No. 73, Docket 80-7348.United States Court of Appeals, Second Circuit.Argued September 12, 1980.
Decided December 17, 1980. As Modified January 27, 1981.
[EDITORS’ NOTE: THIS PAGE CONTAINS HEADNOTES. HEADNOTES ARE NOT AN OFFICIAL PRODUCT OF THE COURT, THEREFORE THEY ARE NOT DISPLAYED.]
Francis J. MacGregor, Asst. Atty. Gen., Hartford, Conn. (Carl R. Ajello, Atty. Gen., Gerard J. Dowling, Asst. Atty. Gen., Hartford, Conn., of counsel), for defendant-appellant.
Jerome M. Griner, West Hartford, Conn., for plaintiff-appellee.
Richard S. Cohen, Atty. Gen. of the State of Maine, Augusta, Me. (James D. St. Clair, William F. Lee, Hale Dorr, Boston, Mass., of counsel), Special Counsel to the State of Maine as amicus curiae.
Thomas N. Tureen, Portland, Maine, Richard B. Collins, Boulder, Colo., Native American Rights Fund, counsel for Passamaquoddy Tribe and the Penobscot Nation as amicus curiae.
Appeal from the United States District Court for the District of Connecticut.
Before FEINBERG, Chief Judge, and LUMBARD and OAKES, Circuit Judges.
LUMBARD, Circuit Judge:
 This appeal requires us to decide whether the Nonintercourse statute, which, since 1790, has prohibited the sale of Indian land unless conveyed by a treaty approved by the federal government applies to land throughout the United States. The appeal arises out of a suit brought against the State of Connecticut in the District Court for Connecticut by the Mohegan Tribe of Indians to regain possession of some 2,500 acres of land in the Town of Montville, Connecticut. In its defense, the State of Connecticut filed a motion to dismiss the suit on the ground that the Nonintercourse statute was intended to apply only to land in “Indian country” and thus did not protect the land at issue here. Judge Blumenfeld denied the State’s motion, holding that the statute was meant to apply to Indian lands throughout the United States. Mohegan Tribe v. State of Connecticut, 483 F.Supp. 597 (D.Conn. 1980). At the State’s request, Judge Blumenfeld certified the question of the statute’s geographic applicability for appeal pursuant to 28 U.S.C. § 1292(b), and we consented to hear it. We find ourselves in substantial agreement with the reasoning of the district court and therefore affirm.
 In the past few years numerous suits have been brought by Indian tribes still residing in the eastern parts of the United States. These tribes have asserted claims to large tracts of land in the East, thereby throwing into uncertainty the validity of land titles throughout the area.
 These suits have been based upon the claim, after a century and a half of occupation by non-Indians, that the states in the East entered into treaties with and purchased land from Indian tribes after the passage of the Nonintercourse statute, which by its terms apparently forbade such transactions without the participation of the federal government. To date, the Indians have been largely successful in their legal battles regarding their claims to the eastern lands. Defenses based upon state
adverse possession laws and state statutes of limitation have been consistently rejected. The only grounds upon which the States have thus far succeeded in defeating Indian claims is in demonstrating that plaintiffs in these suits do not properly represent an existing tribe which can be proved to be the legitimate descendant of the original landholding tribe.
 In these suits, defendant states have marshalled historical evidence which suggests that the eastern Indian tribes and their lands were always understood to be under the jurisdiction of the states. While these arguments have been held to be unavailing in a number of other contexts, such as whether the eastern tribes were properly considered “tribes” under the protection of the federal government, and whether they were considered “tribes” for purposes of sovereign immunity to suit, until this action, no court has had to address directly the issue of whether the Nonintercourse statute was intended to apply to land held by the eastern tribes. The State’s argument is admittedly appealing in that it would explain why both the states and the federal government have ignored so completely what the Indians assert to be the dictates of the Nonintercourse statute. To determine the question of the geographic applicability of the Nonintercourse statute, however, it is necessary to trace the history of the Indian statutes and relevant aspects of Indian land tenure in this country.
 I. HISTORY OF INDIAN LEGISLATION
 In the Royal Proclamation of 1763, the British Crown declared that the power to “grant Warrants of Survey, or pass Patents” for land in the American territory resided solely in the Crown and not in the Governors or Commanders in Chief of the Colonies. The lands were “reserved to the … Indians” unless the sovereign was to decide otherwise. 1 Laws U.S. 443-48. Moreover, the Proclamation, for apparently the first time, established a boundary line between the Indian territory or “Indian country” and the lands of the colonists. Trade which occurred beyond the line could only be conducted with the consent of the Crown. At the same time, land grants throughout the American territories were solely within the authority of the sovereign: grants “upon any Lands whatever, which, not having been ceded to or purchased by us as aforesaid, are reserved to the said Indians, or any of them.” Id. Thus, the policy of the Royal Proclamation was to demarcate an “Indian country” within which trading could only be conducted with the approval of the Crown, and to establish that all grants of land from the Indians would be valid only with the approval of the sovereign.
