Stimulation, Segregation and Scandal: Geographies of Prostitution Regulation in British India, between Registration (1888) and Suppression (1923)*

Abstract This paper explores the regulation of prostitution in colonial India between the abolition of the Indian Contagious Diseases Act in 1888 and the passing of the first Suppression of Immoral Traffic Act in 1923. It challenges the commonly held assumption that prostitutes naturally segregated themselves in Indian cities, and shows that this was a policy advocated by the Government of India. The object was to prevent the military visiting these segregated areas, in the absence of effective Cantonment Regulations for registering, inspecting, and treating prostitutes. The central government stimulated provincial segregation through expressing its desires via demi-official memoranda and confidential correspondence, to which Rangoon and Bombay responded most willingly. The second half of the paper explores the conditions, in both India and Ceylon, that made these segregated areas into scandalous sites in the early twentieth century. It situates the brothel amongst changing beliefs that they: increased rather than decreased incidents of homosexuality; stimulated trafficking in women and children; and encouraged the spread of scandalous white prostitutes ‘up-country’, beyond their tolerated location in coastal cosmopolitan ports. Taken alongside demands that the state support social reform in the early twentieth century, segregation provided the tipping point for the shift towards suppression from 1917 onwards. It also illustrates the scalar shifts in which central-local relations, and relations between provinces, in government were being negotiated in advance of the dyarchy system formalized in 1919.


Introduction
The records of the later ages indicate that public women still occupied a recognized position in the social economy, and further that the system of setting apart definite quarters of the city or town for their occupation had become stereotyped. This arrangement, which persists in India to the present day, was unquestionably followed in the Mauryan age. . .. 1 The isolation and autonomy of regulated places was always [therefore] a convenient imperial fiction. 2 The views of S. M. Edwardes, published in 1924 after years of experience in the Bombay police force, represented a historically common, and persistent, consensus regarding the segregation of Indian common (not military) prostitutes. He reproduced a broader belief that Indian classes and communities resided together '. . .in their own quarters'. Regarding prostitutes specifically, the existence of urban red-light districts in the early 1900s was simply believed to be a legacy of preceding '. . .Hindu and Moslem governments'. 3 This paper explores a hidden history of segregation that exposes the scripting of Edwardes' views by the broader imperial fiction explored in Howell's recent work. 4 It reveals the role of state-induced regulation in the supposedly historical tradition of 'setting apart' prostitutes. Whilst accepting that Indian society did self-segregate to an extent, 5 it will be shown that this segregation was not voluntary, but was fundamentally tied to the fate of regulated and registered cantonment prostitutes. As against the comparative lack of work on prostitutes in Indian civilian spaces, the late-Victorian efforts to regulate military prostitution have been the subject of extensive study and speculation. This paper will show that military concerns came to influence policing and governmental policies in civilian space through the powers of suggestion and stimulation, long after the repeal of the Indian Contagious Diseases Act and the revision of the Cantonment Regulations had ended the explicit and official regulation of prostitution outside of military camps. Repeatedly it was stressed that military authority could not be exerted over urban populations directly, although its influence was felt firmly within municipal and policing policy by the turn of the twentieth century.
The Raj's approach to the problem of prostitution can, without too great a risk of over-simplification, be summarized as an attempt to protect the military from the ravages of venereal disease. The uses of the Cantonment Regulations and the Indian Contagious Diseases Act have been commented on from the perspectives of 'race, sex and class', 6 constitutional crisis, 7 venereal disease, 8 repealist campaigns, 9 international governmentalities, 10 spaces of containment, 11 and of resistance by the prostitutes themselves. 12 Certain cities and regions have also been studied to highlight the regional specificities of the ways in which different laws, traditions, and modes of government were imbricated to tackle prostitution and venereal disease in, for instance, Madras, 13 Bengal, 14 Bombay, 15 and Delhi. 16 These studies fail, however, to present clear links between the end of registration and the emergence of twentieth-century abolitionist mechanisms for tackling brothels and trafficking in India. This paper will, therefore, open with a review of the debates surrounding the the deleterious effects of western modernity on upland India. Taken with the growing concern for child brides and the social welfare of the Indian population, by the outbreak of the First World War, India was at a tipping point that would usher in the abolitionist movements of the interwar, late-colonial period. 18 Re-reading the 1890s: cantonment regulations as 'spatial crisis' In her re-reading of the 1890s debate about India's Cantonment Regulations, Philippa Levine charted out a 'constitutional crisis' that was provoked between Britain and British India, testing the limits of sovereign authority between London and Calcutta's offices of state. 19 Revisiting these debates, this paper will move not 'out' to coreperiphery diplomatic networks, but will focus 'in' on the local and very spatial crisis that the Cantonment Act debate provoked regarding prostitutes who had been ejected from military compounds.
The Cantonment Act (XXII of 1864) was passed in response to the Royal Commission into the Sanitary State of the Army in India (1863), which had documented the high levels of venereal disease amongst British soldiers in India. 20 Larger numbers of British military were arriving in India to increase their ratio to Indian troops, following the uprising of 1857, and they needed to be kept healthy to guarantee the security of the Raj. The Cantonment Act provided for the registration and inspection of prostitutes who served British soldiers. Ballhatchet has shown that these women were registered within the cantonment, examined fortnightly, and detained in a lock hospital if infected. 21 This act was, however, confined within a four-mile radius of the cantonment, prompting consideration within a few years of an Indian version of the British Contagious Diseases Acts for cities without major cantonments. A Home Department circular was issued on 24 July 1867 to the governments of Burma, Bombay, Madras and Bengal to 18 For a discussion and deployment of Malcolm Gladwell's concept of 'tipping point' see Sinha, M. (2006). Specters of Mother India: the global restructuring of an Empire Duke University Press, Durham. p. 1. 19 Levine. Rereading the 1890s, 20 Estimated at an average of 25 per cent over the previous 18 years in Madras solicit opinion on the prevalence of syphilitic disease amongst sailors at Indian ports and the best means to prevent it. 22 The consensus was for legislation, but the resulting correspondence revealed a variety of opinions regarding segregation. An analysis of this discussion provides a route into considering the origins of segregation as a spatial solution to the problem of prostitution, one that was subjacent to the more commonly acknowledged geographies of registration associated with the Contagious Diseases Acts. Whilst the circulated summary of the British Contagious Diseases Act of 1866 had not mentioned segregation, the first reply came from the Chief Commissioner of Burma who claimed that he had established a prostitute quarter in Rangoon on the advice of his Health Officer. The comments from Bengal also stated that: 'It may be noted that the segregation of prostitutes into a separate quarter is quite in accordance with Native ideas and usages'. 23 In general, however, the tone of comment was against segregation. A note from Bengal argued that localizing prostitutes in a particular quarter would produce greater '. . .scandals and evils' than the present flaunting of vice in Calcutta. A submitted note by a Dr T. Farquhar suggested that while in England the prostitute was in an outcast state, in India '. . .the condition of the women is far less degraded, and their influence in the community often considerable, so that unwise interference would be frequently resented even by the influential classes of the people'. 24 This reflects the extent to which the term 'prostitute' was still making its awkward transition from Europe to India. Farquhar's comments would have been referring to what came to be distinguished as 'courtesans', who would later be categorized with prostitutes, though distinguished from the 'common prostitute' that flourished in the urban forms created by the colonial economy. 25 The Indian Contagious Diseases Act (XIV of 1868) did not, when passed, contain provisions for segregation, but did allow the compulsory registration of brothels and prostitutes, periodical medical examinations, and compulsory treatment of prostitutes found to be infected in the areas to which it was extended. Despite the dauntingly 22  oppressive potential of the Act, the archive is replete with instances of subversion and concern over the ability of the government to control these women. By 19 August 1868 the Sanitary Commissioner to the Punjab was already complaining about the ability of cantonment prostitutes to leave that space of regulation and settle beyond its borders. He claimed that experience '. . .has already shown that the women to avoid supervision removed to the bazars and hamlets immediately outside the Cantonment boundary, where they were visited by the soldiery, or made it a practice of coming within the boundary at night to receive their visits'. 26 It is this mobility of both women and men, although only the former shouldered the blame, which would later prompt a turn towards segregation as against the registration of the Indian Contagious Diseases and Cantonment Acts.
