning in 1978, the Supreme Court re-entered the debate about when regulations went 'too far,' and whether the remedy was limited to declaring invalidity or whether the regulation could be upheld in exchange for just compensation. In Penn Central Transportation Co. v. New York City, 438 U.S. 104, 98 S. Ct. 2646, 57 L. Ed. 2d 631 (1978), the Court addressed the first issue: when does a regulatory taking occur. Holmes' formulation gave way to a more elaborate--but perhaps no more certain--three factor analysis: "the character of the governmental action," the "economic impact of the regulation on the claimant," and the "extent to which the regulation has interfered with distinct investment-backed expectations." Id. at 124, 98 S.Ct. at 2659. No taking was found in that case so the second issue, remedy, was not addressed.

Subsequently the Court on four separate occasions accepted but then for various procedural reasons rejected the opportunity to definitively establish whether, as a matter of federal constitutional law, an overly-intrusive regulation entitled the property owner to just compensation for whatever period the regulation was in effect. See Agins v. Tiburon, 447 U.S. 255, 100 S. Ct. 2138, 65 L. Ed. 2d 106 (1980); San Diego Gas & Elec. Co. v. San Diego, 450 U.S. 621, 101 S. Ct. 1287, 67 L. Ed. 2d 551 (1981); Williamson County Regional Planning Comm'n v. Hamilton Bank, 473 U.S. 172, 105 S. Ct. 3108, 87 L. Ed. 2d 126 (1985); MacDonaId, Sommer & Frates v. Yolo County, 477 U.S. 340, 106 S. Ct. 2561, 91 L. Ed. 2d 285 (1986).

In the San Diego Gas case, the Court came close to deciding the issue. Four Justices were of the view, expressed in strong terms by Justice Brennan, that the Constitution meant what it said--a taking, whether regulatory or physical, mandated that the property owner be paid just compensation. After-the-fact invalidation of the offending regulation would not satisfy the constitutional mandate. Chief Justice Rehnquist indicated Justice Brennan was right but the case was wrong.

Finally the issue was resolved in 1987 in First English Evangelical Lutheran Church v. Los Angeles County, 482 U.S. 304, 107 S. Ct. 2378, 96 L. Ed. 2d 250 (1987). The regulation at issue prevented the landowner from rebuilding after a flood destroyed buildings on the property. Since the complaint had alleged and the California Court of Appeals decision under review had assumed a taking, the Court treated the question of whether a regulatory taking had occurred as a given. Id. at 312, 107 S.Ct. at 2384.

The state argued that the landowner's remedy could be limited by state law simply to a determination of validity of the uncompensated regulatory enactment; only if the state insisted on enforcement after the regulation was judicially determined to require compensation would compensation be due. The Supreme Court rejected that reading of the Fifth Amendment. The Constitution, said the Court, requires just compensation for a regulatory taking from the date it occurs until the date of the regulation's rescission or amendment.

For purposes of this opinion we need not explore in detail the various articulations of what constitutes a regulatory taking that have emerged from the cases both before and after Penn Central and First Lutheran Church.7 Much ink has flowed in attempts to critique and explain.8 The Supreme Court itself likes to point out that no set formula exists to determine whether compensation is constitutionally due for a government restriction of property; instead the court must engage in "essentially ad hoc, factual inquiries." See Penn Central, 438 U.S. at 124, 98 S. Ct. at 2659; First Lutheran Church, 482 U.S. at 316, 107 S.Ct. at 2386. But at bottom what emerges is at least the basic notion that the government, under the guise of regulation, cannot take from a property owner the core economic value of the property, leaving the owner with a mere shell of shambled expectations.9

For our purposes here, the important point is that in 1989 when the trial judge dismissed the case, there were two clear lines of Supreme Court authority under which a taking for which just compensation must be paid could be held to have occurred: the traditional physical occupation theory, and the newly-developed regulatory taking theory. And by that time, a series of critical rulings had been made: the 1983 EPA Order alone was not a regulatory taking, although the question of whether that Order plus the subsequent events alleged could constitute a regulatory taking was not foreclosed; there was not yet a taking by physical occupation--the trial judge wanted to know more about the Government's long-range intentions; and the activities for which the Federal Government bore responsibility were only those it undertook directly--the activities of the State of California, although pursuant to the Order, were nevertheless not attributable to the Government. We examine each of these rulings and the impact the rulings had on the dismissal sanction.

