Appellants in this case are two members of the "inactive reserve" of the United States Army and Air Force who voluntarily joined the Armed Forces, but who were permitted to complete their educations before beginning tours of active duty. Prior to entering active duty appellants allegedly became conscientious objectors and under the appropriate military regulations sought release from their military obligations. The Army and Air Force denied appellants' requests for discharge, whereupon appellants petitioned the U.S. District Court for the District of Columbia for a writ of habeas corpus to obtain their releases from the military. The District Judge issued preliminary injunctions to prevent appellants from being required to report for active duty; this injunction has remained in effect pending disposition of this appeal and appellants have never reported for active duty. After hearing argument the trial court dismissed the petitions on the ground of lack of jurisdiction, since appellants' respective "immediate custodians" were not located in the District of Columbia.

We agree that the District Court lacked jurisdiction to grant appellants' petitions. Our conclusion is not, however, based upon a finding that appellants' "immediate custodians" are not present in the District of Columbia. Indeed, we find that the concept of "immediate custodianship" is largely irrelevant to determining which particular federal court may entertain habeas petitions brought by inactive reservists.

Eisel was commissioned as an officer in the United States Army Reserve in June 1964. He was allowed to continue his education and was never attached to an active unit of the military. For bookkeeping purposes Eisel's records were kept at a records center in Indiana, although he was never stationed there or physically present in Indiana. During this period of inactive status Eisel lived several places in the United States and for a time in New Zealand.

While living in Cambridge, Massachusetts, and while still on inactive status, Eisel applied for discharge from the Army as a conscientious objector. Eisel was interviewed in Massachusetts by Army representatives to determine if he should be released, and most of his negotiations with the military took place while he was residing there. At some point after these initial contacts Eisel moved to New York, where he currently resides.

While Eisel's application was still pending he was ordered to report for active duty in Virginia. This order was enjoined by a District Court in Massachusetts and subsequently by the U.S. District Court here. (The Massachusetts action was dismissed on petitioner's motion.) As a result Eisel has never reported to his station in Virginia, and has never physically been in that state.

Appellant Gelber was commissioned in the United States Air Force Reserve in October 1966 and like Eisel has been on inactive status since then. Gelber's records were stored in Denver, Colorado, although he has never been physically located in Denver.

Gelber lived for most of his period as an inactive reservist in Cambridge, Massachusetts, and it was there that he petitioned for release from the military as a conscientious objector. Gelber was interviewed by the Air Force in Massachusetts, although in the course of seeking his release he or his attorney corresponded with Air Force personnel in such diverse places as Colorado, Virginia, Washington, D.C., and Texas. During this entire period and up to the present Gelber has lived in Cambridge, Massachusetts.

Upon being ordered to report for active duty in Alaska (with a stop-off in Texas), Gelber sought and obtained in the District of Columbia a preliminary injunction against the order. Thus, Gelber has never physically been present in either Colorado, Texas, Alaska, or the District of Columbia.

Federal habeas corpus law provides that a habeas action "shall be directed to the person having custody of the person detained."1 Appellants Eisel and Gelber allege that the person "having custody" of them and causing them to be "detained" are the respective Secretaries of the Army and Air Force. Since these "custodians" are officially located in the District of Columbia, appellants argue that Washington is the proper place to seek habeas relief.

The Government vigorously contests the theory that the civilian Secretaries of the Army and Air Force are in any relevant sense the "custodians" of appellants. It contends that only the "immediate custodian" of a petitioner is amenable to a writ of habeas corpus. The Government argues that Gelber's "immediate custodian" is the commander of the record center in Colorado where his records were located prior to being called to active duty. Ironically, with regard to Eisel, the Government argues that the "immediate custodian" is not the commander of the record center in Indiana but rather the commander of the installation in Virginia where Eisel has been ordered to report. The Government offered no real explanation of why the commander of the record center would be the "immediate custodian" of one of the appellants but not of the other.

The District Court declined to hold that the civilian Secretaries of the Armed Services were the "custodians" of Gelber and Eisel. In so doing, the trial judge accepted the Government's contention that only the "immediate custodian" was a proper party to a habeas action and that the action could only properly be brought in the jurisdiction where this "immediate custodian" is physically located. Sensing an inherent contradiction in the Government's contention that the commander of the records center was the "immediate custodian" of one of the petitioners but not the other, the District Judge attempted to be consistent; he held that the "immedite custodians" of appellants Eisel and Gelber were the respective commanders of the records centers where their records were located prior to their being called to active duty.