Ballhatchet has detailed the many failings of the Indian Contagious Diseases Act: their expense; their contested claims to have lowered the rates of venereal disease; the international campaign for their repeal; and the ability of prostitutes to evade registration and inspection. 27 On 27 August 1877 the Officiating Sanitary Commissioner within the Government of India submitted his report on the working of the Lock Hospitals. 28 He urged the detection and punishment of unlicensed prostitutes who dwelled in villages or towns close to cantonments. Civil authorities had apparently reported that there was nothing to be done, besides arbitrary measures which the government could never legalize. The suggestion that the area to which the rules applied should be expanded to cover towns and villages, and to register all prostitutes (and even all women suspected of intercourse with Europeans) was believed to be impossible.
In 1888 the Indian Contagious Diseases Act was withdrawn but a new Cantonment Act (XIII of 1889) was erected in its place, allowing the expulsion from a cantonment of anyone refusing treatment for a contagious disease. 29 Debates about the coercive elements of the latter led to revisions in 1895, 1897 and 1899. These regulations prohibited prostitutes from residing in regimental, but not central, bazars and did allow the inspection of women under suspicion of disease, leading to 26  These twentieth-century forms of regulation were by-products of the late-Victorian debacle over registration. What the Cantonment Act and the furore surrounding it created was a less and less amenable environment for the prostitute within the cantonment, just as urban expansion, social change and commoditized economies began to accelerate across India at the turn of the century. The harassment of the prostitute had driven her from the cantonment, but it has also driven her from the historiography of twentieth-century Indiaher ejection from military space marking her textual exclusion from our archives and research. Following the prostitute into the galis and bazars of the new century necessitates narratives not just of her changing culture 32 and domestic arrangements, 33 but also of the regulations that sought to bring her back within the purview of the colonial state.

Required cooperation: 1899-1913
The opening years of the new century would see two specifically geographical developments in the nature of prostitution and its regulation in India. The first was the definition, and thus creation, of prostitution as an urban and civic problem, rather than as a problem mainly for the military authorities. This resulted from the spatial accumulation of prostitutes in cities and an emergent governmental concern for the health of the broader population, not just of military enclaves. The second development was scalar rather than territorial and saw the central government attempt to sway the politics of provincial governments without centrally legislating itself. Both of these developments centred on the policy of segregation, which emerged as the immediately practicable solution to the urban threat of the prostitute, as propagated by the central government. 34 Whilst epidemiological evidence was proving that the segregation of European populations was ineffective in India, 35 the British clung on to an 'enclavist' mentality into the late 1800s. 36 However, although the administering authority had changed, it was clear that the health of the military was still the key concern of the government. To protect the armed forces the central government attempted to string together the various parts of its domain into a coherent and regulated whole. Its attempts, and failures, to do this highlight the ways in which 'British India' as a territory was itself constantly in a state of construction, competition, and contestation.

From cantonments to municipalities
Four years after the introduction of the 1889 Cantonment Act, and the power to expel prostitutes from the cantonments that it introduced, the Secretary of State for India (the Earl of Kimberley) in London and the Viceroy (the Marquess of Lansdowne) entered into correspondence regarding the rates of venereal disease in British India. 37 Between 1893-1895 this discussion led to recommendations that the police be granted greater powers of supervision over prostitutes, especially those 'loitering' near, and attempting to 'importune' men within, cantonments, although they came to nothing due to fears that the native constabulary would use these powers to blackmail women. 38 This left untouched, however, the problem of prostitutes in towns and cities abutting cantonments, which availed these women not just with physical lodgings, but supposedly with the rights and benefits of civic life.
Between 1897-1898 the government had considered again the impossibility of extending the rules for the prevention of contagious 34 On the shift from viewing prostitution as a problem of surveillance, as in registered and segregated spaces, to one of security and apparatuses throughout broader society, see Howell, P. . . .the mere removal from the precincts of a Cantonment of persons suffering from venereal diseases will be of little avail unless they are prevented from settling elsewhere with the certain result of spreading the disease over a wider and less regulated area. As a first step they suggested that it should be urged on Local Governments that Municipalities and other such bodies should be empowered, if necessary, and required to co-operate actively with Cantonment authorities in their endeavours to check the spread of infectious or contagious disorders, having in view the grave responsibility which attached to every civilized community in a matter so vital to the welfare of the country at large. 40 This statement is fascinating in several respects. First, it extends the legitimate remit of governmental concern with the life of the prostitute beyond her expulsion from the cantonment and into her life beyond it. Secondly, it gives direct evidence of the way in which the central government attempted to manipulate the decisions of local and municipal bodies, whose decisions would later be depicted as beyond the provenance of central government. The pre-dyarchy combination of democratization of rights and government at the turn of the century and the continued sovereignty of the central government is perfectly captured in the notion of being required to co-operate actively. Emphasizing the ongoing influence of the military within broader government, the phrase itself originated in an 1897 note from the Quartermaster General's Department in response to questions about the expulsion of prostitutes from cantonments. Lastly, the statement marks the beginning of the recognition that future policies would move beyond disciplining and curing the military and prostitutes into health and obedience, to a consideration of the biopolitics (or 'broader welfare') of the population through the mechanisms of civil society (or 'civilized community') and preventative medicine. This is in stark contrast to previous and enduring policies and opinions which focused solely on the military. For instance, in response to the suggestion of powers to expel prostitutes from a ten-mile radius around cantonments in the Punjab, P. J. Maitland commented on 6 May 1897 of diseased prostitutes that: Women removed thus far would no doubt be able in one sense to spread the disease over a wider and less regulated area, but on the other hand, they would be practically out of reach of the soldiers, and would be unable to practice their profession until again admitted into a cantonment. 41 This disinterest in the health of the woman herself, or of the population beyond the cantonment, would slowly be broken down, not just by growing pressure for the government to prove its commitment to safeguarding the health of the population, but by the realization that it was almost impossible to protect the soldiery without also regulating the health of the broader public.