The first issue before the trial court in Hendler I was whether there was evidence sufficient to establish a regulatory taking by the Government. The judge was of the view that there was no regulatory taking as a result of the EPA Order itself. That Order mandated that EPA officials and other authorized personnel, including state officials, were to have access to plaintiffs' property for the purposes of installing wells and related equipment such as pipes, tanks, and other storage facilities, and to carry on various activities such as storing and transporting ground water, conducting tests, extraction operations and so on. (The terms of the Order are set out in the footnote below.)10 Plaintiffs were ordered not to interfere in any manner with the EPA/state activities, under pain of significant civil penalties, including punitive damages. Hendler I at 93.

The trial judge construed this Order as "not purport[ing] to dispossess plaintiffs or limit their use of the property.... While [the Order's] terms were expansive, they were not necessarily inimical to simultaneous use of the property by plaintiffs, as long as they did not interfere with defendant's admittedly beneficent activities." Hendler I at 95.

This misconceives the meaning and purpose of the constitutional protections underlying the Fifth Amendment. The Government does not have the right to declare itself a co-tenant-in-possession with a property owner. Among a citizen's--including a property owner's--most cherished rights is the right to be let alone. Olmstead v. United States, 277 U.S. 438, 478, 48 S. Ct. 564, 572, 72 L. Ed. 944 (1928) (Brandeis, J., dissenting); see generally Warren & Brandeis, The Right to Privacy, 4 Harv.L.Rev. 193 (1890).

In the bundle of rights we call property, one of the most valued is the right to sole and exclusive possession--the right to exclude strangers, or for that matter friends, but especially the Government. See, e.g., Nollan v. California Coastal Comm'n, 483 U.S. 825 at 831, 107 S. Ct. 3141 at 3145, 97 L. Ed. 2d 677 (1987); Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 426, 102 S. Ct. 3164, 3171, 73 L. Ed. 2d 868 (1982); In re Etter, 756 F.2d 852, 859, 225 U.S.P.Q. (BNA) 1, 6 (Fed.Cir.), cert. denied, 474 U.S. 828, 106 S. Ct. 88, 88 L. Ed. 2d 72 (1985) ("The essence of all property is the right to exclude...."); Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1548, 220 U.S.P.Q. (BNA) 193, 198 (Fed.Cir.1983) (citing Schenck v. Nortron Corp., 713 F.2d 782, 218 U.S.P.Q. (BNA) 698 (Fed.Cir.1983)).

The notion of exclusive ownership as a property right is fundamental to our theory of social organization. In addition to its central role in protecting the individual's right to be let alone, the importance of exclusive ownership--the ability to exclude freeriders--is now understood as essential to economic development, and to the avoidance of the wasting of resources found under common property systems. See Hardin, The Tragedy of the Commons, 162 Science 1243 (1968); Barzel, Optimal Timing of Inventions, 50 Rev.Econ. & Stat. 348 (1968); Lunn, The Roles of Property Rights and Market Power in Appropriating Innovative Output, 14 J. Legal Stud. 423 (1985).

The intruder who enters clothed in the robes of authority in broad daylight commits no less an invasion of these rights than if he sneaks in in the night wearing a burglar's mask. In some ways, entry by the authorities is more to be feared, since the citizen's right to defend against the intrusion may seem less clear. Courts should leave no doubt as to whose side the law stands upon.

In the case before us, the Order issued by the EPA purported to authorize Government agents, both federal and state, to come on plaintiffs land and to establish a Government presence there. That it was for a beneficent purpose, from the viewpoint of the general public at least, is not at issue; plaintiffs did not contest, nor do we think they could, that the Order qualifies under the "public purpose" language of the Fifth Amendment.

The question addressed by the Claims Court in Hendler I was whether that Order, standing alone, met the tests for a regulatory taking. The court concluded no. On the facts then before the court, and in light of the absence by plaintiffs of proof of facts addressed specifically to the tests for a regulatory taking based on the Order alone, we do not disagree with that ruling.

We note, however, that that ruling says nothing about whether subsequent events, in light of the character of the Government's action and plaintiffs' distinct investment-backed expectations, might have had sufficient economic impact on the plaintiffs to constitute a regulatory taking. Given the fact-specific findings required for determining under current regulatory takings law when such a taking occurs, we understand the trial judge to have refrained from deciding this issue on summary judgment. It remains an issue in the case.