We sympathize with the trial court's attempts to reach the truth through the morass of conceptual contradictions and non sequiturs that abound in this area of the law. We likewise believe that his determination that this action should not be maintained in the District of Columbia was correct. We would prefer not to base our decision, however, upon a determination as to whether the service Secretaries are the "immediate custodians" of the appellants. Where inactive reservists may or may not bring habeas actions is better determined by analyzing the policies for and against allowing an action in a particular jurisdiction, rather than by the blind incantation of words with implied magical properties, such as "immediate custodian."

In determining the most feasible jurisdiction in which to permit habeas actions by inactive reservists, there are at least six distinct interests which should be considered.2 First, the forum should be close to the records of the case and to the witnesses who may have to appear. Second, the choice should promote a fair distribution of habeas cases among the District Courts rather than concentrating the burden on a few. Third, the forum should be one before which it is convenient for the petitioner to appear. Fourth, it should be a forum convenient for the Government. Fifth, and perhaps most important, the forum should be one that is easily determined; the rule for determining its location should be clear and unquestionable. Sixth, to the extent possible, there should be a single, exclusive forum in order to prevent forum shopping among alternatives. It is easy to see that these interests may at times be conflicting and that a rule fulfilling all of them perfectly would be difficult if not impossible to formulate. We believe, however, that the best solution will result from a careful evaluation and weighing of these competing considerations.

Our method of analysis is reinforced by the very recent decision of the Supreme Court in Braden v. 30th Judicial Circuit Court of Kentucky (1973).3 There the Court held that the petitioner, imprisoned in Alabama but subject to a detainer issued by a Kentucky court for an offense charged there, was "in custody" within the meaning of 28 U.S.C. Sec. 2241(c)(3) for purposes of habeas corpus to require a speedy trial in Kentucky. While emphasizing the importance of the location of the custodian in this criminal case, the Court employed practical considerations in determining where an action would lie:

In terms of traditional venue considerations, the District Court in the Western District of Kentucky is almost surely the most desirable forum for the adjudication of the claim. (Footnote omitted.) It is in Kentucky, where all of the material events took place, that the records and witnesses pertinent to petitioner's claim are likely to be found. And that forum is presumably no less convenient for the respondent, the Commonwealth of Kentucky, than for the petitioner. The expense and risk of transporting the petitioner to the Western District of Kentucky, should his presence at a hearing prove necessary, would in all likelihood be outweighed by the difficulties of transporting records and witnesses from Kentucky to the district where petitioner is confined. (Footnote omitted.) Indeed, respondent makes clear that "on balance, it would appear simpler and less expensive for the State of Kentucky to litigate such questions [as those involved in this case] in one of its own Federal judicial districts."4

A. Washington, D.C.

Adopting this functional approach to the problem compels the conclusion that appellants should not be permitted to seek habeas relief in the District of Columbia. Eisel and Gelber are not and never have been residents of the District of Columbia. Their only significant contacts with the military have occurred elsewhere. While there are numerous representatives of the military in the District of Columbia, there is no reason to believe that the military operations in Washington are of any more relevance to appellants than to any other inactive reservist or for that matter to any member of the military, active or inactive.

In short, appellants' relationship to the military located in the District of Columbia, including the various civilian Secretaries, is no different from that of any other member of the military. Thus, if we were to permit Gelber and Eisel to proceed in the District, we would be compelled to permit any inactive reservist, and perhaps any member of the military,5 the same privilege. It is unclear how many reservists are in situations similar to that of appellants or how many of them might be expected to seek release from service by petition of habeas corpus. Undoubtedly, the number of both groups could be expected to fluctuate, with relatively large numbers seeking release during the time bullets are flying and few seeking release when playing soldier looks more fun.

Whatever may be the actual number, it is probable that under some conditions the number could be quite large. Such an influx could seriously overburden an already busy court system. This problem would be compounded by the substantial fluctuations that could be expected between times of peace and war, since it would be difficult for the Congress to ameliorate the burden by creating more judgeships for an influx that could well be short-lived.

An even more compelling reason not to permit this suit is that it may often be inconvenient for habeas petitioners to retain counsel and come to Washington in order to seek release from the military. While it may be most expedient for Eisel and Gelber to bring their action in the District of Columbia, this, of course, would not be universally true. The burden of seeking relief in Washington, D.C., could, for example, be overwhelming for an inactive reservist located in Montana or Alaska.

