‘The Right to Shoot Himself’: Secession in the British Commonwealth of Nations

ABSTRACT The ultimate test of whether an association is voluntary or not is if you can leave it. It is difficult, at this remove, to appreciate how live an issue secession from the British commonwealth of nations was in the 1920s and 1930s. It occupied an inordinate amount of time and negotiation for a doctrine that had been ostensibly conceded in 1920. Yet, much as with the case of the appeal to the judicial committee of the privy council, once the dominions sought to take advantage of the freedom which had been guaranteed by official statements, they found a formidable amount of diplomatic pressure and legal opinion brought to bear to indicate that no such right could be officially declared. This article traces the evolution of the arguments about the right to secede in the 1930s, and examines how the right came eventually to be exercised in the case of the new commonwealth countries in the 1940s. It concludes by examining how the doctrine of secession as developed in the 1930s was abandoned in order to retain Indian membership in the commonwealth.


I. Secession in the Aftermath of the First World War
By the turn of the twentieth century, there was no doubt that it was possible to leave the British empire. The American war of independence had meant the loss of the thirteen colonies and had clearly established that secession by revolution was a clear precedent. What we are concerned with, however, is the question of secession within the empire by non-violent means. The question of whether it was possible to leave the empire by lawful means can be seen on two constitutional planes: political and legal. It could be possible for political actors to concede that it was constitutionally permissible to secede, but for no legal means to exist by which to give effect to this desire.
The classical political constitutional formulation of the British position was by Bonar Law, then leader of the house of commons, who stated in the commons in 1920: There is not a man in this House … [who] would not admit that the connection of the Dominions with the Empire depends upon themselves. If the self-governing Dominions, Australia, Canada, chose to-morrow to say, 'We will no longer make a part of the British Empire', we would not try to force them. Dominion Home Rule means the right to decide their own destinies. 1 This quote was used at the time, 2 and has been used since, to ground a whiggish narrative of inevitable progression towards the recognition of secession. This, however, overlooks the contested nature of the concept, both politically and legally during the inter-war years. The immediate response of the British government was to distance themselves from the implications of this comment. In a telegraph to the governor general of South Africa, Viscount Milner, the colonial secretary, noted that the statement 'does not alter the fact that secession would be unconstitutional and disloyal and that we should be legally and morally justified in resisting it even if we did not choose to do so'. 3 The right to secede from the British commonwealth was to assume a particular prominence in the negotiations which eventually led to the articles of agreement for an Anglo-Irish treaty in 1921. In a reply to an offer of dominion status for southern Ireland by David Lloyd George, prime minister of the United Kingdom, Eamon de Valera stated that dominion status was 'illusory' in light of the geographical proximity of Ireland to the UK and that 'the most explicit guarantees, including the right to Dominions' acknowledged right to secede, would be necessary to secure for Ireland an equal degree of freedom' with more geographically distant dominions. 4 This earned a sharp rebuke from Lloyd George in terms which directly contradicted the position of Bonar Law in the commons: 'we must direct your attention to one point … upon which no British Government can compromise, namely, the claim that we should acknowledge the right of Ireland to secede from her allegiance to the King. No such right can ever be acknowledged by us. The geographical propinquity of Ireland to the British Isles is a fundamental fact'. 5 The British government refused to countenance the possibility of Ireland outside the commonwealth, and it was in fact the first article of the articles of agreement for a treaty agreed between British and Irish representatives. 6 This was also represented in the first article of the new Irish Free State constitution: 'The Irish Free State … is a co-equal member of the Community of Nations forming the British Commonwealth of Nations'. 7 Moreover, when pushed on Bonar Law's point in 1922, Winston Churchill, then secretary of state for the colonies, expressly declaimed the statement, maintaining that the UK had never admitted the right to secede, and the dominions had never claimed it. 8 Thus, we may say that in 1922 the British government maintained that no right to secession existed in theory and the attempt by Irish representatives to establish the point was vehemently and emphatically denied by the British government.

II. Imperial Diplomacy in the 1920s: The Right to Shoot Oneself
In the aftermath of the Irish negotiations, the issue of secession was seized upon most vigorously by the South African prime minister, General James Barry Munik (JBM) Hertzog. Hertzog's national party had a strong Afrikaans base, and was concerned with the establishment of a distinctive course of South African diplomacy. It was Hertzog who insisted on a formula recognising dominion independence in the imperial conference of 1926 which eventually became the Balfour declaration. 9 Hertzog also argued that, in virtue of its independence, South Africa possessed the right to secede from the commonwealth, but he also maintained that South Africa had no intention of exercising the right. It was regarded by Hertzog as a necessary concomitant of independence: if there was no right to secede, then this constituted a substantive limitation on the independence of South Africa.