The ability to extend the act was stressed to depend upon the geography of the areas adjoining the cantonment, whether rural or urban. The government saw no problem in extending cantonment rules to villages outside cantonments, to target women who brought supplies into the cantonment and supposedly seduced soldiers. More caution was required when extending the rules to populous urban municipal areas because a commitment to treat 'contagious diseases' would include cholera, leprosy and other 'infectious disorders', while the military police who enforced cantonment rules could threaten civic liberties. These suggestions were circulated to the provincial authorities for comment and the replies were considered in October 1899. 42 The most general response concerned the geographical specificity of each case, regarding the circumstances of each cantonment, the character of villages and municipalities, and the difficulties of laying down general rules. It was widely agreed that the act could be extended to villages around rural cantonments, as long as a District Magistrate, not the Commanding Officer, executed the orders outside the military zone.
The comments on the urban areas repeated the concerns that no cantonment authority should have power over a municipal district, nor should British military police be responsible for arresting suspects in a town or city. However, as in 1867 when debating the extent of the Indian Contagious Diseases Act, segregation was once again suggested as a solution to the problem of the clandestine prostitute. The reply from the North-West Provinces showed that the provincial government had introduced clauses into its Municipal Bill to prevent solicitation and to compel prostitutes to live in specific localities (the bill became law in 1900). This particular province had long adopted an aggressive stance in pushing for extra legislative powers to tackle unlicensed prostitutes. During the 1877 debate on the working of the lock hospitals the notes from the North-West Provinces and Oudh suggested that low caste women soliciting near cantonments '. . .can only be punished by a fine, which they are too poor to pay, or by a week's simple imprisonment, which they do not mind'. 43 Under the existing laws it had been proven that the women could not be made to attend inspection because they were too numerous, scattered, or were protected by marriages, occupation or doubtful identification; even the use of the special police and spies had failed. The recommendations included not only increased penalties for unlicensed prostitutes but also the prohibition of soldiers visiting city bazars without a pass. One can detect here the genesis of the segregation system that would forbid soldiers from visiting prostitutes, who would be forced to reside in one particular area. The Lieutenant Governor of the North-Western Provinces commented in 1899 that: Meanwhile the local governments should examine their Municipal Acts and take the power where the power does not already exist of compelling prostitutes to live in certain specified localities. This power will be very useful, apart from this particular connection, and will also, it is believed, be consonant with native feeling. It will enable the Magistrates to free the quarters of towns nearest to cantonments from the presence of prostitutes. 44 What a marginal note on the draft North-West Provinces bill also highlights is that the revisions were inspired directly by the Cantonment Code. 45 Section 175 of the Bill (soliciting for the purposes of prostitution) was annotated 'cf Cantonment Code section 209', which had addressed 'Loitering or importuning for sexual immorality'. 46 Section 176 (regarding brothel closure) was annotated 'cf 204/XX', which allowed the military authorities to exclude from the cantonment persons refusing to attend, or remain in, a hospital or dispensary, having been ordered there under section 203 as a person suffering from an 'infection or contagious disorder'.
What is explicitly stated here, yet remained implicit in the segregationist policies which spread through India in the first decade of the twentieth century, was that the segregation of prostitutes originated in attempts to stop soldiers visiting them after the 43  Thus, whilst unwilling to centrally legislate on this matter, the central government was keen to deploy the power of segregation through the power of suggestion, as in the following comment that was circulated as a concluding point to the 1899 debate: 'The suggestion of the North-Western Provinces Government that power should be taken to compel prostitutes to live in localities not adjacent to the cantonments is of special importance, and may be commended to other local Governments'. 47 This was the basis for a letter circulated to provincial governments on 19 September 1899 stating that: The Government of India think it very desirable that [,] where the health of the troops is affected by prostitutes living in Municipalities adjoining Cantonments, an endeavour should be made to abate the evil by enforcing proper measures under Municipal rule, power to do this being taken, where necessary, by an amendment of the Municipal Acts. The Government of the North-Western Provinces and Oudh has shown the way by introducing clauses into its Municipal Bill to prevent loitering and the keeping of disorderly houses to the annoyance of respectable inhabitants. 48 Accompanying this 'desire', the central government circulated 'extract[s] from municipal laws regarding prostitution, disorderly persons, etc', which varied markedly. 49 In 1899 the Central Provinces could merely order a person not to be a public nuisance, 50 whilst Madras only had the power to make bye-laws to secure order in the streets. 51 Both Punjab and Burma could order the closure of a brothel on the complaint of three or more inhabitants, with the latter being able to declare what constituted a 'main street' on which the annoyance of any person could be penalized. 52  prosecute public nuisances there was no specific law for Calcutta. 53 There was similarly a return of 'nil' for Bombay, both Presidency and city.
The Home Department had thus given their suggestion as to the mechanism by which prostitution should be dealt with in urban areas, and circulated extracts indicating that those mechanisms in existence were limited in their extent. The pressure to remedy these deficiencies was increased three months later when, in December 1899, the Secretary of State for India wrote to the Government of India desiring to be informed of the steps taken to suppress contagious and infectious (i.e. venereal) diseases in the Presidency towns (Calcutta, Madras and Bombay). 54 The central government asked its Principal Medical Officer what measures could be adopted to check prostitution, who replied that the only way to prevent the spread of venereal diseases was to place the areas of cities which were frequented or inhabited by prostitutes out of bounds to soldiers, and to organize an efficient system of military police to keep women from public solicitation. However, these measures were said to be impossible in Presidency towns, for three main reasons: 1. The localities of prostitutes are unknown in these towns, thus placing them out of bounds would practically amount to confining men to their barracks, which is impossible.  Burma branch of the British Medical Association calling for measures against public prostitution, the Home Department replied that it was already 'stimulating municipal legislation' to deal with solicitation and brothels. 55 The Police Act in Bombay and the actions in the Burmese Presidency town of Rangoon, responding to central suggestions that both municipal and presidency authorities tackle prostitution, would establish the two models by which segregation would take root in India in the opening years of the new century. They would also, however, become the two scandalous sites that would do most to discredit segregation, after the murder of a vulnerable young woman detained in Bombay's red light district, and a campaign highlighted the immorality of Rangoon's centrally located segregated brothel zone. But these two cities, and the provincial governments based there, would then provide the experimental sites in which alternative powers of regulation would be devised in the post-war years of brothel and trafficking suppression.