Takings under the Traditional Physical Occupation Theory

The second issue before the trial court was whether the Government's actions, in placing wells on plaintiffs' property and engaging in other activities on the site, was a taking--an inverse condemnation--under traditional physical occupation theory. With regard to the wells, the trial judge felt more evidence was needed to establish "whether the devices are truly permanently affixed to plaintiffs' property." Hendler I at 97. But on the facts before the judge, that conclusion again misperceives the thrust of the protections afforded by the Fifth Amendment.

A physical occupation of private property by the government which is adjudged to be of a permanent nature is a taking, and that is true without regard to whether the action achieves an important public benefit or has only minimal economic impact on the owner. Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 102 S. Ct. 3164, 73 L. Ed. 2d 868 (1982) (placement by authority of the government of cable television (CATV) cable and connection boxes on the roof of an apartment building was a taking under the traditional test). As Justice Marshall said in Loretto: "when the physical intrusion reaches the extreme form of a permanent physical occupation, a taking has occurred. In such a case, 'the character of the government action' not only is an important factor in resolving whether the action works a taking but also is determinative." Id. at 426, 102 S.Ct. at 3171.

In this context, 'permanent' does not mean forever, or anything like it. A taking can be for a limited term--what is 'taken' is, in the language of real property law, an estate for years, that is, a term of finite duration as distinct from the infinite term of an estate in fee simple absolute. (While called an estate for years, the term can be for less than a year. See generally Cribbet, Principles of the Law of Property 54 (3d ed. 1989).)

In United States v. General Motors Corp., 323 U.S. 373, 65 S. Ct. 357, 89 L. Ed. 311 (1945), the government's appropriation of the unexpired term of a warehouse lease was a taking; the fact that it was finite went to the determination of compensation rather than to the question of whether a taking had occurred. Accord, United States v. Petty Motor Co., 327 U.S. 372, 66 S. Ct. 596, 90 L. Ed. 729 (1946) (federal government acquired the remainder of a lease for a building); Kimball Laundry Co. v. United States, 338 U.S. 1, 69 S. Ct. 1434, 93 L. Ed. 1765 (1949) (federal government appropriated private business for public use during World War II; a taking).

There is nothing 'temporary' about the wells the Government installed on plaintiffs' property, in the sense in which we used it in referring to the parked truck of the lunchtime visitor. Years have passed since the Government installed the first wells. The wells are some 100 feet deep, lined with plastic and stainless steel, and surrounded by gravel and cement. Each well was capped with a cement casing lined with reinforcing steel bars, and enclosed by a railing of steel pipe set in cement. These surveillance wells are at least as 'permanent' in this sense as the CATV equipment in Loretto, which comprised only a few cables attached by screws and nails and a box attached by bolts. 458 U.S. at 422, 102 S.Ct. at 3168. Nothing in the Government's activities suggests that the wells were a momentary excursion shortly to be withdrawn, and thus little more than a trespass.11 Nor does the Order or the Government's subsequent actions disclose any indication of a timetable for withdrawal.

Part of the difficulty here is the confusion that arises in the cases and commentaries over the use of the term 'temporary taking.'12 The argument in Agins, which was finally laid to rest in First Lutheran Church, was that a regulatory taking, unlike a physical taking, is by its nature 'temporary.' This is because the government, upon being told the regulation was overly intrusive and therefore a taking (by whatever test), could rescind or amend the regulation.

It is equally true, however, that the government when it has taken property by physical occupation could subsequently decide to return the property to its owner, or otherwise release its interest in the property. Yet no one would argue that that would somehow absolve the government of its liability for a taking during the time the property was denied to the property owner. All takings are 'temporary,' in the sense that the government can always change its mind at a later time, and this is true whether the property interest taken is a possessory estate for years or a fee simple acquired through condemnation, or an easement of use by virtue of a regulation. The long drawn out battle of Agins/San Diego Gas/ First Lutheran Church was not a fight over principle, but a dispute over the illogical use of a word.

If the term 'temporary' has any real world reference in takings jurisprudence, it logically refers to those governmental activities which involve an occupancy that is transient and relatively inconsequential, and thus properly can be viewed as no more than a common law trespass quare clausum fregit. Our truckdriver parking on someone's vacant land to eat lunch is an example.

We do not by that example mean to suggest that that defines the boundaries of the case. We need not decide here what physical occupancy, of what kind, for what duration, constitutes a Loretto taking. It is enough to say that, on the facts before the Claims Court on the motion for summary judgment, we conclude that the occupancy by the Government was comfortably within the degree necessary to make out a taking. When the governmental intrusion is as substantial a physical occupancy of private property as this is, Loretto establishes that there is a taking.