Appellants have suggested that this problem could be avoided by making the District of Columbia an alternative location for habeas petitions by inactive reservists. Although it was not clear where else appellants felt that an action might be brought, the implication was that in actions such as this a petitioner would have a choice of seeking a writ in either Washington or some other presumably more convenient location. Indeed, under a recent Supreme Court decision it seems clear that a certain class of inactive reservists-a class which appears to include appellant Gelber-has already been granted the right to bring an action elsewhere than in the District of Columbia.6 Necessarily then, in at least some situations, Washington would be merely an alternative to bringing an action in another jurisdiction. While this proposal may solve the problem of inconvenience, it creates another which is at least as troublesome, since alternative jurisdictions would almost certainly lead to forum shopping by reservists.

It is also unlikely that Washington would necessarily be close to the records and witnesses in a typical inactive reservist situation. Reservists seeking discharge as conscientious objectors are always interviewed by the military personnel in the petitioner's area of residence; the testimony of these interviewers could be critical in some cases. Likewise, the testimony of ministers, employers, teachers, and other acquaintances could play an important role in a proceeding. Locating jurisdiction for all reservists in any one place, including Washington, would necessarily inconvenience witnesses located elsewhere.

For these reasons we find that Eisel and Gelber should not be permitted to maintain these actions in the District of Columbia.

While locating the forum in Washington has distinct disadvantages, establishing jurisdiction at the various centers that store the records of inactive reservists fails on almost every count. Apparently, it is the practice of the various branches of the military to store the records of inactive reservists in a single location. Thus if jursdiction were located in the record center forum, all reservists' cases would be concentrated in a few locations to almost as great an extent as would be the case if Washington were found to be the proper forum. The problem of concentration would be compounded since there are substantially fewer federal District Judges in Indiana and Colorado than there are in the District of Columbia; thus, the burden would be even greater on the judicial system in those places. These isolated record centers will often be remote and inconvenient for both the petitioner and witnesses. Although the Government has at times advocated the record centers as the proper location, it is difficult to imagine that locating all reservists' actions in such places as Denver, Colorado, and Fort Benjamin Harrison, Indiana, would be significantly more convenient for the Government than permitting the action elsewhere. Conceivably, it could lead to even greater inconvenience, since the Government undoubtedly has limited resources in these areas which could be overburdened by the influx of habeas cases.

If it were determined that the record centers were a proper location, it would necessarily be merely an alternative forum. The Supreme Court decision mentioned in discussing the Washington alternative did not, whatever else it might have done, establish that cases such as this must be brought in the location of the record centers.7 Thus, permitting the record center alternative would inevitably lead in some circumstances to the possibility of forum shopping.

Of all the interests that should be considered in determining the proper jurisdiction, the only one that the record center alternative even arguably satisfies is that it is relatively simple to determine. This is, however, a poor dividend indeed for the cost of forfeiting every other interest which a rule should seek to satisfy.

The third possible alternative suggested by the parties, permitting an action in the district to which the reservist has been ordered to report for active duty, is equally unacceptable. As with the other alternatives, the place of reporting would in many cases be inconvenient for witnesses and might be far from the records. It is also clear that such a rule would establish multiple jurisdictions in which actions might be maintained since, as has been noted, the Supreme Court has established at least a limited right to maintain actions in a petitioner's place of domicile.8 Forum shopping is thus a problem with this alternative.

By far the greatest reason not to require actions in the jurisdiction where a reservist is ordered to report is the potential inconvenience that such a rule would impose on the reservist. With this rule the military could determine the forum virtually at will by simply ordering the reservist to a particular location. The services could, for example, select the jurisdiction for cases by issuing active duty orders to a reservist for a designated post, which might be a remote outpost far from the petitioner's home. Such a rule would also permit the military to select a jurisdiction thought to be favorable to the Government by designating it as the place to report. This rule would be both inconvenient and unfair.

For these reasons we cannot accept the place of duty alternative.