In 1929, the conference on the operation of dominion legislation and merchant shipping met to consider the legal implications of the 1926 conference. This conference was essentially a conference of experts below prime ministerial level. The sub-committee concerned with constitutional issues was headed by Sir Maurice Gwyer. The Irish delegation produced a transcript of the meetings based on notes that they made in those meetings. On 14 November 1929, the sub-committee briefly turned its attention to secession. 10 The matter under consideration was how to preserve the crown within the new commonwealth structure. The British side had proposed to place the crown in a protected position whereby no individual state could legally alter the position of the crown, save with the assent of all. In opposition, the Irish Free State and Canadian delegations favoured a political, rather than a legal, limitation on altering the position of the crown. The British delegation then turned to the fact that this would undermine the essential unity of the commonwealth by making the position of the crown dependent on each dominion and the United Kingdom. Sir William Harrison Moore, for the Australian delegation, pointed out that this may implicate what he termed 'the right to secede'. Henry Grattan Bushe, assistant legal advisor to the dominions office, interjected: 'Ah yes, people will say there is a right de facto to secede, but what you are now doing is providing a legal way in which to secede'. 11 The South African delegation immediately proclaimed their ability to legally secede under the provisions of the South Africa Act. The importance of the exchange rests in the increasing acceptance as a matter of political fact that secession could occur. This was ultimately accepted at the highest level the following year in the imperial conference of 1930. At that conference, Hertzog sought to protect his domestic republican flank by securing an announcement on the legal possibility of secession from the commonwealth. From the point of view of the South African government, any attempt to focus on the legal elements of the case were ill-founded. Indeed, a note of a meeting between Hertzog and Patrick McGilligan, minister for external affairs in the Irish Free State, in August 1930 records the South African view: 'General Hertzog … refuses to consider the legal part of the Dominions. For him the basis of Dominion evolution is entirely political and must be divorced from existing law. He brushes aside any reference to the Acts by which the Dominions were established, and concentrates on the principle of co-equality alone'. 12 Hertzog appears to have adopted the rather blasé attitude that the recognition of the right to secede would proceed as a matter of course; the Irish Free State officials concluded, wrongly, on the basis of his confidence in the issue that he had received an assurance from the British side. 13 A consideration of the legal power of South Africa to secede was undertaken in the Irish Free State by John Hearne, legal advisor to the department of external affairs. 14 In the memorandum, Hearne put forward an argument that any attempt to secede from the commonwealth would, under the South African constitution as it stood in 1930, be ultra vires the power of the South African parliament. His argument proceeded on the basis that it would be necessary to remove George V as king before attempting to secede, and that the position of George V in South Africa 11 Emphasis in original. 12 National Archives of Ireland (NAI): Taois/s. 7357 (memo of 1 Sept. 1930). 13 Ibid. 14 NAI: Taois/ s. 6610 (undated memorandum but note in top corner headed 'Imperial Conference 1930'). The memorandum was composed after the meeting between Hertzog and McGilligan as it makes reference to that meeting.
was not dependent on the South Africa Act 1909, which Act alone the South African parliament could amend. 15 Moreover, if the British parliament were to repeal the South Africa Act 1909, this would not put the union outside the empire, and therefore any comparable action by South Africa would have the same effect. The British government were forewarned about Hertzog's ambitions and sought to avoid the matterbetter not to waste time on 'these abominations'. 16 Despite these concerns, the political actors in the conference conceded the secession ground quite quickly. Immediately before the conference met, J.H. Thomas, secretary of state for dominion affairs, was questioned by the media about the right to secede from the commonwealth: 'Mr. Thomas suggested that that was a matter which was causing much more interest to the newspapers than it was to the delegates. No one had questioned anyone's right to secede, any more than one would question his right to shoot himself. But one might question the wisdom of that course of action'. 17 The question of whether or not a legal right to secede existed in the British commonwealth seems, moreover, to have been assumed even by the loyal dominions. James Scullin, the prime minister of Australia, stated that equality of status meant the right to secede, but was not concerned to have a declaration concerning that right as Australia had no wish to exercise it. 18 These declarations may be taken as a significant concession in terms of the political realities, rather than legal niceties, of the commonwealth. It had been commonwealth experience, moreover, that when politics and law collided, it was the legal position which inevitably bent to accommodate the political realities. Notwithstanding this political agreement between all governments, there still existed some considerable scepticism about whether or not secession was legally possible.

III. The Legality of Secession
Constitutional academics doubted whether it was legally possible to secede from the commonwealth. Commentators such as Arthur Berriedale Keith and, somewhat more equivocally, Courtney Kenny 19 expressed a belief that secession was impossible within the structure of the commonwealth of nations. Keith's views on the impossibility of secession were multi-valent.

17
The Times, 30 Sept. 1930, 12. 18 The Times, 26 Sept. 1930, 14;The Times, 28 Nov. 1930, 9. Scullin First, he endorsed the view of secession that 'some Dominions could secede, but others could not, because the right of one Dominion to do what imperils another part of the Empire might be met by the right of the part endangered to protect itself'. 20 Second, secession was impossible under the Statute of Westminster without the assent of all other parliaments of the commonwealth. This was due to the preamble which provided that no alteration in the succession to the throne was possible without the assent of the dominion and Westminster parliaments. 21 Third, he argued secession was impossible internally in both the Irish Free State and South Africa because the power to amend the constitution of each was limited by its membership in the empire; once the dominion sought to remove itself from the empire, it exceeded its vires. 22 In the case of the Free State, this was an explicit limitation by virtue of the terms of the Anglo-Irish treaty, in the case of South Africa; it was implicit in the grant of power itself.