 During the Revolution and prior to the Articles of Confederation, the Continental Congress created a Department of Indian Affairs in 1775. The jurisdiction of the Department was divided into three areas, again all lying along the border of so-called “Indian country.” 1 Laws U.S. 597. It was not until after the adoption of the Articles of Confederation in 1781, however, that the policy of the colonies toward Indian trade and, in particular, toward alienation of Indian lands was codified. First it is essential to note that Article IX of the Articles of
Confederation gave the central government exclusive power over Indian affairs with the limitation that such power could not be exercised in such a manner as to infringe upon the “legislative right” of any state. Thus, in the Resolve of 1783, 1 Laws U.S. 607-08, Congress affirmed the prior policy of the Royal Proclamation of 1763 by establishing that all land transactions with the Indians would be invalid unless approved by the federal government, but it limited the effect of the Resolve to lands “without the limits or jurisdiction” of the states. Id.
 Federal policy during the Confederation thus continued the requirement that dealings with the Indians respecting land sales would only be valid if accomplished with the authority and approval of the central government. However, due to the division of authority between the states and the central government as established in the Articles of Confederation, federal authority was limited to transactions with Indians outside the “limits or jurisdiction” of the states. On this point the Resolve of 1783 specifically referred to the division of powers established in the Articles of Confederation. See id.
 The limitations on federal authority to deal with Indian affairs contained in Article IX of the Articles of Confederation created uncertainty over the relative spheres of federal and state authority and were removed in the Constitution. Instead, the Constitution granted Congress the authority to “regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes.” U.S.Const. art. I, § 8, cl. 3. Thus, Congress was granted plenary authority to regulate trade with the Indian tribes throughout the United States.
 In 1790, Congress enacted the first of the Indian Trade and Intercourse Acts. Act of July 22, 1790, ch. 33, 1 Stat. 137 (“1790 Act”). The majority of sections of the Act established a system of licensing for trade with the Indians and imposed federal authority over crimes committed on Indian property. Section 4 of the Act contained the first Nonintercourse statute:
And be it enacted and declared, That no sale of lands made by any Indians, or any nation or tribe of Indians within the United States, shall be valid to any person, or persons, or to any state, whether having the right of pre-emption to such lands or not, unless the same shall be made and duly executed at some public treaty, held under the authority of the United States.
 The statute thus continued the policy of placing authority to extinguish Indian land rights in the sovereign.
 In order to appreciate the meaning of the first Nonintercourse statute, the peculiar nature of Indian land tenure in the United States must be noted. In accordance with the “right of discovery” of the European settlers, native Indians found in this country were granted the “right of occupancy” to their lands. That is, the natives were allowed to remain upon their lands, but their freedom to alienate those lands was restricted. The land could be sold only to the European settlers or the governmental authority representing those settlers. This would prevent, for example, Indians from selling their lands to another foreign government, hostile to the settlers. Thus, while the Indians retained the “right of occupancy,” the settlers retained the “fee
interest” in the land and retained a “preemptive right” to purchase the land from the Indians. After the Revolution, this “pre-emptive right” lay in the individual states — at least in the already settled part of the country. And when the states joined the Union, unless they ceded the lands, they retained their “pre-emptive rights.” Nevertheless, the right to purchase Indian lands was not inconsistent with federal control over the extinguishment of Indian occupancy. Thus, the first Nonintercourse statute provided that even where the states retained “pre-emptive rights” to purchase the land, the federal government was responsible for overseeing any transfer of land from the Indians to the states.
 The provisions of the Indian Trade and Intercourse Act of 1790 were amended and reenacted five times. In 1793, Congress supplanted the Act of 1790, and continued the regulation of trade with the Indian tribes, providing in more detail for the licensing of such trade and for the enforcement of criminal law on Indian property. More important for our purposes, section 8 of the 1793 Act continued the Nonintercourse statute with minor changes:
And be it further enacted, That no purchase or grant of lands, or of any title or claim thereto, from any Indians or nation or tribe or Indians, within the bounds of the United States, shall be of any validity in law or equity, unless the same be made by a treaty of convention entered into pursuant to the constitution … Provided nevertheless, That it shall be lawful for the agent or agents of any state, who may be present at any treaty, held with Indians under the authority of the United States, in the presence, and with the approbation of the commissioner or commissioners of the United States, appointed to hold the same, to propose to, and adjust with the Indians, the compensation to be made for their claims to lands within such state, which shall be extinguished by the treaty.
 When reenacted in 1796, the Act, for the first time, contained in its first section the description of a boundary line “established by treaty between the United States and various Indian tribes,” extending from Lake Erie down the Cayahoga and Ohio Rivers, through Kentucky and eventually to South Carolina, delineating settled territory from “Indian country.” The first section also contained a proviso that:
if the boundary line between the said Indian tribes and the United States, shall, at any time hereafter, be varied, by any treaty which shall be made between the said Indian tribes and the United States, then all the provisions contained in this act, shall be construed to apply to the said line so to be varied, in the same manner, as the said provisions now apply to the boundary line herein before recited.
 1796 Act, § 1. Again a licensing system was imposed for trade beyond the boundary, and crossing the boundary for certain purposes was prohibited. The Nonintercourse statute, with essentially the same language as in the 1793 Act, became section 12 of the 1796 Act. The Trade and Intercourse Act, as a whole, was reenacted with similar language in 1799 and 1802.