This court is satisfied that it would be improper to permit appellants to seek habeas relief in the District of Columbia. We are aware, however, that our comments are of little solace to Gelber and Eisel who must now set off in search of the proper forum. This unfortunate consequence may be compounded by the possibility that upon instituting an action in another jurisdiction the Government may deny that the new location is the proper forum. It has been suggested that in future proceedings, by taking contradictory positions on which jurisdiction is proper, the Government might engage in a form of inverse forum shopping, i. e., whatever court is hearing the case will always be argued to be the wrong jurisdiction. Such tactics, if successful, could prevent the case from ever being heard. While we certainly do not ascribe such unworthy motives to the Government, we are concerned by the apparent contradictions in the Government's position in this case and the unwillingness of counsel on appeal to articulate a clear-cut rule for determining just where inactive reservists might bring habeas actions. To help prevent the Kafkaesque specter of supplicants wandering endlessly from one jurisdiction to another in search of a proper forum, only to find that it lies elsewhere,9 we think it proper to carry our reasoning to its ultimate conclusion and determine where we believe the proper forum to be.

As seen above, under a functional analysis approach none of the alternatives advanced by the parties in this case is acceptable. The true answer lies in an entirely different direction. There are, however, numerous guideposts in existing law that are of help in determining the answer.

Appellants' action here is based upon a remedy granted by a federal statute.10 Although the statutory scheme designates particular forums as proper in some situations,11 it is silent regarding the appropriate jurisdiction for this type of action. Further, while the statute does provide that the action shall be against the "person having custody of the person detained,"12 it does not define "custody" or specify who the person having "custody" will be. Nowhere does the statutes speak of an "immediate custodian" or intimate that an action must necessarily be instituted in the location of such an "immediate custodian," even if it were possible to grant substance to the vague concept of "immediate custodianship." Confronted with this vacuum the courts have attempted to formulate rules governing where habeas actions might be maintained.

Until relatively recently, only persons who were actually physically restrained could seek habeas relief. The concept of "custody" was narrowly limited and only those who were literally in someone's physical custody could obtain a writ.13 Since habeas petitioners were actually imprisoned or otherwise prevented from moving about, if one had territorial jurisdiction over the petitioner one necessarily had territorial jurisdiction over the person who was actually holding him captive-his "immediate custodian," if you will. This was true because if the petitioner was physically restrained, there must be someone in the same location who was doing the actual restraining. In such situations the issue of whether a court had jurisdiction over a custodian was relatively unimportant; if one had proper jurisdiction over the petitioner, one necessarily had jurisdiction over the custodian.14

In the leading Supreme Court decision of Ahrens v. Clark,15 petitioners confined in New York brought suit in the District of Columbia against the Attorney General of the United States to obtain their release. The Court decided that Washington was not the proper forum because it did not have territorial jurisdiction over the petitioners, although the Court has recently explained that the lack of territorial jurisdiction in Ahrens was important because of the risk and expense attendant upon the production of 120 prisoners from New York.16 There was nothing in the opinion that would intimate that the Court felt that the action would fail because the Attorney General was not the "immediate custodian" of those being held "in custody."17 Indeed, there is language that intimates that the Court felt that the issue of who had "immediate custody" was largely irrelevant to where the action should be maintained.18

Thus according to the Supreme Court, it was the location of the petitioner, not the location of the "immediate custodian," that was determinative of the outcome. In cases of physical custody, however, the locations of petitioner and "immediate custodian" are necessarily the same. Thus it is possible that one could sufficiently blur the distinction and argue that the location of the "immediate custodian" rather than the petitioner was outcome determinative. This misinterpretation is aided by the existence of a number of cases decided prior to Ahrens that held, in effect, that habeas actions would only lie in the jurisdiction of the "immediate custodian."19

Within the last decade habeas jurisdiction has expanded greatly and the remedy has been applied to situations in which the petitioner has not been physically restrained in the traditional sense. In the landmark case of Jones v. Cunningham,20 the Supreme Court retreated from its prior requirement of strict custody and recognized that legal as well as physical restraints could be properly regarded as a "custody." In holding that a prisoner released on parole was sufficiently in "custody" to seek habeas relief, the Jones Court listed a number of factors which it felt entitled the parolee to maintain the action. Jones was subject to four types of restraints: (1) he was restricted in his physical movement, unable to leave his community or change his residence without special permission; (2) he was subject to other regulations which did not specifically restrict his movement but had that practical effect, including the requirements that he obtain special permission before operating an automobile and that he report regularly to his parole officer; (3) he was threatened with reincarceration for violations of his parole; and (4) he could be ordered back to prison without a court hearing.