Philip Noel-Baker, then professor of international relations at the London school of economics, also questioned whether or not the king could assent to a bill if secession affected the other members of commonwealth on the grounds that it was unclear on whose advice the crown was bound to act in such a circumstance. 23 Noel-Baker was ultimately less strident than Keith and concluded that it was practical, rather than legal, considerations which were ultimately decisive. This view of the legality of secession was not confined to the UK. In South Africa, H.J. Schlosberg argued that there could be no right to secede, as the UK did not have the right to declare itself a republic. Therefore any attempt to pass a bill for such a purpose by a dominion would have to be vetoed. 24 Moreover, Schlosberg argued that the introduction of a secession bill 'would be actual treason' as it would destroy the crown. 25 The view put forward by the South African delegation in 1929 was that the power contained in section 152 of the South Africa Act allowed the parliament to repeal any provision of the Act, including the abolition of the crown. 26 Maurice Gwyer argued that this was not a position entertained by all South African jurists, with an oblique reference to what was presumably Schlosberg's book. As we have seen, the Irish analysis of the South African position argued  London, 1929, 258-268. 24 Schlosberg, The King's Republics, 41-42. Schlosberg was a self-declared 'average Nationalist', see ibid., vii. 25 Ibid., 42. This view was also put forward by Leo Amery in relation to the equivalent position of the viceroy of India: see British Library, India Office Records (IOR), IOR/L/PO/6/50(ii) Amery to Hoare (24 Nov. 1933 from the British position that the South African parliament did not possess the power to secede. 27 In contrast, the argument in favour of secession rested on two major prongs. First, that the commonwealth was a voluntary association of nations. If it was voluntary, it followed logically that one could leave. Second, in the aftermath of the Balfour declaration, it was clear that each of the dominions was 'equal in status' with the United Kingdom. Any attempt by the UK to limit the power to secede would mean some form of inferiority on the part of the dominions, and would therefore be inconsistent with the doctrine of co-equality. The Balfour declaration was given legal form in the Statute of Westminster, and this provided the dominions with the legal power to repeal or amend any legislation, including legislation passed by the Westminster parliament. Although this did not specifically confer a legal power to secede, it followed by logical consequence of equality of status that there was no legal limit to prevent secession.
In the debates about the enactment of the Statute of Westminster in 1931, Winston Churchill adopted one of Arthur Barriedale Keith's argumentsnamely that the reservation of matters affecting the crown in the preamble meant that the assent of all members of the commonwealth was necessary. 28 Churchill did explicitly draw attention to the questionable legal use of the preamble, but he certainly appears to have believed that it bound the dominions in terms of constitutional convention. 29 This point was also seized upon by General Smuts in South Africa to argue that the terms of the Statute of Westminster limited South Africa's freedom of action. Within the dominions office in London, the legal means by which secession could be achieved was also questioned. 30 It was conceded, on the basis of Bonar Law's statement that the dominions had a political right to secede, although the legality of this was questioned. The memorandum ultimately concluded that this was an issue best avoided, and Harry Batterbee noted: 'these questions at the moment are academic'. 31 In South Africa, however, this academic argument was spurring on legislation in the form of the Status of the Union Act which we will consider later.
As late as 1942 in the United Kingdom, the king's bench division expressed their doubts as to whether the Statute of Westminster actually enshrined a right to secede in Murray v. Parkes. 32 Although the comments were obiter, 27 Ibid., 'The South African Claim to the Legal Right to Secede'. 28 Parliamentary Debates, series 5, vol. 259, col. 1196, 20 Nov. 1931. 29 In fact, John A. Costello, attorney-general of the Irish Free State, had explicitly argued against the inclusion of the convention in statutory form as the title of the crown would give rise to religious illfeeling in the Free State; see notes on meeting of Gwyer Committee, NAI: Taois/ s. 6610 (14 Nov. 1929 both Viscount Caldecote CJ and Singleton J claimed that the Statute of Westminster did not provide either expressly or by implication for secession 33 and questioned whether secession could take effect without the involvement of the other members of the commonwealth. 34 This, of course, tallied with Keith's view of the Statute of Westminster. Notwithstanding these legal objections, the political concession of the power to secede in 1930 meant that a country could, more than likely, secede thereafter. This raised the question: how could one do so?

IV. Methods of Seceding
In the 1920s and 1930s, a number of theoretical means of secession were mooted. These methods varied considerably and gave rise to considerable confusion as to what was, in fact, necessary in order to effect a secession from the British commonwealth of nations. These operated on two planes: public international law and imperial constitutional law. Within these planes, it is important to distinguish between voluntary secession and automatic secession. The former assumed the state concerned wished to secede, while the latter contended that in certain circumstances secession might occur as a consequence of some other action by the state.
At the field of public international law, the first question arose as to whether or not there existed a right of dissolution of treaty obligations between the constituent members of the commonwealth. The relations of the members of the commonwealth between each other suffered, however, from one difficulty: although the members, with the exception of Newfoundland, were also members of the league of nations, their relationship between themselves was not governed by treaties. Instead, it was regulated by imperial constitutional law, and the constitutions of those dominions were based on British statutes. 35 There was one exception: the Irish Free State. The Irish Free State had concluded articles of agreement for a treaty with the United Kingdom and had subsequently lodged this agreement with the league of nations. 36 Moreover, as this treaty specifically made reference to the Free State's membership of the commonwealth, it was open to the Free State to denounce this treaty and ipso facto secede from the commonwealth and empire. This was the view of Irish legal officials. 37 Nonetheless, it faded as a matter of legal importance due to the relative elasticity of imperial constitutional doctrine after the Statute of Westminster.