 In 1834, however, certain changes relevant to the questions before us were enacted. First, the description of the boundary line between the United States and Indian country along with the proviso regarding changes in the line was eliminated and replaced with the description of Indian country as “all that part of the United States west of the Mississippi, and not within the states of Missouri and Louisiana, or the territory of Arkansas, and, also, that part of the United States east of the Mississippi
river, and not within any state to which the Indian title has not been extinguished.” Thus, Indian country did not include any lands within the borders of these states. At the same time, the Act eliminated the phrase “within the bounds of the United States” in the Nonintercourse statute, so that no words of geographical application existed. The trade restrictions, contained in the other sections of the Act, were expressly limited to Indian country. However, section 29 of the 1834 Act stated that the repeal of the former sections of the earlier Acts would not “impair or affect the intercourse act of eighteen hundred and two, so far as the same relates to or concerns Indian tribes residing east of the Mississippi.”
 The present Nonintercourse statute, 25 U.S.C. § 177, states simply that:
No purchase, grant, lease, or other conveyance of lands, or of any title or claim thereto, from any Indian nation or tribe of Indians, shall be of any validity in law or equity, unless the same be made by treaty or convention entered into pursuant to the Constitution.
 One other section of the Act, important to the question of the Nonintercourse statute’s applicability to eastern lands, should be mentioned. Beginning in the 1793 Act, all of the Acts contained a section which provided that:
nothing in this act shall be construed to prevent any trade or intercourse with Indians living on lands surrounded by settlements of the citizens of the United States, and being within the jurisdiction of any of the individual states.
 1793 Act, § 13. This “surrounded by settlements” exception was contained in all of the Acts except that of 1834 when it was eliminated. No comparable provision exists today.
 The State of Connecticut makes two arguments based upon the Acts. First, the State argues that the Nonintercourse statute was never meant to apply to land outside of Indian country but was designed only to assure that land on the Western frontier of the country would be obtained from the Indians solely through federal treaties. Indeed, the State asserts that all of the provisions of the various Acts were meant to apply only in Indian country. Second, in the alternative, the State argues that the “surrounded by settlements” provisions were meant to exclude from the operation of the Nonintercourse statute and the Act as a whole, Indian lands in the already developed states. We now turn to these contentions.
 II. STATUTORY LANGUAGE
 As the district court noted, the proper place to begin in construing a statute is with the language of the statute itself Touche Ross Co. v. Redington, 442 U.S. 560, 99 S.Ct. 2479, 61 L.Ed.2d 82 (1979). At first blush, it is apparent that the language of the Nonintercourse statute makes its dictates applicable to Indian land located anywhere in the United States. Section 4 of the Act of 1790, the original Nonintercourse statute, quoted above, contains no language suggesting any geographical limitation. In addition, section 4 expressly applies to “any state, whether having the right of pre-emption to such lands or not.” This language makes plain that the statute was intended to proscribe land conveyances to the states, and therefore was not limited to protection of land outside the boundaries of any state. As noted earlier, only the original thirteen colonies had rights of pre-emption to Indian lands, yet the existence of this right of pre-emption did not exclude the existence of federal authority to determine the conditions under which Indian title to land within those states would be extinguished.
 The 1793, 1796, 1799, and 1802 Nonintercourse statutes contain the language that no purchase of Indian lands “within the bounds of the United States” will be
valid unless accomplished by a federal treaty. This is especially significant considering that, in these Acts, a detailed boundary was drawn which divided settled areas from Indian country. While this language was eliminated from the 1834 Act (along with the detailed description of the Indian boundary), section 29 of the Act specified that the provisions of the 1802 Act, which contained the “within the bounds” language, were still in force with respect to tribes east of the Mississippi. In any event, the 1834 Act and the present statute also contain no language limiting the applicability of the land transfer provisions.
 The State of Connecticut contends, however, that the language in the original Non-intercourse statute referring to sales to states with rights of pre-emption and language in subsequent statutes referring to sales to authorized state agents was meant solely to include within federal control land purchases by those of the original states through which the Indian boundary passed, so that those lands within Indian country yet also within the bounds of the original states would be covered. Further, the State argues that the “within the bounds of the United States” language was merely meant to exclude land transactions with Tribes residing outside of the United States who claimed land within the United States. The second contention is simply historically inaccurate, since the United States in fact entered into treaties with tribes under the jurisdiction of foreign countries respecting land within United States borders, and these treaties complied with the requirements of the federal statutory restraints against alienation. See e. g., Treaty of May 31, 1796, with the Seven Nations of Canada, 7 Stat. 55, which conformed to the requirements of section 12 of the 1796 Trade and Intercourse Act. See generally, R. Clinton M. Hotopp Judicial Enforcement of the Federal Restraints on Alienation of Indian Land: The Origins of the Eastern Land Claims, 31 Maine L.Rev. 17, 30 n. 71 (1979).