It is unclear from the Court's opinion which of these factors it felt were determinative. The Court appeared to rely heavily, however, on the limitations placed upon the petitioner's physical freedom:

History, usage, and precedent can leave no doubt that, besides physical imprisonment, there are other restraints on a man's liberty, restraints not shared by the public generally, which have been thought sufficient in the English-speaking world to support the issuance of habeas corpus. . . . Such restraints are enough to invoke the help of the Great Writ. Of course, that writ always could and still can reach behind prison walls and iron bars. But it can do more. It is not now and never has been a static, narrow formalistic remedy; its scope has grown to achieve its grand purpose-the protection of individuals against erosion of their right to be free from wrongful restraints upon their liberty.21

Since physical restraint was a part of the test, the nature of the "custody" concept was not altered. Physical restraints similar to those in Jones have generally existed in the areas to which Jones has been extended: probation;22 escape from detention;23 conditional suspension of a sentence;24 and, in some cases, bail.25

By like token, the expansion of the habeas remedy under Jones and its progeny did not alter the location where such actions would lie. These cases all involved criminal detentions in which the proper forum is designated by statute.26 Since the nature of the evil (i. e., the "custody") that was sought to be remedied had not been altered, there was no basis for the courts to depart from the established statutory and common law.

The inactive reservist presents a situation in which the concept of "custody" has been stretched to the extreme. The petitioner is as mobile as any other member of our society and conceivably can be in "custody" anyplace he happens to be at the time. Indeed, it was not until the Supreme Court's decision in Strait v. Laird,27 handed down May 1972, that it was established conclusively that an inactive reservist was actually in "custody" and therefore able to seek habeas relief.

Holding that inactive reservists are "in custody" creates several conceptual and practical problems. The conceptual problems are discussed completely in Justice Rehnquist's dissent in Strait.28 Since Strait is now clearly the law, these conceptual problems are not of primary concern for this court. The fundamental practical problem of where an inactive reservist may bring suit, however, is.

In this area of administrative as opposed to criminal custody there are no statutory guidelines. As we noted above, the place where the petitioner was in "custody" has been the touchstone for this determination.29 Since the inactive reservist is not "in custody" in any paricular place this test is no longer particularly helpful.

In casting about for another test, it is not surprising that one might fix upon the other factor that has occasionally been mentioned in previous actions30-the location of the "custodian." While it may be true that the petitioner is not "in custody" in any particular place, it is superficially reasonable to conclude that if the inactive reservist is in "custody" at all, as the Supreme Court has said that he is, then he must have a "custodian." The reasoning continues that if one could but find this "custodian" all questions would be solved. We have seen the problems that this sort of reasoning produces.31 It is not surprising that one who is "in custody" only in the most metaphorical sense would also have a "custodian" that partakes of the chimerical.

This search for the proper "custodian" is not merely unproductive, it is not required by law. There has never existed in this Circuit an absolute requirement that a habeas action be brought in the location of the "immediate custodian." For example, servicemen stationed abroad have been freely granted the right to bring habeas actions in the District of Columbia even though their "immediate custodians" were not in any sense located here.32 In such situations the "immediate custodian"-presumably the petitioner's commanding officer-was not within the territorial jurisdiction of any federal District Court. Thus, to the extent that the concept of "immediate custodian" has ever had any viability, it has readily given way to the practical impossibility of obtaining such jurisdiction.

B. The Nature of Custody for the Inactive Reservist

The answer to the issue of the proper forum must begin with an understanding of the significance of the altered nature of "custody" in the context of the inactive reservist. Unlike an incarcerated prisoner, an inactive reservist is in "custody" only in a highly metaphysical sense. The chains imposed by the Government bind only his sense of wellbeing and if he is a prisoner at all his confinement emanates from his own mind.

The altered nature of "custody" necessitates a reconsideration of the nature of "custodianship." In situations of physical custody it is both natural and helpful to think of an actual person charged with the physical restraint of the prisoner. Where the petitioner is at large but the restraints deny him freedom of movement it will still be proper to think of his "custodian" as being the person who is causing the restraint.

In the case of the inactive reservist, however, this type of thinking is entirely inappropriate. The instant case illustrates the point. Here there is no showing that any of the suggested "immediate custodians" have exercised any real control over the petitioners. The commanders of the record centers and the bases to which petitioners have been ordered to report have never had anything to do with the denial of the requests for release. The record centers' commanders have been mere shadows through which others worked and the relationship of the base commanders is at best potential. By like token, the civilian service Secretaries' relationship to the petitioners is extremely remote.