33
Ibid., 128, 134. 34 Ibid. 35 Their accession to the optional clause of the PCIJ specifically reserved inter se matters. 36 The UK objected to this on the grounds of the inter se doctrine. 37 NAI: Taois s. 6110, memo entitled 'The Constitutional Right of Secession as an Issue at the Coming Conference'. This memorandum owed a great, albeit unacknowledged, deal to the Schlosberg analysis of secession.
The starting point as to the inter-war commonwealth constitution in relation to secession is the Balfour declaration which defined the relationships between the members of the commonwealth: 'They are autonomous Communities within the British Empire, equal in status, in no way subordinate one to another in any aspect of their domestic or external affairs, though united by a common allegiance to the Crown, and freely associated as members of the British Commonwealth of Nations'. 38 The development of inter-commonwealth relations tended to proceed on the basis of the concept of 'equality of status', but the consideration of what was necessary to effect secession focussed on those elements without which a dominion did not meet the minimum standards of 'common allegiance' or 'free association' required to remain a member of the commonwealth. We will deal with different methods in turn: a declaration of neutrality, a declaration of a republic, and the removal of the crown from the internal constitution of the state.
Under the doctrine of the British commonwealth of nations, once one element of the commonwealth went to war, all other members were automatically at war. This doctrine prompted academics commenting on secession to posit that any declaration of neutrality in an armed conflict would ipso jure operate to effect secession. The reasoning is relatively simple: one could not be a member of the commonwealth and not at war; a declaration of neutrality would mean that one was not a member of the commonwealth; therefore the state in question had seceded. As Malcolm Lewis stated: '[i]t would be open to any part of the Empire to pursue a policy of passive as opposed to active belligerency, but a declaration of neutrality would imply a secession'. 39 Although it might be thought that the developments in imperial constitutional law in the 1920s and 1930s would call this doctrine into question, in 1935 E.C.S. Wade was still convinced of the correctness of the 'orthodox' view. 40 A second method of secession was to declare a republic. 41 This would break the bond of allegiance between the king as head of the commonwealth and the dominion concerned, as well as removing the link between citizen and monarch. During the 1930s, de Valera expressed his wish to declare a republic but remain within the commonwealth. Thomas Inskip, the attorney-general, expressly disavowed any possibility of an agreement between the Free State and the United Kingdom which attempted to proceed on this basis. 42 An intermediate position, and one which the Irish Free State tested in the 1930s, was that it might be possible for a dominion to secede by removing the crown from the internal constitutional machinery of the state. This was based on the idea that removal of 'common allegiance' to the crown would put the state outside the commonwealth. The difficulty with this position was that it was never clear where the definitive rupture would occur. So, removing the oath which members of the Irish parliament were required to swear, as de Valera did in 1933, could be seen to provoke such a crisis. However, the Free State constitution still contained a number of monarchical elements at that time, including a governor general. The Canadian consideration of the position of the Irish Free State in 1932 reasoned that these monarchical elements meant the Free State had not seceded; the memorandum which considered the issue concluded 'the oath vanishes, but allegiance remains'. 43 A similar view was taken in London; removal of the oath did not a secession make. Nonetheless, the question of what exactly would give rise to this was not clarified, it was simply noted: 'further action or threatened action by the I.F.S. (including probably the removal of the Governor General), from which the proper inference to be drawn would be that the I.F.S. had definitely seceded from the Commonwealth'. 44 It was unclear what form the 'further action' could take.

V. Secession in the 1930s
The imperial conference of 1930 and subsequent passage of the Statute of Westminster indicated a growing acceptance within the commonwealth of the possibility of secession as a matter of political fact. The inter-war 1930s brought three notable developments in relation to secession: the opportunistic diplomacy of the Irish Free State, the Status of the Union Act in South Africa, and the Government of India Act 1935.
In the Irish Free State, 1932 saw the replacement of the Cumann na nGaedheal government with the Fianna Fáil party under Eamon de Valera. This party was formed primarily of individuals who had fought a civil war rather than accept the constitutional settlement of 1922. Their governmental policy focussed on constitutional symbolism and the need to ensure that the constitution of the state was autochthonous. This was influential across two dimensions in relation to secession. First, each successive development brought with it the question of whether secession had been effected. The Irish position was that they did not want to leave the commonwealth, therefore putting the onus on the other commonwealth members to positively expel them. Second, in 1933 de Valera sought an assurance from the British government that, in the case that the Free State did secede, force would not be brought to bear on the state. This gambit was designed to frustrate the British cabinet. This was for two reasons. First, an admission that secession was possible would reverse Lloyd George's position from 1921 and be a considerable diplomatic coup for de Valera's government. Contrariwise, a declaration that secession would be met with military force would both be met with a furious reaction by the South African government, and potentially put the British government in breach of the Kellogg-Briand pact which prohibited war in the resolution of disputes between signatories. 45 The British government, alive to these controversies, sought to defuse the controversy by refusing to comment on a hypothetical situation. These considerations were also present in the India office, as the secretary of state for India, Samuel Hoare, wished to know what the position was in relation to secession. 46 However, the dominions secretary made it clear that even asking for the law officers' opinion in this regard would either bolster the Irish case, or irritate the South Africans. 47 Accordingly, Hoare agreed not to ask for their formal opinion.