 The contention that the statute as well as all the provisions of the various Acts was meant to apply only to those parts of the original states which were considered to be within Indian country, while plausible, is amply refuted by inspection of the language of other sections of the various Acts and the structure of the Acts as a whole. To begin with, it is plain from the other sections of the various Acts that Congress was quite explicit when it needed to refer either to Indian country or to the Indian boundary. For example, section 3 of the 1790 Act refers to criminal penalties against those who “attempt to trade with the Indian tribes, or shall be found in the Indian country.” The 1793 Act contains the same language in its section 3. Section 6 of the 1793 Act refers to purchases of “any horse [from] an Indian, or [from] any white man in the Indian territory.” In the 1796 Act, where Congress was dealing with lands lying west of the described border of Indian country, it did so specifically: section 2, for example, refers to penalties against citizens who cross over, or go within the “said boundary line.” And section 14 refers to Indians who cross the “said boundary line.” Section 16 of the Act of 1799 refers to persons “found in the Indian country over and beyond the said boundary line.” The 1834 Act states specifically in section 2 that trade with the Indians “in the Indian country” is prohibited unless a license is first obtained from the government. The term “Indian country” also appears in sections 3, 4, 6, 7, 10, 16, 17, 20, 23, 24, and 25 of the 1834 Act. Yet as noted earlier, the Nonintercourse statute contained in the 1834 Act makes no mention of any geographic limitation.
 Aside from the plain language of the Nonintercourse statute and the fact that the term “Indian country” does appear in other sections of the Act, other elements of the structure of the Act also rebut the State’s argument. First, as noted by the district court, the “surrounded by settlements” exception would make no sense if the Acts as a whole were meant to apply only to Indian country, since no territory within Indian country could have been so surrounded. Second, as also noted by the district court, the provision respecting the President’s power to take measures to prevent the sale of alcohol to the Indian tribes, originally adopted in the 1802 Act without any language of geographical limitation, was amended in 1822 to state that the President would be given the power in particular to authorize searches of the wares of traders upon information that “ardent spirits” were being “carried into the Indian countries by said traders.” Act of May 6, 1822, ch. 581, 3 Stat. 682. When incorporated into the 1834 Act, the provisions relating to alcohol were specifically limited to “Indian country.” 1834 Act, § 20. These changes would have been unnecessary had the 1802 Act applied only in Indian country.
 Third, it is particularly striking that despite the large number of references to Indian country and Indian territory contained in the Acts, mentioned above, a number of other provisions including the Nonintercourse statutes contain no language of geographic limitations. These provisions contain restrictions respecting trade simply with “any Indian tribe” or respecting “lands belonging to any Indian tribe.” For example, section 5 of the 1793 Act imposes penalties for making a settlement upon “lands belonging to any Indian tribe.” Similarly, section 9 of the 1793 Act provides for furnishing services to “the friendly Indian tribes” without geographic limitation. Yet Section 6 of the 1793 Act specifically prohibits horse trading without a license “in the Indian territory.” Thus examination of the provisions of the 1793 Act suggests that Congress was careful to distinguish between regulations applicable only to Indian country and those applicable to all Indian tribes and their lands. Further examples could be drawn from the subsequent Acts. We conclude that, by failing to use language restricting applicability of the Nonintercourse statute, Congress intended the statute to apply throughout the United States.
 Indeed, as noted by the district court, the only arguments with respect to the language of the Acts that the State can muster in support of its view that the Nonintercourse statute applied solely in Indian country are rather tenuous. The State first points to the titles of the various Acts. For example, the 1796 Act is entitled, “An Act to Regulate Trade and Intercourse with Indian Tribes and to Preserve Peace on the Frontiers.” From this, the State argues that the statute was concerned only with land transactions on the frontiers. Apart from the purposes of the Acts, which we will discuss below, it is sufficient to note here that the “Peace on the Frontiers” language was not contained in the titles of the 1790 or 1793 Acts, and that, although useful in interpretation of ambiguous provisions, the title of an Act cannot be held to contradict the more specific language contained in the body of the Act.
 The only other argument based upon statutory language concerns the provisos contained in the 1796, 1799 and 1802 Acts’ description of the Indian boundary. These provisos stated that if the boundary is subsequently altered by treaties with the Indians, “all the provisions contained in [the Acts] shall be construed so as to apply to the said line so to be varied in the same
manner, as said provisions apply, by force of [these Acts] to the boundary line hereinbefore.” (Emphasis added.) The State emphasizes the first few words of the proviso, but when read as a whole it is quite plain that the proviso applied if subsequent treaties should alter the boundary and provided that those sections which did apply to the boundary would similarly apply to any new boundary established by those treaties. Hence, those sections that did not apply to the boundary would be left unaffected.
 Thus, the conclusion to be drawn from the language of the Nonintercourse statute and the various Acts is that the statute was meant to apply to Indian land throughout the United States. In regulating relations with the Indians, Congress imposed certain restrictions on transactions occurring in Indian country and certain others to those involving Indians situated throughout the United States. The Nonintercourse statute, containing no language of limitation, must then be read as applying to all Indian lands.
 III. LEGISLATIVE AND JURISDICTIONAL HISTORY
 Connecticut, however, makes a number of arguments for its interpretation of the statute based upon external evidence: first, the State argues that the purposes of the Trade and Intercourse Acts were both to preserve peace on the frontier and to enforce treaty obligations and that neither of these purposes would require protection of lands outside of Indian country. Moreover, it finds support for these views in the history of English and colonial legislation regarding Indian lands as well as federal Indian policies during the Confederation. Finally, the State points to evidence of executive branch interpretation of the Acts also indicating that the Acts were understood to leave land transactions with eastern Indian tribes to the control of the states. We are not persuaded.