The morality of the municipality: the Rangoon Police Act (1902)
The Government of Burma was the first to respond to the central government's suggestion that prostitution be dealt with via amendments to municipal rules. 56 On 12 July 1901 it wrote to the central government in Calcutta forwarding proposed amendments to the Rangoon Police Act (Burma IV of 1899) and the Burma Municipal Act (III of 1898) along the lines of the clauses included in the North-Western Provinces and Oudh Municipalities Act (I of 1900). 57 Two separate bills were required following the reorganization of the Rangoon Police into an independent force, as in other Presidency Towns, in 1899. 58 While the outcome of these amendments would prove to be scandalous in their own right, there had been previous requests to take similar action which had, themselves, mobilized the power of the open prostitution and solicitation in the city, which was said to be to the detriment of public health and morals, and to the safety of the women who were bought and sold for 'immoral purposes'. 59 The 1104 memorialists sought the suppression of prostitution '. . .or at the very least that the said evils may be confined within the narrowest possible limits'. The President commented that the location of the 'evil' was in the centre of town, and that every class, office and religion was represented in the memorial. He admitted the impossibility of suppressing all prostitution, but agreed that it could be confined. However, five years after the decline of the Indian Contagious Diseases Act how, he asked, could one avoid the stigma of 'state regulation of vice'? It would be 'scandalously wrong' if the municipality could not prevent brothels from opening near to girls' schools.
After investigating the issue in December 1893, the municipal committee found that in 1884 there had been unsuccessful attempts to confine prostitutes to one part of the port using the Indian Contagious Diseases Act. An appendix listed 107 brothels, detailing their address, owner's name and nationality, and classed the prostitutes by race and nationality. Of the listed occupants there were 58 Burmese, 53 natives of India, six Chinese, three Japanese and three 'Asiatic', two European, with one separately listed woman from Spain, and one Shan, an ethnic Buddhist of Burma. A map was also produced by the municipal committee which picked out the brothels in lusciously alarming red, highlighting their proximity to schools (the numbered circles) and civic institutions ( Figure 1). Anecdotes were also used to prove that '. . .the behaviour of the women has long been an open scandal'. The Punjab Municipal Act (1891), which had been circulated by central government in 1899, was cited as a way of overcoming the burdensome need for evidence of promiscuity demanded by the Indian Penal Code.
The drafted bill did not, however, make it into law. The Deputy Commissioner of Rangoon insisted, in April 1894, that it would be impolitic to legislate against solicitation as the remedy would be worse than the problem, putting into the hands of the police an 'engine of oppression' against respectable women. On this basis the Chief Commission of Burma rejected the proposed amendment to the Lower Burma Municipal Act (1884). Similar complaints about the danger posed to the morals of school girls were made in 1896. 60 But it was not until the clear signal of support from central government that the Rangoon authorities instituted their segregated system.
The 1901 bills were explicitly acknowledged to be a response to the attempts to diminish the evil of prostitution in cantonments, which had created difficulties in adjoining municipal areas. In the interest of public decency and the health of the troops the powers of the municipality would be extended from the main streets, and from cases of annoyance, to the whole municipality and to decisions made by the local government condemning brothels, loitering or soliciting. The bill would give the Commissioner of Police power to prohibit the residence of prostitutes to within specific areas. Reaffirming the twoway surveillance that was the key motive behind segregation, the note continued: It is hoped that, by means of these provisions, prostitutes, or at least the more objectionable of them, will be removed from the immediate neighbourhood of military barracks, and that the parts of the town in which they eventually congregate may be successfully placed out of bounds for the troops'. 61 The Burma Municipal Bill extended similar powers to the rest of the province. The specific mechanism for segregation consisted of three additional sections in the Rangoon Police Act: 43A: Penalty for continuing to keep a disorderly house: on the complaint of three or more inhabitants that a house is used as a brothel, or by disorderly persons, to the general annoyance of inhabitants of the vicinity or persons using any main street, the discontinuance of this use of the house could be ordered within five days. The local Government could declare by notification which streets would be deemed main streets.
43B: Power to prohibit brothels and the residence of prostitutes in certain areas: on one complaint or police report that a house is being used as a brothel or lodging place for more than one prostitute within an area of Rangoon Town as decided by the Police Commissioner could be given one months notice to vacate, and forbid any prostitute thereby evicted to dwell in the stated area.
43C: Penalty for soliciting or loitering for purposes of prostitution: Whoever, in any street, thoroughfare or place of public resort, loiters or solicits for purpose of immorality or commission of immorality respectively, could be fined or imprisoned.
The selection of the areas to which these sections were applied after 1902 created three segregated zones in the heart of Rangoon. Section 43A was applied to all of Rangoon, such that any brothel or prostitute could be moved on after three complaints. The more drastic powers of 43B were applied to an area delimited by Montgomery Street to the north, Thompson Street to the east, the Rangoon River to the south, and Godwin Road to the west (see Figure 2). This grid pattern section of town effectively constituted the European and governmental quarter of the city, and the powers effectively allowed for brothels and prostitutes to be removed from public view in this area. However, within this zone five areas were excepted from 43B, being the traditional prostitute quarters of town (the two major ones correlating to clusters of brothels identified in 1894, see Figure 1). E. Shuttleworth, District Superintendent of Police in Rangoon, would refer in November 1917 to these areas as being close to Sulé Pagoda Road, the 'exact centre of Rangoon'; 62 27 th -29 th streets lay to its west, 33 rd -34 th street just to its east. Although the powers that the acts would bestow were clear, the geography of their enactment was not mentioned in the legislation, disguising the extent to which a segregationist policy was central to the new laws. This became clear in practice, and was suspected by many during the debate within government over whether the Act should be passed.