By like token, the concept of permanent physical occupation does not require that in every instance the occupation be exclusive, or continuous and uninterrupted. The evidence before the court was that Government vehicles and equipment entered upon plaintiffs' land from time to time, without permission, for purposes of installing and servicing the various wells. They remained on the land for whatever duration was necessary to conduct their activities, and then left, only to return again when the Government desired.

In Kaiser Aetna v. United States, 444 U.S. 164, 100 S. Ct. 383, 62 L. Ed. 2d 332 (1979), the property owners had formed a marina by dredging a shallow lagoon and a connecting outlet into contiguous navigable waters. An exclusive subdivision community was then built around the marina. The lagoon had been private property before the development, and the property owners continued to deny access to the public after the development.

Subsequently, the government claimed that the property owners were required to open the lagoon to members of the public who might choose to visit by boat, since it was now subject to the "navigational servitude." The Court held that if the Government wished to impose public use--even intermittent public use--of the lagoon upon the property owners, it was required to pay just compensation. The Court explained:

In this case, we hold that the "right to exclude," so universally held to be a fundamental element of the property right, falls within this category of interests that the Government cannot take without compensation.... the imposition of the navigational servitude in this context will result in an actual physical invasion of the privately owned marina.... And even if the Government physically invades only an easement in property, it must nonetheless pay just compensation (citations omitted).

Id. at 179-80, 100 S.Ct. at 393.

The principle of Kaiser Aetna was reinforced in Nollan v. California Coastal Comm'n, 483 U.S. 825, 107 S. Ct. 3141, 97 L. Ed. 2d 677 (1987). Justice Scalia, writing for the Court, said:

"To say that the appropriation of a public easement across a landowner's premises does not constitute the taking of a property interest but rather ... 'a mere restriction on its use,' ... is to use words in a manner that deprives them of all their ordinary meaning.... We have repeatedly held that, as to property reserved by its owner for private use, 'the right to exclude [others is] "one of the most essential sticks in the bundle of rights that are commonly characterized as property." ' [citing Loretto and Kaiser Aetna ] ... We think a 'permanent physical occupation' has occurred ... where individuals are given a permanent and continuous right to pass to and fro, so that the real property may continuously be traversed, even though no particular individual is permitted to station himself permanently upon the premises." Nollan v. California Coastal Comm'n, 483 U.S. at 831-32, 107 S.Ct. at 3145-46.

The evidence before the court in Hendler I reflected a situation in which the Government behaved as if it had acquired an easement not unlike that claimed in Kaiser Aetna. Pursuant to the easement, the Government at its convenience drove equipment upon plaintiffs' land for the purpose of installing and periodically servicing and obtaining information from the various wells it had located there. Kaiser Aetna and Nollan would seem to leave little doubt that such activity, even though temporally intermittent, is not 'temporary.' It is a taking of the plaintiffs' right to exclude, for the duration of the period in which the wells are on the property and subject to the Government's need to service them.13

We emphasize that the issue is not whether the Government had the right to impose itself and its activities on these plaintiffs. Whatever right the plaintiffs had to be let alone was overcome by the Government's need in the interest of public health and safety to monitor the ground water contamination. Indeed, plaintiffs expressly concede that point.

The issue before the court in Hendler I was whether, on the facts before it, the Government took any property by permanent physical occupation, thus obligating it to pay plaintiffs just compensation. The trial judge thought not, absent more facts; we think nothing more needed to be shown. The trial judge denied plaintiffs' motion for summary judgment on this point; he should have granted it.

The Activities Conducted on the Land by the State of California

Finally, the question was raised whether the activities of the State of California were attributable to the Federal Government. This question bears on two important issues. One, whether the Federal Government caused and is liable for the totality of activity. This is important for determining the damages owed by the Government, and also for determining, should there be any doubt, whether there was enough activity by the Government to constitute a taking. Second, it goes to the issue of what it is plaintiffs had to prove--was it enough to show total activity, or did they need to show which of the actors did what.

The trial judge was of the view that it was necessary to establish that a common law agency relationship existed between the Government and the State of California in order for the Government to bear responsibility for the state's activities on the land. He concluded that there was insufficient proof of such a relationship. Hendler I at 99-100.