 As the Supreme Court has noted with respect to Indian legislation and treaties, “[t]hese instruments . . . cannot be interpreted in isolation but must be read in light of the common notions of the day and the assumptions of those who drafted them.” Oliphant v. Suquamish Indian Tribe, 435 U.S. 191, 206, 98 S.Ct. 1011, 1019, 55 L.Ed.2d 209 (1978); accord, Wilson v. Omaha Indian Tribe, 442 U.S. 653, 666, 99 S.Ct. 2529, 2537, 61 L.Ed.2d 153 (1979). Certainly, courts should never ignore strong extrinsic evidence which may serve to explain the meaning of statutory enactments, particularly when the statutes are as deeply embedded in American history as are those relevant here.
 Legislative history is practically non-existent on the particular issue of geographic applicability. There is only a single sentence in the House Report on the 1834 Act which suggests that that Act was meant only to apply outside the jurisdiction of any of the states. However, as the district court recognized, the effect of this isolated statement is undercut by the subsequent amendment to the Act which added section 29 — the section which continued in effect provisions of the 1802 Act insofar as they applied to Indians east of the Mississippi.
 Apart from this, the State’s argument as to Legislative intent is not so much directly on the point of geographic applicability as it is on the motivation for granting federal protection to the Indians. The evidence suggests that federal Indian policy was based upon the need to prevent Indian uprisings. Accordingly, certain concessions were granted to the Indian tribes in order
to avoid the necessity for large-scale military intervention.
 Thus, the State directs our attention to statements by President Washington, and his Secretary of War, Henry Knox, both of whom played a large part in developing the original Trade and Intercourse Act, regarding the necessity of restraining settlers from encroaching upon Indian lands on the frontier, since these encroachments were a constant source of Indian unrest. The State also points to the writings of several historians respecting Washington’s and Knox’s desire to satisfy the concerns of the Indians over the activities of settlers along the frontier. While the evidence rather convincingly demonstrates that the nation’s early leaders were perhaps not so charitable toward the Indians as we have come to view them, and although this evidence quite readily demonstrates that contemporary attitudes have colored our views of the original motives behind American Indian policy, we do not believe that the evidence in any way rebuts the Nonintercourse statute’s applicability to Indian lands throughout the United States. Of course, Washington and Knox and their countrymen were concerned not only with protecting the Indians, but with preventing the onset of overt hostilities. Moreover, it is true that peace along the frontier, and in particular the prevention of encroachment by non-Indian settlers on Indian lands along the frontiers, were primary objects of the Act’s land provisions. Nevertheless, there is nothing inconsistent with these views in reading the Nonintercourse statute, as its language suggests, to include encroachment upon Indian lands throughout the United States. Such a promise may well have been thought best to mollify the apprehensions of Indians on the frontier as well as those in the settled regions. Surely it would have furnished evidence to the Indians of the seriousness of the federal government’s intention to regulate land sales, a matter of great concern to the Indians.
 The State also points to considerable evidence that the Acts were designed to enforce various obligations imposed by treaties with the Indians entered into by the federal government, and therefore the Acts were not concerned with the eastern tribes with whom no treaties were held. One of the most prominent commentators on Indian Law, Felix Cohen, has written that “each of the substantive provisions of the first Indian Trade and Intercourse Act fulfilled some obligation assumed by the United States in treaties with various Indian tribes,” and that this analysis “would probably apply equally to each of the later trade and intercourse acts.” F. Cohen, Handbook of Federal Indian Law
69-70 (1945). However, this does not support the view that the framers of Indian legislation did not also wish to protect Indian tribes located throughout the United States. Indeed, it may well have been to placate tribes located on the frontier or with whom the United States had entered into treaties that federal protection was granted to all Indian lands.
 We thus agree with the district court that there is no evidence demonstrating that peace on the frontier and enforcement of treaty obligations were the sole purposes of the various Acts. We share the district court’s intuition that if the states were meant to retain control over eastern Indian lands, some evidence of such authority would exist. Finally, as just noted, it would seem consistent with the goals of the early administrations to convince the Indians both on the frontier and elsewhere that the federal government meant to take into its own hands the problems of intrusions upon Indian property wherever they might occur.
 As for the evidence regarding Indian policy preceding the various Trade and Intercourse
Acts, we believe that examination of this history, outlined above, rebuts rather than supports the State’s position. As noted earlier, from the time of the Royal Proclamation of 1763, it was the policy of the sovereign to retain control over land transactions with the Indians in the central government, first the British Crown and later the federal government. The only period during which this was not the case was during the Confederation. During that period, as outlined above in discussing the Resolve of 1783, federal Indian policy was constrained by the limits imposed by Article IX of the Articles of Confederation. It was dissatisfaction with the uncertainty created by these limitations that led to the expansive definition of federal power to deal with the Indians under the Commerce Clause of the new Constitution. The State’s argument that the new Congress declined to exercise its expanded powers and chose instead to maintain Indian policies established during the Confederation is highly implausible. Aside from lacking any support in the language of the enactments subsequent to the Constitution, or in any decisions of the courts, it is also rebutted by the history recited above suggesting that under its expanded powers Congress reverted to the policy established long before the Confederation that the sovereign would control all land transactions with the Indians.