More significantly, perhaps, the constitutional changes by de Valera's government successively challenged the outer boundaries of what one could do and yet remain in the commonwealth. In 1935, the Irish Free State passed a citizenship Act which explicitly prevented the extension of British citizenship to citizens of the Free State. 48 This was a direct challenge to the concept of personal allegiance underlying the 1926 declaration and the response of the British government was simply to treat it as inapplicable outside the territory of the Free State. In 1936, the Free State used the occasion of the abdication crisis to eliminate any references to the king or governor general in the internal constitutional machinery of the Free State. The Free State was content to provide that the king could act as a constitutional 'organ' in certain matters of external affairs under the simultaneously passed External Relations Act 1936. Again, the question arose as to whether the Free State had seceded. Again, the answer given was that the commonwealth was prepared to treat the matter as not changing the relationship between the Free State and the commonwealth. appointed secretary of state for dominion affairshe was the most far-sighted British politician of the time in regard to commonwealth affairs. He was less concerned with the constitutional structure of the empire and more concerned with the means by which nationalist movements could be reconciled to their existence within it; Ireland would provide an example of the complete freedom that a state had within the commonwealth. Accordingly, he advocated that the British government not treat either the abdication crisis or the constitution of 1937 as putting Ireland outside the commonwealth. 49 As we shall see, MacDonald was to make another telling intervention in the aftermath of the second world war.
Prior to the passage of the Statute of Westminster, the legislative competence of the South African parliament to secede from the commonwealth was questionable. We have seen that John Hearne questioned whether such an Act would be within the power of the South African parliament in 1930. More important than that purely legal question, however, was the operation of the signification of royal assent to bills. Kennedy and Schlosberg noted that the powers of reservation and negativity meant that the operation of dominion legislation was subject to overriding control by the British cabinet. 50 In 1934, the South African parliament passed the Status of the Union Act which removed this executive check on their power. Moreover, the Act declared that the union was 'a sovereign, independent State' in the preamble and gave the force of law to the Statute of Westminster in South Africa. 51 These steps allowed Hertzog to claim that the dominion status had progressed to the stage where secession had been implicitly recognized as all institutional structures which would have prevented it had been eliminated.
The position of the British government in relation to secession can further examined in the context of South Asia. In the case of India, progression towards dominion status had been indicated since the Montagu-Chelmsford report in 1919. The Indian national congress party was committed by article one of their constitution to attain 'Poorna Swaraj (Complete Independence) by all legitimate and peaceful means'. 52 However, any agreement that secession was possible for a dominion would in theory free India to choose its own course; a point not lost on British officials. 53 Moreover, such an agreement would seem to en passant raise the question: who should determine the constitutional future of the country? Should a constitution be framed by Indians, or instead by the The question of the legislative competence of the union parliament was to recur in the Harris case, but as that case falls outside the timeline under consideration here, it has been omitted. British government? If it were to be the former and they determined to leave the commonwealth, what could induce them to stay? They would not be in the position of the Irish delegates in 1921-1922, when the compulsory clauses of the 1922 constitution were introduced. The British government were aware of this problem and sought to curb the movement towards independence through legal means in the Government of India Act 1935 which provided for a legislature of prescribed powers which could not repeal any Act of the British parliament without the approval of the governor general 54 or make a law in relation to the sovereign in any circumstance. 55 In the inter-war years, therefore, the doctrine of secession had been implicitly recognized in the imperial conference of 1930 and through the passage of the Statute of Westminster in 1931. However, secession remained implicit and the attempt by the Free State to force the issue in 1933 had been ignored. The actions of the Free State demonstrated the difficulty with establishing how exactly it was to operate within the commonwealth. The position in relation to citizenship or the constitutional response of the Free State to the abdication crisis both provided for a potential breach, but these were ignored. The question in relation to neutrality was to be tested by the outbreak of the second world war.

VI. The Second World War: Neutrality and the Concession of Secession
The idea that neutrality could give rise to an automatic secession from the commonwealth has already been canvassed. It arose as a result of the commonwealth doctrine that a declaration of war by the king in one of the constituent elements of the commonwealth automatically placed the other elements of the commonwealth at war. H.D.J. Bodenstein, secretary of the department of external affairs in Pretoria, disclosed that he believed the South African vote in favour of war was based on the idea that there was 'a moral, or even a legal, obligation to support the United Kingdom whenever she went to war'. 56 The reference to a legal obligation is a little obscure, and it is not clear whether Bodenstein was referring to the commonwealth position or the particular situation that South Africa found itself in as a result of the Simonstown agreement. Schlosberg was clearer: the dominions had the right to remain neutral both by virtue of their co-equality with the United Kingdom and by virtue of their international personality as evidenced by their membership of the league of nations. 57 However, General Smuts certainly believed South Africa was legally obliged to enter the war as a result of allegiance to the crown, as he was to prove in 1939.
In 1937, Ernest Lapointe, Canadian minister for justice, addressed the question of secession in the context of a debate on '[strict] neutrality' in the case of any war. 58 Lapointe pointed out that 'according to all constitutional writers, this would mean the secession of Canada from the commonwealth of nations'. 59 Lapointe wished to preserve the Canadian position to adapt in the light of the practicalities of the case, which he believed would prove important in any decision Canada would be forced to reach; he quoted General Smuts to the effect that 'sufficient unto the day is the evil thereof. Wise men leave these things alone'.