Having been received in Calcutta, the proposals were passed through the relevant sections of government for vetting, where members of the Home Department quickly noticed that the proposed legislation would effectively give legal sanction to a segregationist policy. This policy had long been favoured in Burma, dating back to the Chief Commissioner's claim in 1867 that a prostitute's quarter had already been established in Rangoon. As with that debate, there were voices within government both for and against segregation. On 16 August 1901 an internal comment by A. Williams, Deputy Secretary of the Home Department, provided an insightful summary of the logic of the proposed amendments, claiming that they: . . .seem to me to go considerably beyond the provisions in force in existing Acts by permitting certain authorities to appoint local areas in which prostitutes are not to dwell, the principal object being to exclude them from municipal areas which abut on cantonments. But it is clear that the provisions can be so worked as by a series of excluding orders to constitute a prostitutes' quarter or 'chaklah', the desirability of which seems very questionable. 63 This legislation was stressed to go much further than existing laws, which allowed certain streets to be declared out of bounds for disorderly houses. The amendments also found their backers, however. F. Cowie of the Home Department commented on 3 September 1901 that the suggestion of fines of Rs. 25 for every day after which an order to discontinue a disorderly house was contravened was more suitable than the Bombay Police Act, which was the other legislative innovation being worked through at the time and which imprisoned offenders for short periods. Cowie then went on to show that these amendments, as with those from the North-Western Provinces and Oudh, were very similar to those of the Cantonment Code. Cowie argued that these provisions had been fully debated when the new Cantonments Act was passed in 1899, and that the approving committee had commented that: 'It is possible that the setting apart of specified localities in a cantonment as the residence of prostitutes might be held to give legal sanction to the practice of prostitution' but the rule was allowed to stand. It had also been commented, however, that a more arbitrary regime was permissible in a cantonment than in a municipality. Cowie concluded, rather weakly, that: I do not see that the draft sections now under consideration need necessarily result in the establishment of a prostitutes' quarter, but they might be used in a way which would tend in that direction-or to augment an existing one. The chief object perhaps is not the morality of the municipality but the health of troops, and it would perhaps be desirable to make it plain that the only 'local areas' to be declared to come under the operation of sections 43B or 181A should be such as adjoin cantonment areas. 64 This as good as gave official sanction to a provincial government that had already declared itself willingly segregationist. The covering letter from the Chief Secretary to the Government of Burma that proposed the amendments admitted that the committee amending the Burma Municipal Bill in 1897 had been in favour of segregating prostitutes, but the Lieutenant Governor had been doubtful that it would have been acceptable to central government at that time. The statement above also reconfirmed the central government's sole concern with military, not public, health. Deputy Secretary Williams responded on 6 September that he felt it best that the power to exclude prostitutes be restricted to areas abutting cantonments, leaving evictions further afield to rely upon initiation by neighbours of disorderly houses. This recommendation was made not on the grounds of expediency, but of public relations in the face of ongoing purity campaigns against toleration of brothels: Administratively this may not be the most convenient route; this would be to get the sore to a head and confine it to a particular locality. But the present tendency is to regard any sort of regulation as tantamount to sanction of prostitution, and I think it desirable, therefore, to interfere with it as little as possible. 65 But opinion was clearly against Williams within the Home Department. For instance, J. P. Hewett added to the internal debate, on 8 September, arguing that: 'I do not see how it can be argued with any show of reason that because you give power to exclude prostitutes from particular parts of a town you are encouraging prostitution by driving them to congregate in other parts of it'. 66  amendments on 16 September 1901, creating a dense concentration of prostitutes in the centre of Rangoon after the Act was passed in 1902. This segregation would ultimately be of monumental consequence for the way in which prostitution was regulated in interwar India.
Questioned segregation: diffusing the Bombay Police Act (1902) The same year, but 1,540 miles from Rangoon, the City of Bombay Police Act (IV of 1902) was passed. 67 It addressed various elements of policing in the city, but clause 28 specifically provided the 'power to segregate prostitutes and keepers of brothels'. 68 The act listed the inspiration for this clause as having come from previous Bombay and Madras Police Acts but there were also explicit links to previous legislation which had targeted prostitution. Just as the Cantonment Code had inspired the North-Western Provinces and Oudh Municipalities Act (I of 1900) amongst other citations from All-India and provincial acts, clause 28 of the Bombay Police Act cited sections 2 and 20 of the Indian Contagious Diseases Act. 69 This clause gave the Commissioner of Police the power to target the managers or occupants of any residence used for the business of a common prostitute forbidding them after seven days from residing in, using, or frequenting any street or place specified in the notice. Clause 120 provided even more extensive powers for combating 'soliciting for purposes of prostitution and indecent exposure of person'. Anybody within, or within sight of, public space who by words, gestures or otherwise attracted attention for the purposes of prostitution, or solicited or molested a person for the purpose of prostitution, or wilfully and indecently exposed their person, could be punished with imprisonment for up to eight days, or a fine of up to Rs. 50. Anybody refusing to pay could be imprisoned for a month for a first offence, and for up to three months for subsequent offences. The powers had the potential to be used towards segregation, and this is how they were put into effect, with Grant Road near the Railway Station becoming the main locus for prostitutes in the city. Tambe has shown that this Kamathipura area had previously been a centre for marginal castes and trades and that prostitutes had started to congregate there, but this conglomeration was fortified and rigidified by the police act. 70 Commenting on the Act more broadly, Prashant Kidambi has shown that it '. . .extended police jurisdiction over a range of activities and sites that had hitherto been outside its purview and concentrated enormous discretionary powers in the hands of the Commissioner and his deputies'. 71 This was obviously to the liking of central government, which was still under pressure from the Secretary of State to reduce rates of venereal disease in the Presidency towns. It had circulated the Bombay Police Bill to Madras and Bengal in August 1901, but had found those Presidencies unwilling to introduce further legislation. With the bill having been passed in 1902 the Home Department reconsidered their objections between January and March 1903. 72 It agreed that Madras had similar counterparts to the clauses passed by Bombay in its own law, and also that it did not have the same 'evils'. However, the objections of Bengal were treated with greater scepticism. These were that: (1) such powers would drive prostitutes from one part of the city to another, a responsibility to which the Commissioner of Police was opposed; (2) it was felt to be unsafe to entrust constables with the power to arrest women, because they could use these powers for blackmail. A note on file by the Home Department suggested, in response: 'As regards (1) the effect will no doubt be as stated, but there appear to be certain advantages arising out of segregation'. 73  powers, but that the police Commissioner was resisting because he would have to enforce them. 75 However, on 3 January 1903 the Chief Secretary to the Government of Bengal wrote to the Home Department of the Government of India in defence of Calcutta's Commissioner of Police. It was argued that the law in Calcutta (Bengal Act IV of 1866) allowed for the closure of brothels, on the basis of complaint or the decision of the executive, and to act against solicitation on complaint. Whilst appreciating elements of the Bombay Act, it was not desired that the Police Commissioner be burdened with the responsibility of relocating the prostitutes, while the municipality had failed to suggest principles of guidance by which such powers could be enforced.
A final attempt was made to sway Calcutta in the direction of segregation by the Home Department later that year. 76 Madras and Bombay had shown that they were having no trouble in enforcing the relevant sections of the police acts, especially those that tackled soliciting through allowing constables to arrest on initiative without any complaint from the aggrieved. The Deputy Secretary wrote to the Chief Secretary of Bengal on 1 July stressing again that the Government of India did not see the difficulty of segregating to a great degree the prostitutes of the southern part of the town in localities that could then be placed out of bounds to soldiers. The power to arrest for solicitation without a complaint being made was also stressed to be a most useful 'repressive measure'.