 Lastly, the State asserts that evidence of the interpretation of the Acts by the executive branch supports its position. Admittedly, “[l]ongstanding, contemporaneous executive and administrative interpretation by those entrusted with the enforcement of ambiguous legislation may also shed light upon an appropriate construction” of statutory language. Leary v. United States, 395 U.S. 6, 25, 89 S.Ct. 1532, 1542, 23 L.Ed.2d 57 (1969). The State points out that President Washington himself wrote in 1792 that the New England tribes were “so situated as to be rather considered a part of the inhabitants of the State of Massachusetts than otherwise, and that State has always considered them under its immediate care and protections.” Letter to Archbishop John Carroll, quoted in P. Guilsday, The Life and Time of John Carroll 607 (1954). In more recent times, the federal government repeatedly disclaimed any responsibility for managing the affairs of the eastern tribes. See Letter, W. A. Jones, Acting Commissioner of Indian Affairs to Francis M. Morrison, Esq., July 10, 1899; Letter, Commissioner of Indian Affairs C. J. Rhoades to Mrs. Edith M. Smith, October 12, 1931.
 Nevertheless, the fact that the federal government disclaimed responsibility for these tribes is not determinative here. We believe that, although considerable evidence amassed by the State supports the proposition that the federal government did not avail itself of the provisions of the Nonintercourse statute and appeared to leave management of the affairs of the eastern tribes to the individual states, it does not follow that the federal government had no obligation to do so, or that the states had the authority — unimpeded by the Acts — to buy land from the eastern tribes without federal approval. We note, moreover, that there is no mention of any reliance by the federal government upon any geographical limitations contained in federal Indian legislation in the evidence offered. We also note that none of the evidence concerns land transactions.
 IV. CASE LAW
 The State, finally, relies upon statements in the case law which are said to determine the question of geographic applicability of the various Nonintercourse statutes and the Trade and Intercourse Acts as a whole. In our view, however, the case law is inconclusive on the issue.
 The State points first to early decisions containing language suggesting that the Trade and Intercourse Acts were meant to be confined to Indian country. In American Fur Co. v. United States, 27 U.S. (2 Pet.) 357, 369, 7 L.Ed. 450 (1829), for example, the Supreme Court stated that “all the provisions contained in [the 1802] Act, and consequently, those contained in the amendatory
act of 1822, are by [the first section] expressly confined” to “Indian country.” The Court, in American Fur, however, was concerned with the question of penalties for those found bringing “ardent liquors” into Indian country for trade with the Indians, and it held that only those goods of the traders found within Indian country were to be forfeited. The Court was construing only that section of the Act, added by amendment in 1822, respecting importation of liquor into Indian country and the penalties for doing so. Similarly, in Bates v. Clark, 95 U.S. 204, 205, 24 L.Ed. 471 (1877), the issue was whether goods seized by American military officers after finding that the owner intended to import liquor into Indian country for trade purposes, were properly seized inside of Indian country. While the Court used broad language respecting the applicability of the Act, it was only interpreting provisions relating to penalties for importation of alcohol, by that time contained in section 20 of the 1834 Act, and further amended by statute in 1864. While admittedly, the statements made in both American Fur an Bates suggested that all the provisions of the Acts, including the Nonintercourse statute, were limited in operation to Indian country, the Court in those cases was concerned with particular provisions of the Acts and suggested only that the trade and intercourse provisions were generally limited to Indian country.
 The State also points to language, generally dicta, in various decisions recognizing state control over the eastern tribes See, e. g., Worcester v. Georgia, 31 U.S. (6 Pet.) 515, 580, 8 L.Ed. 483 (1832) (McLean, J., concurring). It should be noted, however, that these comments were made in the context of constitutional decisions respecting the reach of federal control over the Indian tribes. On these issues, Justice McLean’s views, while shared by certain state courts, see, e. g., State v. Doxtater, 47 Wis. 278, 2 N.W. 439 (1879), never commanded a majority of the Supreme Court. Instead, the majority view was that stated by Chief Justice Marshall in the Worcester opinion, noting that federal authority over Indian affairs was plenary, and that the grant to regulate “Commerce” among the Indian tribes gave the federal government authority to regulate practically all aspects of Indian affairs.
 Third, the State cites a number of early decisions respecting Indian land tenure. See, e. g., Mitchel v. United States, 34 U.S. (9 Pet.) 711, 9 L.Ed. 283 (1835); Cherokee Nation v. Georgia, 30 U.S. (5 Pet.) 1, 31, 8 L.Ed. 25 (1831) (Baldwin, J., concurring); Johnson v. M’Intosh, 21 U.S. (8 Wheat.) 543, 5 L.Ed. 681 (1823); New Jersey v. Wilson, 11 U.S. (7 Cranch) 164, 3 L.Ed. 303 (1812); Fletcher v. Peck, 10 U.S. (6 Cranch) 87, 3 L.Ed. 162 (1810). Most of these cases are concerned with various conveyances by the states of lands formerly held by the Indians. None of them suggests limitations on the applicability of the Nonintercourse statute.