The theory of automatic secession as a result of neutrality was put to the test by the second world war in South Africa and the Irish Free State. In South Africa, the cabinet split on the question of whether or not the country was automatically at war when the rest of the empire was. The prime minister was of the opinion that legally speaking South Africa was not automatically at war, where other ministers in his cabinet believed it would involve 'a breach of the association of the Union with the Commonwealth'. 60 Hertzog's view was defeated in the South African house of assembly and Smuts formed a pro-war coalition. In contrast, Ireland did declare neutrality during the second world war and pursued the policy until its completion. This led to a somewhat confused situation in the aftermath of the second world war. From the point of view of the Irish government in August 1948, Ireland was not a member of the commonwealth, and had not been so for a number of years (although the exact date on which it ceased to be a member was not canvassed in official documents). 61 However, from the point of view of other governments, Ireland was still a member of the commonwealth and had yet to secede. The confusion about Ireland was widespread. 62 Under British law, the matter was only clarified by s. 1(1) of the Ireland Act 1949. 63 The difficulty of clarifying exactly when secession would take place had been foreseen in the late 1920s and early 1930s, but it was only with the accumulation of events that it became clear that there would be no automatic secession on foot of another event - an accretion of instances once labelled constitutionally impossible had become gradually accepted within the imperial constitution.
In the aftermath of the second world war, therefore, the position, as noted by K.C. Wheare, was that the simplest method of secession from the British commonwealth of nations was the declaration of a republic. 64 This final element of secession was to come under strain with the Asian move towards independence after the second world war.

VII. The Practical Concession of Secession: Burma
From a practical, rather than a theoretical, point of view, the most significant advance in establishing the right to secede from the commonwealth was secured by India in the course of the efforts of the second world war. In August 1940, Leo Amery, secretary of state for India, put forward the British government's scheme for a new constitutional settlement for India in the aftermath of the war. This settlement would be as a dominion which would be 'within the British Commonwealth of Nations'. 65 Nicholas Owen demonstrates that it was a fear of losing the war which led to a groundswell of British popular opinion for the grant of immediate dominion status that resulted in the Cripps proposals. 66 The secretary of state for dominion affairs, Clement Atlee, proposed on behalf of the Indian cabinet committee the creation of a 'new Indian Union which shall constitute a Dominion, equal in every respect to the United Kingdom and the other Dominions of the Crown, and free to remain in or to separate itself from the equal partnership of the British Commonwealth of Nations'. 67 This was subsequently amended in the final Cripps proposals to emphasize the 'common allegiance to the Crown' but again the right of the dominion to decide its future relationship to the commonwealth was also guaranteed; this was an implicit guarantee of the right to secede. 68 The drafting process makes it clear, in particular the 2 March draft, that the implicit guarantee was appreciated by the drafters. Although the Cripps mission to India was ultimately unsuccessful in garnering Congress party support for the war effort, it marks a distinct change in terms of the political concession that secession was implicitly acceptable from the point of view of the commonwealth and, more significantly still, that the right extended to India. This change extended to all major political parties within the UK. In June 1945, for example, Amery stated: 'no limit is set to India's freedom to decide herself her own destiny, whether as a free member and partner in the British Commonwealth or even without it'. 69 This principle, once conceded, was never again questioned. Attlee's government after the conclusion of the war proceeded on the same basis: India herself must choose what will be her future Constitution; what will be her position in the world. I hope that the Indian people may elect to remain within the British Commonwealth. I am certain that she will find great advantages in doing so. In these days that demand for complete, isolated nationhood apart from the rest of the world, is really outdated. Unity may come through the United Nations, or through the Commonwealth, but no great nation can stand alone without sharing in what is happening in the world. But if she does so elect, it must be by her own free will. The British Commonwealth and Empire is not bound together by chains of external compulsion. It is a free association of free peoples. If, on the other hand, she elects for independence, in our view she has a right to do so. It will be for us to help to make the transition as smooth and easy as possible. 70 The pronouncement by Attlee was seized upon with particular vigour in 1946 by the Anti-Fascist People's Freedom League (AFPFL) led by Aung San in Burma. 71 The secession question was complicated. A memorandum prepared by U Tin Tut in November 1946 specifically requested a declaration by the British government that secession was consistent with dominion status; however, this paragraph did not feature in the memorandum later submitted under Aung San's signature to the executive council. 72 The league was invited in December 1946 to London for talks about the future of Burma as part of a wider Burmese delegation. In response, the working committee of the AFPFL demanded that there be an acceptance that the talks would proceed on the basis of four main points. The most important from the point of view of this paper were the contentions: '[i]mmediate steps to be taken from now to prepare the way for a free united Burma' and 'a categorical declaration to be made forthwith that Burma would get complete independence within a year'. 73 It is clear from the briefing documents prepared by Lord Pethick-Lawrence that the point was conceded in principle by the British cabinet: The right to secede is implicit (and is well known by the more experienced Burmese politicians to be implicit) in the Dominion status which we have promised Burma (in so many words in 1940); it would be impossible to refuse formally to assure Burma that she would be free to elect independence outside the Empire now that India has had a formal assurance to that effect. 74 The question of whether it would be possible to retain Burma within the commonwealth faltered on the British insistence on dominion status in order to remain a member. In this matter, it was the British insistence on a link with the crown, which had formed the basis of the commonwealth since the Balfour declaration, which made it impossible to reconcile with a republican constitution. Within the British cabinet, the wisdom of holding firm to this policy was questioned, but ultimately the will did not exist in 1947 to push the issue. 75 As in the 1930s, the most far-sighted British official was Malcolm MacDonald, who was then governor general of the Malayan Union and Singapore. In a memorandum composed in June 1947, MacDonald drew on his knowledge of the development of the imperial constitution to outline the opportunity that Burma presented for 'the creation of a British Commonwealth of Nations including nations and peoples of many races, colours and civilizations': Our genius for political government enabled our predecessors in the last generation to transform a large part of the Colonial Empire into a Commonwealth of free and equal Nations. The ruling peoples in these nations, however, are white men, most of them of British stock. The test now is whether we can transform the 'coloured' parts of the Colonial Empire also into a Commonwealth of free and equal Nations. This is obviously a more difficult task. 76 MacDonald advocated the use of Ireland as a model for the development of a new commonwealth model where a republican government recognized the crown as the head of the commonwealth (though not internally) and for certain functions in external affairs. In the subsequent India and Burma committee discussion of the memorandum it became clear that India's willingness to accept dominion status, even for an interim period, meant that the British government would insist upon it in the case of Burma: the British government were prepared to sacrifice Burma from the commonwealth in order to preserve India as a dominion. 77 As the Burmese were not prepared to concede on republican status, they would have to leave the commonwealth, notwithstanding some misguided offers by Britain to send experts to Burma to explain dominion status.