Yet again, however, the suggestions of the central government were rebuked. Replying on 15 September 1903 the Chief Secretary to the Government of Bengal, C. Macpherson, defended the resistance of the Commissioner of Police to segregation on the grounds that there were well known rules for the troops in Calcutta which practically placed out of bounds the localities where prostitutes could congregate. 77 It was also argued that congregation had been happening naturally anyway and that, except by complaint of nuisance from neighbours, it seemed unnecessary to expedite spontaneous movement. Providing a forceful conclusion to his reply, Macpherson echoed Williams' earlier comments on the Rangoon Police Act amendment, commenting that segregation, though common in Japan and Europe, would be foreign to a national of Great Britain and a novelty in Calcutta that would lead 75  to opposition '. . .and with a repetition of the well known charges that the State was recognising and lending its aid to organised vice'. 78 In response to the letter, a note on file in the Home Department rather timidly decided it would not pursue 'localization' in Calcutta further.
What is apparent from the discussions above is that outside the operation of the Indian Penal Code, the Criminal Procedure Code, or All-India legislation, British India functioned as a network of legislative spaces that drew upon each other for ideas and innovations with regard to specific problems of law. The bureaucratic fiction of India here fragments into differentiated blocks. Historically these had divided into the Presidencies of Madras, Bombay, and Bengal. However, as the colonial state developed into increasing numbers of provinces (9 by 1901, 11 by 1935) through the periods of dyarchy (1919-1935) and into fuller provincial autonomy (1935)(1936)(1937)(1938)(1939)(1940)(1941)(1942)(1943)(1944)(1945)(1946)(1947), they created their own legislation, increasing the variety of innovations that could inspire other provinces, as the earlier example of the North-Western Provinces has shown. This network was by no means harmonious, of course. The example of Calcutta has shown how certain nodes, especially those with imperial pedigree and influential figureheads, could resist the suggestion and stimulation of central government. Whilst effective in creating solutions to the problem of post-Cantonment Act municipal prostitutes, the network was itself also vulnerable to internal ruptures and crises, as the scandals in Rangoon and Bombay would illustrate in 1917.
While Calcutta had resisted the urge to segregate, this did not prevent other provinces in India from following the examples of Bombay and Rangoon. As the discussion of the Rangoon powers had suggested, the ability to ban the existence of brothels near cantonments slowly morphed into the power to segregate brothels in one part of the city. This was not a process of identical reproduction, but more one of normalization and infusion, just as Flavia Agnes described the way in which legal changes to the matrimonial status of Christian and Parsi women came to influence uncodified Hindu and Muslim personal laws. 79 For example, the Bengal Disorderly Houses Act (III of 1906), which also applied to all the municipalities in the Province of Bihar, was followed in the next year by identical powers in 78 Ibid. 79  the Eastern Bengal and Assam Disorderly Houses Act (II of 1907). 80 The act allowed first class magistrates to order a brothel or house used for habitual prostitution or by disorderly persons to close if it was in the immediate neighbourhood of a cantonment. However, the same order could be passed if the house was in the vicinity of an educational institution, boarding house or hotel used by students, or if the house was used to the annoyance of 'respectable' inhabitants in the vicinity. Whilst not instigating such a strict model of segregation as Rangoon or Bombay, these powers could be used to protect not only the military, but also the youth and the respectable classes of a municipality from social and biological threat of prostitution. Similar powers were extended through the British Baluchistan Bazars Regulation (V of 1910), enabling respectable inhabitants of a vicinity to get a brothel closed down. 81 Comparable processes were also underway in Ceylon. In 1899 an ordinance had been passed to suppress brothels in Colombo, although they continued to exist all over the city. As the Inspector General of Police wrote to the Colonial Secretary on 27 March 1916, in 1900 the brothels were highly conspicuous, existing on prominent street-corners in residential and business districts: 'These houses were so conspicuous that soldiers and sailors were frequently seen adjusting their dresses, even urinating on the road'. 82 The problem was, therefore, not just the influence of brothels on the area, but also the men who visited them, the effects of which being augmented by the dispersed nature of the brothels. This led to unfortunate cases of mistaken identity, the Inspector General continued, such as when a bunch of soldiers and sailors entered the Ladies' College close to a favourite haunt in Union Place, and refused to believe it was not a brothel.
As such, a new ordinance was passed in 1907 that restricted immigration to Ceylon of destitute or 'vicious' persons and allowed the closure of a brothel in a residential area after local complaint. Discreet brothels continued to exist in residential quarters if normal residents could remain untroubled by their existence. This policing of visibility was said to have met with success, as the General of Police continued in his letter of 1916, '. . .the brothels were out of sight where they could corrupt no one and could not be seen by persons who did not want to find them; the houses were quite orderly and few thefts occurred'. 83 The protection here was not of a cantonment, but of the respectable residents and areas of the city, by effectively segregating prostitution into 'Reclamation Road'. 84 However, the segregated zone soon befell the fate that lay in store for those within India. In 1909 agitation started against segregation, as the General of Police put it, '. . .by persons who were foolish enough to think, or pretend to think, that the immediate driving away of all prostitutes would result in the total absence of immorality'. 85 In 1912 the Scottish professional abolitionist John Cowan arrived in Colombo and started the sort of effective campaigning that had worked so well against the Indian Contagious Diseases Act in the past and would be put to such good use in the future by bodies like the Association for Moral and Social Hygiene, by whom Cowan was later supported. He devoted himself entirely to his work and cut an impressive physical figure; as Alison Neilans later remarked: 'John Cowan always looked about as poor as anybody could look; he expected the Lord to provide, and he did!. . .he certainly did most wonderful work in the Far East'. 86 After an intensive campaign of preaching and petitioning in the segregated zone, Cowan met with success. Within a year the segregationist policy was abandoned. Though Cowan had found Colombo fit to campaign against, it was not as stark a segregated system as had emerged in British India, as he would find when he disembarked at Rangoon in 1914. There Cowan would take part in making this local site of segregation into a site of scandal that would help overturn segregationism and open the way for abolitionism in interwar India. But this was dependent upon broader developments that had been slowly stigmatizing the brothel. These constituted scandalous geographies across a range of scales: from the micro and the brothel as a local site of sexual perversion; to the meso and the coastal/inland practice of prostitution; to the macro and the permeability of India's borders to trafficking in prostitutes. 83

1913-1918
The foregoing analysis affirms the ongoing revival of interest in the role of the state, and of governance more broadly within studies of colonial India. But lavishing attention on the ambitions of governors and the intricacies of regulations can detract from the effects of these measures when they are put into practice. 87 Such a bias also deflects attention from the agency by which governing is materialized, internalized, contested and altered. 88 It is at this scale that governmentalities become 'problematized' and it becomes apparent that policies are most often devised in response to problematization of existing practice, rather than out of some essential urge to improve, civilize or initiate 'Progress'. Outside of political theory and the academy, in social practice and cultural context, these 'problematizations' are often encountered as 'scandals'. While many would argue that the very concept of colonial occupation was a continuous scandal, there have been studies of specific scandalous events and their consequences. Dirks has written of the Scandal of Empire in his study of the East India Company and the British negotiation of its excesses and exploitations. 89 Arnold has asked: 'What kinds of crises-whether social, political, or epidemiological in nature-precipitate the formation of a system of pubic health, or bring about its reform and reconstruction?' 