 Most of the decisions cited involved transfer of a state’s “fee” interest subject to the right of Indian occupation Fletcher v. Peck, supra; Johnson v. M’Intosh, supra. In New Jersey v. Wilson, supra, the Indians had abandoned the land in question and the only issue was whether the land was still non-taxable. Justice Baldwin’s comments in Cherokee Nation, supra, also refer to the fact that, although the states were “seized in fee,” the states’ interests were subject to the Indian rights of occupation. Finally, in Mitchel, supra, the Court was analogizing to certain decisions holding that state land grants would be held valid as to land within the states, but invalid as to lands in Indian country. In the cited decisions, however Winn v. Patterson, 34 U.S. (9 Pet.) 663, 9 L.Ed. 266 (1835) Danforth v. Wear, 22 U.S. (9 Wheat.) 673, 6 L.Ed. 188 (1824) Patterson v. Jenks, 27 U.S. (2 Pet.) 216, 7 L.Ed. 402 (1829), not only did all of the state land
grants predate the Indian Trade and Intercourse Act of 1790, but one of the decisions, Danforth v. Wear, supra, made clear that even as to those transactions the state’s interest could only be conveyed subject to the extinguishment of Indian title. Thus, the early decisions of the Supreme Court do not support the State’s position that the Nonintercourse statute applied solely in Indian country.
 We turn then to the two most recent opinions bearing upon this subject, each of which contains language providing some support for the positions urged by the two parties in this appeal. I Oneida Indian Nation v. County of Oneida, 414 U.S. 661, 94 S.Ct. 772, 39 L.Ed.2d 73 (1974), the Court held that there was federal jurisdiction to hear a claim for ejectment based upon Indian title, since that title rested upon federal law. In the course of its opinion, the Court noted that:
The rudimentary propositions that Indian title is a matter of federal law and can be extinguished only with federal consent apply in all of the States, including the original 13. It is true that the United States never held fee title to the Indian lands in the original States as it did to almost all the rest of the continental United States and that fee title to Indian lands in these States, or the preemptive right to purchase from the Indians, was in the State, Fletcher v. Peck, 6 Cranch 87, 3 L.Ed. 162 (1810). But this reality did not alter the doctrine that federal law, treaties, and statutes protected Indian occupancy and that its termination was exclusively the province of federal law.
 414 U.S. at 670, 94 S.Ct. at 778 (footnote omitted).
 Most recently, in Wilson v. Omaha Indian Tribe, 442 U.S. 653, 99 S.Ct. 2529, 61 L.Ed.2d 153 (1979), the Court held that the present-day descendant of section 22 of the 1834 Act, which establishes the burden of proof in land disputes between Indians and “a white person,” did not apply to a suit between the State of Iowa and an Indian tribe. As one of the reasons in support of its conclusion, the Court noted that the original section 22 was incorporated into the 1834 Act which, according to the Court, was meant to apply only “to the whole Indian country.” 442 U.S. at 667, quoting H.R. Rep. No. 474, 23d Cong., 1st Sess. 10 (1834). Thus, the Wilson court believed that in adopting section 22, the Congress had in mind only land disputes arising outside the jurisdiction of any state.
 Although the district court found Oneida’s language more closely on point than Wilson’s, we believe that neither is determinative here. Regardless of whether the quoted language from Oneida is dictum or not, it suggests only that all Indian land transfers were subject to federal control and that extinguishment of Indian title is a federal question. It does not necessarily hold that the specific provisions of the Nonintercourse statute applied to all of those transfers. Instead, it would be consistent with Oneida to hold, as the State argues, that land transfers in Indian country were to be governed by the statute, but that the exclusion of eastern lands from the Nonintercourse statute meant that the “federal law” as to those lands was to allow the states to purchase them.
 At the same time, Wilson does not persuade us to find that the Nonintercourse statute was so limited. Wilson involved the construction of a provision enacted independently of other Indian legislation and added to the Trade and Intercourse Act in 1834. Once again, it must be recalled that section 29 of the 1834 Act provided that the 1802 Act would remain in effect with respect to tribes located east of the Mississippi. Thus, while section 22, relating to the burden of proof in land disputes, may have been meant to apply only within the Indian territory specified in section 1 of the 1834 Act, there is no reason to conclude that the Nonintercourse statute, contained in both the 1802 and 1834 Acts, was also so limited. Nothing in Wilson reflects at all upon the geographic applicability of the previous
Trade and Nonintercourse Acts, and therefore, even if it were to be held that the 1834 Nonintercourse statute was limited territorially, the Nonintercourse statute contained in the 1802 Act remained applicable by virtue of section 29 of the 1834 Act to the tribes east of the Mississippi.
 We conclude that the case law does not support the State’s contention that we should ignore the plain language of the statute and limit the Nonintercourse statute’s applicability. Moreover, while no case is controlling on the issue, Oneida
does make clear that the extinguishment of all Indian title was meant to be a matter of federal concern. Since we have found no evidence that Congress intended to treat eastern Indian lands in a different manner, it would seem reasonable to believe that Congress intended a unified federal policy toward land acquisition from the Indians. In any event, we find nothing in the case law which dissuades us from our conclusion that Congress intended the Nonintercourse statute to apply throughout the United States.