The fact is that petitioners have been denied release from the service by none of these people. Their true "jailer" is that faceless amalgam of decision-making process known as the "bureaucracy." Perhaps it would be possible to trace through the baroque interior of the petitioners' cases and discover the precise point that it was decided that they would not be released from the service, and designate that moment, the point at which they were placed "in custody" and the particular decision-maker as their "custodian."33 Yet such a process would often be impossible and would not be particularly helpful. It is far more helpful simply to accept reality and think of the "custodians" of inactive reservists as being, in a general sense, the Armed Services of which they are a part.

In the extreme example of non-physical "custody" presented by the inactive reservist, the petitioner's "custodian" accompanies him everywhere he goes. If the restraints in this case are largely psychological, they will exist in the petitioner's mind regardless of his location.

C. The Location of the Proper Forum.

Accepting our analysis of the nature of the inactive reservists' "custody" and his "custodian" does not mean however, that he may choose his forum by simply appearing personally in a jurisdiction, thereby bringing his mental "jailer" along with him. To think in such terms is to make the same error that the parties here make in keying jurisdiction to the location of the "custodian." This whole issue simply cannot be allowed to turn on the location of an entity whose existence is such a source of controversy.

Indeed, the Supreme Court has specifically held that the mere presence of a petitioner in a forum, standing alone, is not sufficent to support a court's jurisdiction. In Schlanger v. Seamans34 the Court was asked to consider a denial of habeas relief to a serviceman who had obtained a temporary leave of absence from the service to attend college in Arizona. The Court held that the mere presence of the petitioner in Arizona to attend school was not enough to permit him to bring the action there.

While the situation of a serviceman on leave to attend school is not entirely analogous to that of an inactive reservist, we think that the principle established in that case is applicable here; mere presence in a jurisdiction is not enough to support a habeas action. The Schlanger result is, of course, consistent with the functional method of determining the proper forum for habeas actions. If one were permitted to establish jurisdiction by merely being before the court, a petitioner could select any forum he wished; this would be forum shopping in its most extreme form.

There is much language in the Schlanger opinion that indicates the Court felt jurisdiction would not lie in Arizona because the petitioner's "custodian" or commanding officer was elsewhere.35 We do not feel that this language, even assuming that it was determinative in that case, is of relevance here. The petitioner in Schlanger was not an inactive reservist, but rather was on leave of absence. His commander was not the keeper of some records in an obscure place where the petitioner had never been. His commander was the commanding officer of petitioner's last duty station. Schlanger's commanding officer had been in charge of his conduct in a very physical and immediate way and the petitioner had remained subject to the commanding officer's control while on his leave of absence. All of Schlanger's contacts with the Army had been through his chain of command, and there is no suggestion that confusion existed regarding to whom the petitioner was responsible.

On these facts we see that Schlanger's condition was more analogous to that of a parolee who, although free, has retained continuing contacts with a particular, easily identifiable entity. In such a situation the concept of "custodianship" may retain viability since the "custodian" is not an object of speculation as with the inactive reservist. In short, the nature of the "custodian" in Schlanger was similar to the traditional definition of that term. As we have shown, the "custodian" in the inactive reserve context bears little resemblance to its traditional counterpart. We therefore do not feel that the Supreme Court's discussion of "custodianship" in Schlanger is relevant to this case. And, as the Court plainly stated in Strait, "[t]he jurisdictional defect in Schlanger, however, was not merely the physical absence of the Commander of Moody AFB from the District of Arizona, but the total lack of formal contacts between Schlanger and the military in that district."36

The ultimate answer to where jurisdiction does lie for the inactive reservist may be found in a careful consideration of the Supreme Court's opinion in Strait v. Laird.37 That case involved an inactive reservist who, unlike the petitioner in Schlanger, had never been on active duty and who sought release from the military as a conscientious objector. The petitioner in Strait had been a longtime resident of California and it was in that state that he sought release from the military, first by application to the military and failing that by petitioning for a writ of habeas corpus in the federal court. The Government argued that the California courts had no jurisdiction because Strait's "custodian," supposedly the same commander of the record center in Indiana involved in this case, was not within the court's jurisdiction. The Government's argument was based on an interpretation of Schlanger that would require actions always to be brought in the jurisdiction of the "immediate custodian." The Strait Court wisely rejected this mechanistic method of solving the problem and inquired into the factors that were relevant to a functional determination of the proper forum.