90 The three crises he identified, relating to cholera in the 1860s, the plague in the 1890s, and legitimation by a nationalist government in the 1940s, embed the influence of scandals and crises within the lurching development of the colonial state. Similar work on contemporary India has been done by Rajan, who explored The Scandal of the State through examining the points of crisis that linked conflict, controversy and contradiction over women in custody, in law, and in death. 91  mentioned the colonial scandals resulting from the sexual escapades of 'privates off parade' and explored the international class-racesex dynamics of an 1860s military scandal. 92 Rao and Pierce have commented upon the scandal of racialized violence and torture in colonial India, 93 just as Kolsky has covered the 'judicial scandal' of the acquittal of the majority of 'white violence', 94 while Sinha has excavated the global scandal provoked by Katharine Mayo's analysis of gender relations in late colonial India. 95 These studies attest the power of scandals to travel and scandalize outside their place of origin, becoming part of what Sinha termed 'imperial social formations'. 96 Whilst attention to this mobility is vital, and is central to explaining the collapse of segregationism in India, it must not detract attention from the complex social networks, cultural norms, and geographies of assemblage 97 through which scandals emerge in particular sites. Scandals can become one of the complex mechanisms by which a series of spatial configurations and events coalesce into the enigmatic and excessive property of place. Places can thus come not only to act as aides-mémoire of particular events, but can also become synecdochic, representing a larger chain of elements of which they are only a part. 98 For instance, in the years after the uprising of 1857 Delhi came to stand for the betrayal of Indian Muslims and the supposed despotism of the last Mughal, 99 just as Morant Bay came to symbolize the potential vulnerability of Empire after the Jamaican uprising of 1865. 100 Port Said was perhaps the most scandalous international site of prostitution, geographically and racially bridging the east and west in a site of supposed orientalist depravity. 101 It was this process of becoming synonymous with immorality that colonial governors feared, in addition to the Secretary of State's concern, expressed in 1899, that the Presidency towns should not become vectors of contagious disease. 102 The following section will frame the scandals that emerged in 1917 Rangoon and Bombay. These discredited segregationism as a means of regulating prostitution, due to its effects on wider civic society and on the occupants of the segregated zone respectively. These scandalous sites also had other geographical consequences. The central government would continue to work to create a scalar division between local initiatives against scandals involving prostitution and a benign yet uninvolved higher tier of government, despite its ongoing attempts to steer local policies. The scandals also opened up vistas onto other potentially more diffuse geographies that plagued the insecurities of the official mind. Three in particular will be explored below: first, the ever-present threat of the emergence of lower class homosexuality, should spaces for the release of libidinal energy not be permitted; secondly, coastal borders that were permeable to external trafficking; and, finally, a coastal cosmopolitanism that threatened to spread inland. Central to all these debates was the brothel, the segregation and sanctioning of which came to be seen as the core of the problem of, not the solution to, prostitution.

Barriers or roads to sodomy?
Colonial policy regarding prostitution was driven by fears for the imperial civilizing project, haunted by the spectres of a worldwide spread of British 'syphilization', 103  It is widely accepted that sexuality was medicalized and categorized in the nineteenth century, and that individuals who engaged in same-sex encounters were replaced by the 'homosexual' as an abnormal and pathological sexual identity. 105 Yet the opinions of Raj officials attest to a much more disconcerted belief in the fluidity of sexual mores. Section 377 had been introduced into the Indian Penal Code in 1860 to punish the offence of sodomy, buggery and bestiality. However, despite this deterrent, Hyam has shown that there were concerns that without sexual contact with women the Imperial Army would become 'replicas of Sodom and Gomorrah', or would pick up 'special Oriental vices'. 106 In 1894 Viceroy Elgin claimed that having no prostitutes would lead to more deplorable evils, with an increase in unnatural crimes already reported, 107 while in 1913 the Governor of Burma opposed the deportation of European prostitutes from India for fear it would lead to an increase in '. . .the crime of rape and the secret vice of the Oscar Wilde type'. 108 In the same year the Home Member, Sir Reginald Craddock, presented evidence for such vice amongst native prisoners at Port Blair's penal settlement in the Andaman Islands. 109 The spatial conditions (isolation from women) and social routine (lack of discipline) were said to condemn convicts to a life of abstinence or sodomy. The latter was used to curry favour with petty officers, presenting '. . .the royal road to privileges' for '"habitual recipients" as the passive agents are termed'. The latter were widely known but only came to official attention when they became diseased, or were murdered by former recipients who had been supplanted in favour by new inmates. Though offering fascinating insights 'beneath the surface' of what Craddock termed sexual morality, the relative paucity of such anecdotes affirms Peers' suggestion that the priority was to deny, rather than prosecute, homosexuality in the British Army in India. 110 At various times the threat of homosexuality was mobilized to justify the segregation of prostitutes, so as to make them available for the armed forces and the population more broadly. In his 1920 commentary on the abandonment of segregation in 1912, the Inspector General of Police of Colombo listed the effects of the delocalization of prostitutes. He claimed they were now scattered all over Colombo, annoying local residents, pestering people in the streets, visiting private houses, and causing disturbances as strangers to town tried to locate them. A 'more serious aspect' which he raised, however, was the fact that: On any night there may be seen in public places a number of boys from 14-18 years of age walking about singly about the hours of 10-11pm 'waiting to be spoken to'. I have two cases in hand of young men who, prior to 1912 would have visited the houses referred to [brothels], have been found in the company of native youths for improper purposes. 111 The Ceylon Vigilance Society report of 1915 confirmed that 'unscrupulous villains' were continuing to trade on boys, an old practice but one that was growing around the Fort, further suggesting that the military were a willing market for homosexual acts should heterosexual prostitution not be readily available. Similar fears were evoked in November 1917 when the segregation system in Rangoon was under debate. A review committee on a report into prostitution by Rangoon Deputy-Inspector Shuttleworth in 1917 insisted that: '. . .throughout history the vice of homo-sexualism has invariably made its appearance among collections of male human beings deprived of access to women. This vice also exists after its cause, as shown by Greek literature and the Arabian Nights, as translated by Richard Burton. We are satisfied it exists in Rangoon and have no desire by undue repression to replace promiscuous intercourse by unnatural vice '. 112 This was argued to be the case in rural as well as urban environments, where the lack of anonymity made such transgressions politically, as well as socially, dangerous. On 25 November 1918 the British Chief Commissioner in Baluchistan, H. R. C. Dobbs, wrote to central Government requesting the power to segregate prostitutes in 110 Peers. Privates off Parade, p. 843. 111 NA/Home(Police)/1920/24-29A. 112 NA/Home(Police)/1920/24-29A/January. the town of Quetta so as to place them out of bounds to soldiers. He could not, however, suppress a series of doubts over the proposals: From a political standpoint, however I confess to some fear lest the soldiers, deprived of their formed normal and recognised opportunities for the indulgence of their passions, may turn secretly much more than before to the women of the villages and tribesmen in the neighbourhood of Quetta and may thus not only excite the animosity of the tribesmen, but contract and spread venereal disease in a manner which may be even more difficult to cope with than before. There would also seem to be some danger that, if intercourse with women is made very difficult for troops, the practice of unnatural vice with the youths and boys of Quetta and of the surrounding villages may to some extent take its place (this being a vice by no means uncommon on the frontier). Such a result might damage the reputation of the British Government and have a bad political effect. 113 It was, therefore, acknowledged that segregation could have varying effects on sexual habits and desires, and on the abominable recurrence of homosexual acts. Should it be designed to facilitate safe and easy access to prostitutes, as in Colombo, segregation could supposedly stem homosexual deviance by satiating libidinal demand, but if it were designed to restrict military access to prostitution, it could risk driving the men into the frontiers, of both the countryside and of sexual morality.