Common law agency is one test for determining the Government's responsibility, but it is not the only basis for establishing the Government's liability for the State's activities. The activities of EPA and of California were two coordinate and coordinated parts of the same undertaking. It is a basic principle of eminent domain law that government officials--executive branch officials--derive their power to take private property for a public purpose by a grant of authority from the legislature. Once legislative authority is obtained, authorized members of the government may determine how and when the authority will be exercised. See 1A J. Sackman, supra, at § 3.2.

Government officials could not, absent proper legislative authority, lawfully enter plaintiffs' property and begin drilling wells and installing equipment. Likewise, any administrative order issued to allow such an entrance would have to be based on proper legislative authority. In this case, such authority flows from CERCLA, and it was under the authority of that federal statute that the EPA Order was issued.

The Claims Court found that the EPA Order authorized access to plaintiffs property for "EPA officials and other authorized personnel, including state officials" and that such authority was based on 42 U.S.C. §§ 9606(a), 9604(a), (b), and (e). Hendler I at 93. Thus, California state officials who entered onto plaintiffs' land did so under the authority granted by CERCLA.

It follows that their activities within the scope of the Order are attributable to the Federal Government for purposes of takings law just as are the activities of EPA itself. As the Court in Loretto observed, "[a] permanent physical occupation authorized by [a government] is a taking without regard to whether the [government], or instead a party authorized by the [government], is the occupant." 458 U.S. at 432 n. 9, 102 S. Ct. at 3174 n. 9. In Loretto the authorizing government was a state, rather than the federal government, but the principle remains the same.

There is no question but that California officials were acting under the authority of the 1983 Order, and in pursuance of the State's formal Cooperative Agreement with the Government to assist in carrying out Superfund activities, including this one, and with substantial funding from the Government. That the State had authority to act on its own initiative, as the Government contends, is immaterial. It is no defense to a charge of authorizing someone to violate another's rights that the perpetrator might have done so on his own.

The Discovery Requests and the Ensuing Dismissal

We turn now, in light of the law applicable to the issues just discussed, to the question of whether plaintiffs should be sanctioned for their conduct of the case, and whether dismissal with prejudice is a proper sanction under the circumstances. A plaintiff with a proven meritorious claim may have that claim taken away if the plaintiff's conduct--or more accurately the conduct of the lawyer for the plaintiff--is so offensive to the court and so violative of established norms as to justify this severe a punishment. See Nat'l Hockey League v. Metro. Hockey Club, 427 U.S. 639, 96 S. Ct. 2778, 49 L. Ed. 2d 747 (1976); see also Rule 37.

The controversy over the adequacy of plaintiffs' responses to the Government's propounded interrogatories turned on whether the answers provided "the essential factual bases for plaintiffs' claims, information necessary for defendant to prepare a defense and to pursue meaningful discovery." Claims Court Order, April 30, 1987. Following the extended discovery period, during which the parties fought over whether various interrogatories were proper and whether answers should be compelled, the Government moved the court to dismiss Hendler's complaint under Rule 37. The Claims Court's Order of April 30, 1987, stated that plaintiffs' supplemental responses had failed to respond fully to the Government's interrogatories, and gave the plaintiffs fourteen days to comply. The court warned that failure to comply would result in dismissal of the case. Hendler timely responded, but the Government deemed the response inadequate. In its December 11, 1989 decision, Hendler II, the court granted the Government's motion and dismissed the case with prejudice.

The court deemed the plaintiffs to have acted willfully in not responding fully and precisely to the Government's interrogatories. "All of the information the government seeks is within the exclusive knowledge and possession of plaintiffs. Only plaintiffs know how the EPA activities on their land have deprived them of any economically viable use of the subject property, yet they refuse to tell defendant and the court." Hendler II at 31 (footnote omitted). The court noted that Hendler's defense to the Government's motion to dismiss is that "they have 'told all they know.' " Hendler II at 31 n. 9. The court concluded that the Government was clearly prejudiced in its defense of the suit: "Defendant has no means of anticipating how plaintiffs will attempt to prove that there is no economically viable use of their property. At some point in the proceeding, plaintiffs must marshal facts which prove how the EPA activities have deprived them of an economically viable use of their property." Hendler II at 32 (emphasis in original).