 V. “SURROUNDED BY SETTLEMENTS” EXCEPTION
 As we conclude that the Nonintercourse statute was intended to apply to lands throughout the United States, we must consider the State’s contention that the “surrounded by settlements” exception was intended to exempt from the various provisions of the Trade and Intercourse Acts transactions, including those involving land, with Indians residing in the settled areas of the eastern States. The district court disposed of this issue by adopting the view put forth in Narragansett Tribe v. Southern Rhode Island Land Development Corp., 418 F.Supp. 798, 808-09 (D.R.I. 1976). 483 F.Supp. at 599, n.9. In Narragansett, the district court held that the “surrounded by settlements” exception applied only to trade with individual Indians who had left their tribes and chose to reside in non-Indian settlements. This interpretation was also accepted by the district court in Mashpee Tribe v. Town of Mashpee, 447 F.Supp. 940 (D.Mass. 1978), aff’d on other grounds sub nom. Mashpee Tribe v. New Seabury Corp., 592 F.2d 575
(1st Cir.), cert. denied, 444 U.S. 866, 100 S.Ct. 138, 62 L.Ed.2d 90 (1979).
 We find this issue more troublesome than did the district court. It accepted the reasoning of Narragansett that the exception applied only to individual Indian land transactions because the exception was eliminated in the 1834 Act at the same time that the Nonintercourse statute was amended to exclude coverage of land transactions with individual Indians. 418 F.Supp. at 808-09. There are difficulties with this reading of the exception. The most serious one would seem to be that there is no necessary connection between the elimination of the exception, which did not contain any language suggesting that it was an exception solely to the Nonintercourse statute, and the amendment of the Nonintercourse statute itself. Indeed, the exception provided for “trade and intercourse” with Indians surrounded by non-Indian settlements, and thus, it must have applied to the trade and intercourse provisions as well. Yet examination of the 1834 Act reveals that the Act continued to proscribe certain types of transactions even with individual Indians or upon property owned by individual Indians. See, e. g., § 7 (“purchase or receive of any
Indian”); § 9 (“on any land belonging to any Indian or Indian tribe”); § 20 (“sell . . . any spiritous liquor or wine to an Indian”).
 Further doubts about the Narragansett court’s approach are raised by considering the origin of the “surrounded by settlements” exception. The exception derived from a proviso in section 1 of the original 1790 Act. That proviso stated that “the President may make such order respecting the tribes surrounded in their settlements by the citizen of the United States, as to secure an intercourse without license, if he may deem it proper.” While in subsequent Acts, the exception stated merely that it applied to transactions “with Indians,” it does not appear that Indian tribes surrounded by settlements were meant to be excluded. Indeed, the term “Indians” standing alone in the Acts has generally been interpreted to include both individual Indians as well as Indian tribes. Wilson, supra, 442 U.S. at 665-66, 99 S.Ct. at 2356-2357. It thus appears that the exception was meant to apply to transactions both with individual Indians and with tribes whenever either was situated on lands surrounded by non-Indian settlements.
 While these considerations suggest that the district court was incorrect in its reading of the exception, they do not compel acceptance of the State’s position that the exception was intended to exclude land transactions by eastern tribes. Instead, we find most persuasive the position advanced here by the Mohegan Tribe: that the “surrounded by settlements” exception was not meant to apply to land transactions at all. As just noted, the exception first appeared as a proviso in the first section of the 1790 Act. That section required that all “trade and intercourse” with Indian tribes be undertaken only by those possessing licenses from the federal government. The land provisions, which applied at the time to both individual and tribal lands, were contained in a separate section. In subsequent Acts, the only change was to place the exception in a separate section. Thus, it would appear that no substantive changes were intended by this alteration beside the important one of withdrawing the matter from Presidential discretion and making it mandatory. Moreover, the change from “tribes” in the 1790 Act to “Indians” in subsequent Acts is explained simply by the fact that section 1 of the 1790 Act itself applied only to trading with “the Indian tribes,” while the subsequent Acts also restricted certain types of trade with individual Indians. See, e. g., § 6 of the 1793 Act and §§ 9-10 of the 1796 Act.
 The land provisions of the Acts, however, always remained distinct from those regulating “trade and intercourse.” Thus, it appears that the “surrounded by settlements” exception was meant to apply only to the latter. The elimination of the exception in the 1834 Act is more plausibly explained by the fact that the 1834 Act specifically limited its trade restrictions to Indian country, see e. g., §§ 2, 6, 7, although once again, it must be recalled that section 29 of the 1834 Act continued in effect those provisions of the 1802 Act applicable to tribes east of the Mississippi. Thus, the exception would seem to have remained applicable to trade and intercourse with certain of the eastern tribes, but not to any land transactions with the states. That the 1834 Act eliminated land transactions with individual Indians from the coverage of the Act is of no significance.
 We agree with the district court that the Nonintercourse statute was not intended to be applicable solely in Indian country. We also hold that the “surrounded by settlements” exception was not meant to apply to
Indian land transactions. The order denying defendant’s motion to dismiss is affirmed.
and, quite recently, of the Maine claims at issue i Passamaquoddy.