74
Ibid., 'Burma: Constitutional Position' (9 Dec. 1946 The fact that the legal right to secede had been conceded still raised the issue of how secession could be legally effected. In the case of Burma, this was accomplished by the first article of the Burmese constitution of 1947. The treaty signed between Burma and the UK in 1947 explicitly guaranteed the independence of Burma as a republic, in contrast to the agreement between Ireland and the United Kingdom in 1921. The assumption underpinning the Burmese negotiations also largely dictated Ireland's withdrawal from the commonwealth as a result of its wish to declare itself a republic. 78 As we have seen, the position of India had shaped the development of the imperial politics in relation to secession in the 1920s, 1930s and 1940s. It was the threat of secession by India that, in the end, prompted a re-consideration of the nature of the commonwealth itself.

VIII. Secession and the Re-Conceptualization of the Commonwealth
The UK government were keenly aware of the possibility that India could follow Burma into seceding from the commonwealth. The geopolitical importance of India meant that they were determined that this should not occur. We have seen how Atlee's pronouncement in relation to India guaranteed final say to India about membership of the commonwealth. The Indian Independence Act 1947 provided for dominion status and recreated, in section 6 of the Act, the substantive provisions of the Statute of Westminster. 79 The use of the phrase 'independence' in relation to India had been a fraught one, and the British government took a conscious decision in 1946 to ensure that references to Indian 'independence' did not preclude independence within the commonwealth. 80 This made the approach of the British government far different when compared with the corresponding Irish Free State Constitution Act in 1922. Article 1 of that constitution had provided for the Free State's membership of the commonwealth; there was no corresponding provision in the 1947 Indian Independence Act.
Notwithstanding this grant of independence, it became clear that India intended to declare a republic. However, this did not mean that India was committed to exiting the commonwealth. Michael Brecher points out the pre-eminence which Nehru had in foreign affairs. 81 It is important, therefore, 78 See W David McIntyre, '"A Formula may have to be Found": Ireland, India, and the Headship of the Commonwealth', 91 The Round Table ( 2002), 391;D. Coffey, '1916D. Coffey, ' , 1921  to consider Nehru's reasons for staying in the commonwealth. Brecher contends that there were seven: (i) there would be no limitation of India's independence, (ii) the danger of Indian isolation outside the commonwealth, (iii) the commonwealth would prove useful for enhancing the prestige of India, (iv) the promotion of world stability, (v) continuity with the British state, (vi) the promotion of Indian interests in Asia, (vii) the promotion of Indian economic interests and (viii) the protection of Indians in other countries in the commonwealth. 82 A problematic legal area was the possibility of a 'most favoured nation' suit against the United Kingdom. This referred to clauses in international treaties which provided that the signatories thereto were to be accorded any further preferential treatment which might be accorded in a subsequent treaty. Importantly, however, the United Kingdom had treated these clauses as not being effected by commonwealth agreements on the theory that the sovereignty of the crown extended to all areas of the commonwealth and therefore commonwealth countries were not 'foreign' countries to which a most favoured nation suit could be applied. 83 This made the extension of some element of monarchical constitutionalism to India imperative in the eyes of the British negotiators.
The theory of 'latent' or 'dormant' sovereignty in Nehru's ten points appears to have been developed by Sir B.N. Rau in his discussion with Lord William Jowitt, the lord chancellor, in October 1948. 84 Rau's argument was twofold. First, although the king did not enjoy any sovereignty in respect of India, this could be made active by an amendment to the constitution. Second, the extinction of sovereignty over a territory required parliamentary authorization to be effective and this had not occurred in the case of India, in comparison to, for example, Burma. Rau also stressed that this was a legal argument and may not have proven politically acceptable.