As a general rule, trial courts are given wide discretion to manage the course of a trial, and to direct the conduct of counsel. But this discretion is not without limits. We review the trial court's dismissal decision for an abuse of discretion. Nat'l Hockey League v. Metro. Hockey Club, Inc., 427 U.S. 639, 642, 96 S. Ct. 2778, 2780, 49 L. Ed. 2d 747 (1976); Adkins v. United States, 816 F.2d 1580, 1581-82 (Fed.Cir.1987). An abuse of discretion may be found when (1) the court's decision is clearly unreasonable, arbitrary, or fanciful; (2) the decision is based on an erroneous conclusion of the law; (3) the court's findings are clearly erroneous; or (4) the record contains no evidence upon which the court rationally could have based its decision. Western Elec. Co. Inc. v. Piezo Technology, Inc., 860 F.2d 428, 429-30 (Fed.Cir.1988).

The interrogatories at issue are nine in number. They break down into two general categories; (1) interrogatories concerning the Government's own activities, and (2) interrogatories seeking to establish economic impact.

a. Interrogatories concerning the Government's own activities

Interrogatories 8, 13, 14, and 16-18 deal with the particulars of the Government's activities on the plaintiffs' properties. The Government's argument is that Hendler has simply chosen to be uncooperative in responding to its interrogatories and supplying the evidence to enable the court to determine if a taking did occur, and if so, what would be just compensation. Interrogatories 13 and 17 are illustrative.

What is the factual basis for plaintiffs' contention that EPA has taken plaintiffs' property by driving heavy machinery, trucks, and other vehicles onto the Subject Property?

What is the factual basis for plaintiffs' contention that EPA has taken plaintiffs' property by detonating explosives on the Subject Property?

Appendix at 76, 87.

If the questions are intended to ask, when and how often did the Government drive heavy machinery and detonate explosives on plaintiffs' property, it is hard to know what more plaintiffs should have said. Plaintiffs' final answer to interrogatory 13, for example, listed the exact parcels on which the driving occurred, listed driving-related activities such as "monitoring wells," and stated that they did not know for certain which driving was EPA's and which was California's, but that discovery was continuing. Plaintiffs then added that "Property Owners have scheduled the depositions of EPA employees to learn about EPA activities on the Subject Property."

Neither party has suggested that some private individual surreptitiously entered onto plaintiffs' land to install the wells. Nor is there any suggestion that plaintiffs themselves installed the wells. Thus, the wells must have been installed pursuant to the EPA's order. The EPA, one must assume, is aware of what wells and support equipment were installed, the methods by which the excavation occurred, and the extent of ongoing intrusion on plaintiffs' property to monitor and maintain the wells.

In response to the Government's interrogatories, plaintiffs contacted State of California officials, and scheduled depositions of EPA employees. For the Government to make requests which would require plaintiffs in turn to seek information from the Government itself, and then to seek dismissal with prejudice when plaintiffs failed to supply the Government with the information that the Government already had, seems a cruel joke.

Whether funny or not, the Government's requests clearly exceed the bounds for which discovery is provided to litigants. See Reichert v. United States, 51 F.R.D. 500, 503 (N.D.Cal.1970); see generally 8 C. Wright & A. Miller, Federal Practice and Procedure § 2174 (1970 & Supp.1991). Plaintiffs' complaint, amended twice, put the Government fully on notice of the actions about which plaintiffs complained. C.f. Ingalls Shipbuilding, Inc. v. United States, 857 F.2d 1448, 1452 (Fed.Cir.1988). To use discovery as an alternative to its own preparation of a defense or to harass comes close to governmental abuse of the judicial process. See Olmert v. Nelson, 60 F.R.D. 369, 370 (D.D.C.1973); Kainz v. Anheuser-Busch, Inc., 15 F.R.D. 242, 251 (N.D.Ill.1954); see generally 8 C. Wright & A. Miller, supra, § 2174. There is little question that a dismissal under Rule 37 based on these interrogatories was an abuse of discretion.

b. Interrogatories seeking to establish economic impact

The Government repeatedly pressed for more specific answers to the interrogatories propounded regarding economic impact. A look at two of these interrogatories suggests the difficulty plaintiffs faced in answering them.

What is the factual basis for plaintiffs' contention that EPA has interfered with plaintiffs' legal ability to sell the Subject Property?

What is the factual basis for the plaintiffs' contention that EPA has interfered with all intended development of the Subject Property? What specific plans had been made for such development? When did plaintiffs intend to commence such commercial and residential development? ... What specific barriers has EPA created that prevented [sic] the development of the land at this time?

Appendix at 93-95. Plaintiffs responded to I 26, in part, that the combination of EPA's physical occupation and the Order made "sale of the Subject Property to a knowing buyer impossible, as a knowing buyer would not know what use could be made of the property." Id. Plaintiffs responded to I 27, in part, by stating that they were not developers, that they did not intend to develop the property and that they simply planned to sell or lease the property. Id.