The Court in Strait felt there were essentially four factors relevant to its determination that California was the proper forum for the petitioner's action: (1) California was Strait's "home"; (2) he had never been on active duty; (3) California was where he had received his commission; and (4) California was where petitioner had made his application for release from the military and where the application had been processed. The Court concluded with the observation that "[i]t was in California where . . . [petitioner] had his only meaningful contact with the Army . . . ."38

Under these criteria it is quite clear that appellant Gelber may bring his action in Massachusetts. Massachusetts is Gelber's domicile and it is where he has applied for release and engaged in negotiations with the military to obtain release. As in the Strait case, Massachusetts is where the petitioner has "had his only meaningful contact with the Army . . . ."

The meaning of the Strait opinion for appellant Eisel is, however, less clear. Eisel moved from Massachusetts to New York after his initial application for release and interviews with service personnel. Thus, while Eisel had his "most meaningful contact" with the service in Massachusetts, his "home" is in New York. The Court in Strait did not tell us which of the several factors it considered were determinative there. Undoubtedly, it considered all of them relevant and did not wish unnecessarily to decide which should prevail when they might be in conflict. Here two of the factors are in conflict and we must determine which should be allowed to control.

It would be possible to establish a rule that held the forum to be where the serviceman had his most significant contacts with the military. Such a choice has the rather obvious disadvantage, however, of being vague and easily subject to misinterpretation. It may well be that an inactive reservist may have contacts with the military in a number of different places. Deciding where the place of the "most significant contacts" is might be as difficult as finding the mythical "immediate custodian."

Employing the petitioner's home or domicile39 as the determining factor does not have this disadvantage. If we apply the six criteria used to analyze the suggested "custodian" tests in Part II above, we see that a standard based on one's domicile is virtually free of objections. First, witnesses such as ministers, employers, and friends who would testify regarding the petitioner's status as a conscientious objector are more likely to be located in the jurisdiction of the petitioner's domicile. Second, such a rule will promote an even distribution of this type of habeas cases throughout the country; the number of such actions in any given jurisdiction should be roughly proportionate to the size of the population. Thirdly, it is clearly a convenient forum for the petitioner. Fourthly, there does not appear to be any reason to think that a domiciliary test would work an inconvenience on the Government. Fifthly, the forum under a domiciliary test would be relatively easy to determine. There are of course situations in which the exact location of a person's domicile is unclear. Such situations are relatively uncommon; they are certainly not as difficult of resolution as the alternatives that have been suggested to us. In addition, our judicial system is familiar with the concept of domicile-the problem has arisen in the context of decedents' estates, divorce, and taxation-and there will be a body of law that courts can draw upon to resolve controversies. Finally, such a rule would establish a single exclusive forum, thereby preventing forum shopping. Traditionally a person has been able to have only one domicile. It is also not possible to change one's domicile by simply moving from place to place.

We believe that our proposed rule of domicile is both correct under existing law and consistent with the policy considerations relevant to determining the proper forum in habeas cases. Admittedly, there is language in the Strait opinion which might be read as an indication that the Court there considered the concept of "custodianship" in concluding that the action was properly maintainable in California.40 This language is more properly read, however, as a statement that the "custodian" was in California because that is where appellant had his "home" and his most significant contacts with the military. There is no suggestion in the opinion that California is the proper forum because petitioner's immediate custodian is located there. The Court did conclude that the custodian was present in California; that presence, however, was of a highly metaphysical nature. The discussion was no doubt included in order to eliminate the possibility that the Court's opinion might be misinterpreted as doing away with the need for a "custody" and a "custodianship." Likewise, in this case we are not saying that appellants are not in "custody" or that their "custodians" are not located within the jurisdiction of the forum. We are simply saying that the location of the "custodian" in this situation assumes such an uncertain character that its location should not be made the test for determining the proper forum.41

Based on the above, we conclude that appellant Eisel may bring his action in the jurisdiction of his domicile, New York.

The injunction of the District Court has remained in effect pending the disposition of this appeal. Since we have denied appellants permission to bring their action in the District of Columbia and have suggested the jurisdictions that we believe are proper, the issuance of the mandate in this case is stayed for a period of thirty days to permit appellants to take whatever action they deem appropriate in light of our decision.42

The action of the District Court in dismissing petitioners' suits for lack of jurisdiction, for the reasons stated in our opinion herein, is

Affirmed.