This interpretation of the Indian constitution, described as 'as subtle as it is novel', was not accepted by the British law officers. 85 First, they rejected the contention that there was no municipal law extinguishing sovereignty as they pointed out that such a law would have to be passed shortly to regulate the matter. British law distinguished between dominions and 'foreign countries', but admitted of no intermediate state. Second, they considered the nature of the commonwealth relationship might change; and that it might no longer be based upon the concept of allegiance to the crown. However, they believed that it would be necessary to point to 'a continuous and substantial tie between members of the Commonwealth, subsisting after, as well as before, the repudiation of allegiance to His Majesty'. This was problematic as there was not, in the view of the law officers, such a basis. It would, in their opinion, require (a) de facto acceptance by civilized nations of the commonwealth as such a community, (b) a declaration by the prime ministers of the commonwealth of such a relationship and (c) common citizenship. The final point was the most difficult to satisfy. Although commonwealth citizenship conferred real civic benefits in the United Kingdom, 86 it did not do so uniformly across the commonwealth. In light of this, the law officers concluded that the commonwealth would struggle to contradistinguish itself from South American countries, who could make a much more substantiated claim to a shared common heritage.
The attitude of the law officers was viewed with considerable irritation by Nehru and Menon. 87 Nonetheless, they must be taken to express a key idea about the continuity of a legal relationship between commonwealth members based on the crown which was shared also by political leaders, in particular Evatt of Australia. 88 This was evident when the British cabinet took the decision to make the issue a political, rather than legal, one on 12 November 1948, but was immediately thwarted by the insistence of the representatives of Canada, Australia and New Zealand on the retention of some link to the crown. 89 Despite this initial opposition, it was the decision to focus on the political, rather than legal, elements of the decision that was the crucial turning point in the discussions with India. If the legal position was subordinate, it allowed the crafting of a new commonwealth relationship which was cognisant of India's wish to become a republic and yet not secede. This focus on the political realities of the case was carried on into the cabinet discussions in January 1949 and guided British policy thereafter. 90  of the legal monarchical links within the commonwealth was ultimately subordinated to their political wish not to be seen to exclude India from the association if it did not wish to be so excluded. 91 It was ultimately successful in garnering the support of the commonwealth members and resulted in the London declaration of 28 April 1949 wherein India was recognized as a member of the commonwealth and a republic. In legislative terms, the method hit upon in the UK was to pass the India (Consequential Provision) Act 1949 which provided that, unless the contrary was subsequently provided for, any reference to India was to be treated as if India had not declared a republic. 92 In India, the Citizenship Act 1955 recognized commonwealth citizenship in India for citizens of other commonwealth countries. 93 It did not, however, recognize Indians as commonwealth citizens. 94 Moreover, it was clear that citizenship was a subordinate link compared to the wish to be included in the association.
The feared most favoured nation suit failed to materializethere is no record of an official dispute involving the issue of commonwealth membership between 1947 and 1960 under the General Agreement on Tariffs and Trade [GATT]. In fact, the issue became even slightly more complicated on 24 October 1953, when the United Kingdom was allowed to waive the strictures of Article 1, paragraph 4(b) of the GATT agreement for 'items … which have been traditionally admitted free of duty from countries of the commonwealth'. 95 Article 1(4)(b) set the margin of preference on products in respect of duties and charges between the most favoured nation and preferential rates existing on April 10 1947. The UK did this with reference to annex A of the GATT agreement, which exempted a list of countries from the general operation of the GATT, including India, Burma and Ireland. The 1953 agreement simply referred back to annex A, without any acknowledgement that the constitutional relationship between the countries had, in the meantime, changed with the declaration of republics in Burma, Ireland and India.

IX. Conclusion
The history of secession as a legal concept also makes clearer the contingent nature of legal and political analysis. It was the advent of the second world war, and specifically the early British losses in that war, which was ultimately to drive the key changes in approach by the British government to the question. In the absence of these changes, it is at least questionable whether it would have proceeded on the same timetable. The history of secession is a product of historically contingent events, as opposed to the inevitable progression towards more enlightened sunlit uplands.
The commonwealth experience in relation to secession was noteworthy for a number of different reasons. First, it demonstrates clearly how commonwealth law was subordinated to politics on a number of occasions. The commonwealth members showed a willingness to undermine legal doctrine in order to avoid results that were politically problematic. This was the case with the Irish Free State in the 1930s and subsequently with India in the 1940s. Second, the doctrine was at the core of the development of the institution from the post-1918 grouping of countries on the basis of common allegiance to the crown to the post-1945 grouping as a free association of states with the crown as the head of that association. Third, the experience with the doctrine also demonstrates the difficulty with maintaining a hard line in relation to international associations based on consensus. The negotiations involving the declaration of an Indian republic demonstrated the problem of co-ordination amongst an organization with diffuse interests and the manner in which a country with a strong commitment to an ideal could be relatively successful against such a grouping. Fourth, it was noteworthy that when Britain dropped its typical constitutional pliability that it was unsuccessful. The insistence on dominion status as a sine qua non of membership of the commonwealth in 1947 meant that Burma and Ireland were lost to the organization, before the point was conceded in the case of India shortly thereafter. While the initial reaction of the settler dominions to the possibility of a change may be seen as a vindication of the British government's reticence on the issue in relation to Burma, their eventual acceptance of the re-casting of the organization makes it difficult to avoid the conclusion that the insistence on dominion status in 1947 was misguided. It is notable that Malcolm MacDonald was more prescient than his colleagues on this matter. Once the rigidity of 1947 was abandoned, the commonwealth was free to renew its commitment to constitutional suppleness with characteristic élan.