On appeal, the Government sought to support the Claims Court's decision to sanction plaintiffs by arguing that the facts requested in the interrogatories were essential to the plaintiffs' claims, and unless substantiated, the court could not conclude that they constituted a taking. Quoting from the court's opinion: "If the evidence shows the absence of a permanent taking, to decide whether a temporary taking has occurred the court will need to know the character and expected duration of the government activity, and the degree of interference with the private property," Hendler I at 97. The Government argued that "[o]n the record before it, the court found 'no hint of how defendant's activities may have affected [plaintiffs'] investment backed expectations, or even if they had any.' " See Hendler I at 97.

The difficulty with this argument is that it assumes that to establish a taking of property under the Fifth Amendment, it is necessary that the owner of the locus in quo establish investment backed expectations, and that anything less than a taking of all economically viable use of the land is 'temporary' and not entitled to just compensation. While those are the buzz words dominating the current literature of takings law, they relate to the jurisprudence of regulatory takings, and have no relevance when the cause of action is premised on an actual physical occupancy. As Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 437, 102 S. Ct. 3164, 3177, 73 L. Ed. 2d 868 (1982), demonstrates, the extent of occupation was only relevant to compensation, not liability.

We have before us, then, a dismissal sanction based at least in part on an alleged recalcitrance which goes at most to proving the quantum of damages resulting from the physical taking. Plaintiffs had satisfactorily met their burden of proof on the liability issue--that there indeed had been a permanent physical occupation by the Government. If plaintiffs failed to aggressively develop a strong factual basis for calculation of just compensation, that was their right--they are free to choose a course of action which may turn out not to maximize their recovery. Since the plaintiffs repeatedly indicated their willingness to be limited to their answers, the Government certainly cannot be said to be prejudiced by such a litigation tactic.

A Rule 37 dismissal is a harsh remedy, which should be reserved for only the most severe abuses of the discovery process. Societe Int'l v. Rogers, 357 U.S. 197, 212, 78 S. Ct. 1087, 1095, 2 L. Ed. 2d 1255 (dismissal appropriate only for cases of willfulness, bad faith, or fault of the petitioner); Nat'l Hockey League v. Metro. Hockey Club, 427 U.S. at 639, 96 S.Ct. at 2778 (dismissal appropriate where respondents demonstrated "flagrant bad faith" and "callous disregard"); Adkins v. United States, 816 F.2d 1580 (Fed.Cir.1987) (dismissal appropriate for complete failure to respond). As the Claims Court itself noted, "the draconian sanction of dismissal is to be used only as a last resort, in extreme circumstances." Hendler II at 30.

Flaunting an order of a trial judge cannot be condoned, and the judge has wide discretion in imposing sanctions. However, dismissal is authorized only when the failure to comply with the court order is due to willfulness or bad faith and not from the inability to comply with the order. Nat'l Hockey League, 427 U.S. at 640, 96 S.Ct. at 2779. We recognize that the trial court made a finding of willfulness "based upon its review of the record, its personal impressions of the attorneys, and the history of this case[.]," Hendler II at 31, but this conclusion is based on an erroneous over-emphasis on the importance of economic valuation in takings jurisprudence, and little else. We do not condone the plaintiffs' unseemly delay in responding to the January 1987 order, but the dismissal was not based on their failure to submit a timely response, rather it was based on the alleged inadequacy of the response.

Had the case not gone off on the wrong footing the controversy over the discovery orders would never have arisen. This was not entirely the judge's doing. The Government at trial and in its brief and in argument before this court showed a remarkable inability to comprehend the fundamental difference between a takings claim based on a regulatory taking theory and one based on physical occupation. For example, the Government's brief on appeal: "Both counsel for the United States and the Claims Court believed that the facts requested were essential to the plaintiffs' claims, and therefore to the government's defense." Appellee's Brief at 17. But given the factual stipulation in the record before the court establishing the nature and extent of the Government's physical activities on the land, at best an arguable defense might have been that these years of physical invasion were somehow 'temporary.' None of the disputed interrogatories spoke to that question, even assuming it was legitimately before the court.

In Ingalls Shipbuilding, Inc. v. U.S., 857 F.2d 1448 (Fed.Cir.1988), this court reversed a similar Rule 37 dismissal. The government had provided only the most generalized responses to a contractor's interrogatories seeking the particulars of a fraud claim. The court supported its holding in part with the trial court's failure to give plaintiffs ample notice of its intent to dismiss the case, whereas here the plaintiffs had a clear warning of the trial judge's intent. Nonetheless, what was said there holds true here:

As we see it then, the court meted out a sanction tantamount to dismissal not because of any willful failure to answer Litton's interrogatories, but because of the government's continued pursuit of a case the court felt had no merit. Discovery sanctions, however, are meant to deter intentional abuse of the discovery process, not as a means to resolve the merits of a case. Where the problem is a perceived inadequacy of proof, not the failure to provide it, dismissal as a sanction under Rule 37 is not warranted (citations omitted).

Ingalls Shipbuilding Inc. v. United States, 857 F.2d at 1451.

As we stated above in setting out our standard of review, we may reverse a trial court's Rule 37 dismissal as an abuse of discretion when that dismissal is based, among other things, upon an erroneous conclusion of law. Our examination of the record before us leaves us with a firm conviction that this has occurred in at least three ways.14

First, the original trial judge's refusal to grant the plaintiffs' summary judgment motion on the issue of a permanent physical occupation slanted the posture of the case to issues relevant only to regulatory takings theory, with its attendant balancing and emphasis on economic impact.

Second, the successor trial judge's view of the prejudicial impact of interrogatories 24, 26, and 27 was skewed by the first trial judge's conclusion that the activities of the State of California were irrelevant to the case. To the extent that the issues remaining were focused on regulatory takings theory, the second trial court could not have properly weighed this question when the bulk of the activities were carried out by California. Also, plaintiffs' inability to distinguish in its responses to the interrogatories between the activities of EPA and of California, see e.g. I 13, supra page 28, no doubt weighed heavily in the second judge's impression of prejudice to the Government. As a matter of law, this inability in plaintiffs' responses is no defect at all; State and Federal activities were fungible goods in a confused mass.

Thirdly was the error with regard to interrogatories 8 and 11-17, discussed above. These interrogatories sought detailed information already in EPA hands. The Government was on notice that all instances of physical invasion provided factual grounding for plaintiffs' takings claim. Furthermore, we note that the property in question was unoccupied and plaintiffs lived some distance away. The Government will not be heard to complain that plaintiffs failed to monitor their property around the clock to discover each and every instance of intrusive Government activity, and then failed in some duty to report these intrusions accurately to the Government during discovery.

The cumulative effect of these errors compel us to conclude that plaintiffs' suit should not have been dismissed on the grounds given; to do so was an abuse of discretion.

Nothing we have said will preclude the Government from presenting evidence during the damages phase of this case on the intended duration of its occupancy. The trial court will of course consider all relevant evidence on that question, including events since the original hearing. For example, plaintiffs have alleged that a joint DHS/EPA cleanup plan was issued June 30, 1988. Plaintiffs further alleged that the plan proposes to install extraction wells which will operate for at least ten years, and that at least one such well has already been installed. Appellants' Opening Brief at 9 n. 3. Plaintiffs accordingly in September 1988 moved to suspend proceedings and reopen discovery. The second trial judge deemed this motion moot in light of his dismissal. Hendler II at 29.

In view of our decision today, that issue is now properly before the trial judge. Even if the trial judge chooses to hold plaintiffs to the scope of their interrogatory answers on the specific issues addressed, newly available information relevant to what is just compensation must be considered. That evidence will go to the question of whether the taking constitutes the acquisition of an interest in fee, or something less such as a term of years, or some combination thereof.

And of course, once a taking is adjudged, plaintiffs will have the opportunity to establish their severance damages, the damages accruing to their retained land as a result of the taking.15 Since this case involves claims of both a taking by physical occupation and a regulatory taking, we leave to the trial judge on remand the question of whether plaintiffs' total damages for the physical taking--including damages for both the property taken and the injury to retained land--constitute a complete remedy, thus obviating the need for determining whether a regulatory taking also may have occurred.

The dismissal sanction against plaintiffs is reversed. The grant of partial summary judgment in favor of the Government on the question of California's role is also reversed. The denial of plaintiffs' motion for summary judgment on the question of a taking by physical occupancy is reversed; summary judgment for plaintiffs on that motion is ordered. The case is remanded to the Claims Court for further proceedings consistent with this opinion.

Costs are awarded to appellants Hendler et al.

AFFIRMED-IN-PART, REVERSED-IN-PART, AND REMANDED.