An alternative view, however, was deployed by abolitionists in the campaign against tolerated brothels, representing a broader shift against the brothel as a site of visibility and safety, into one of danger and occlusion. An explicit rendering of this argument was provided by John Cowan in his exposé of the segregated quarter in Rangoon, to which he had travelled after his campaign in Colombo. His report, published by the Association for Moral and Social Hygiene on 9 June 1916, included a section on 'PROSTITUTION AND SODOMY'. He described venues which were 'centres of sodomy', and then went on to dismiss the logic that dictated that brothels prevented an increase in sodomy: My belief is that sodomy exists side by side with prostitution, particularly where prostitution is most rife. It is not an alternative vice to prostitution, but a sequel to it, a stage lower, in fact the last stage in the descent of the sensual man. With the possible exception of some special cases, men proceed through prostitution to sodomy, not from chastity to sodomy. The public brothel is an easy stepping-stone to sodomy, as boys are frequently kept there, or thereabouts; it is the road to sodomy, never a barrier against it. 114 This was in line with some of the emerging sexology literature which identified brothels not only as sites of disease, but also of inversion. 115 Cowen claimed that sodomy in Colombo had existed alongside prostitution, before segregation, and this case was still being cited in the 1920s. At the first Imperial Social Hygiene Congress, organized as part of the British Empire Exhibition at Wembley, London, in May 1924, the issue was raised by Sybil Neville-Rolfe, General-Secretary of the National Council for Combating Venereal Disease. 116 She noted that the case of Colombo was widely cited, in terms of the relation between abolition and the rise in unnatural vice. But she claimed that a special Committee of Inquiry had shown that the practice existed quite independently of 'ordinary prostitution' and had not been affected by the closure of brothels. Similarly, on 20 January 1929, the Army Department sent to the Military Secretary the results of a committee that had been formed to consider whether the regulations in place to check venereal diseases were too restrictive. 117 In support of this argument was the fact that Courts Martial for unnatural offences were seriously increasing, being three times as numerous as they were three years previously, although the committee dismissed the link between unnatural vice and the campaign against venereal diseases. Whilst the abnormality of homosexual desire threatened to cause a social and political scandal, the trafficking and occupation of white prostitutes resulted in even clearer scandalous geographies, that also spanned the rural and the urban.

A white slave trade?
Prostitution had long focused around port towns, taking advantage of mobile capital and labour. 118 But as technological advances eased the cost and strain of travel, so prostitutes themselves became more mobile; able to search out more profitable markets, but also vulnerable to deception and coercion into sexual economies not of their own choosing. 119 It was this permeability of their coastal borders that came to vex Indian officials in the early twentieth century, with this vexation focusing on the figure of the 'pimp', 'souteneur', or 'trafficker'. Their mobility made it difficult to gather enough evidence to have them arrested and deported but, under pressure from the League of Nations, trafficking would emerge as the key signifier under which a host of actions against prostitutes would be justified in the 1920s-1930s. 120 But these concerns had also surfaced earlier with regard to the relative geographical proximity but administrative autonomy of Ceylon. Just as the provinces of British India compared notes, inspired, and conspired against each other, so these networked geographies stretched outside the Indian Empire to its colonial neighbour. 121 During a review of the effectiveness of its regulations in April 1903, the Bombay authorities commented on the difficulties of using the Foreigners Act (III of 1864) against foreign pimps. Because they were aware of how long it took the police to receive orders for action, they only stayed long enough to settle the women they brought to brothels. Commenting on these proposals, the Home Department noted that: 'It is stated that these foreigners make Colombo their headquarters. The Bombay Government suggests that it would probably be of considerable benefit to Calcutta, Rangoon and Bombay if measures could be taken to render Colombo untenable for such purposes'. 122 But it also noted that more information would be needed about what it termed 'prostitute farmers' before any proposals would be put to the Ceylonese government. As shown above, increased powers against pimps were introduced, alongside the segregationist orders, in Ceylon, in 1907. Yet concerns over border crossings would continue to problematize the boundaries of the colonial state in relation to what Helen Wilson termed 'cooperation of HM dominions' in March 1913. 123 Wilson was writing to the Secretary of State for India as Honorary Secretary of the British Committee of the International Federation for the Abolition of State Regulation of Vice (the forerunner to the Association for Moral and Social Hygiene), and complained that brothel-keepers and foreign prostitutes who were expelled from Ceylon were simply shifting to Bombay. She included the following report from her correspondent, dated 5 February 1913: Mrs Caroline Goldenberg, owner and keeper of one of the worst of these houses, also owns a brothel in Bombay, and is said to have another at Port Said. She imports girls directly from Germany, has been engaged in the White Slave Trade for 20 years, and boasts that 1000 girls have passed through her hands. She has amassed a large fortune. She is now being allowed to move at her leisure to Bombay. In view of the feeling in England against the White Slave Traffic, I am surprised that the authorities here have thought fit to let this woman loose on other British territory.
This report strings together some of the many discourses, and inconsistencies, that informed discussions about trafficking in the early twentieth century. The report clearly links brothels with trafficking, which would become the main point of focus of the League of Nations' investigation of trafficking in the 1930s. It deployed the concept of the 'white slave trade', the hybrid discourse of myth and fact that reached a crescendo in Britain between the Pall Mall Gazette exposé of 1885 and the Criminal Law Amendment Act of 1912. But it also attests to the agency of women, albeit here as brothel keeper, within these networks as 'let loose' on British territory. Whilst much attention has been lavished on transgressions of international borders, there has been less emphasis laid on the transgressions that these prostitutes could commit inland over the social landscape.

Coastal cosmopolitanism
It was this sense of prostitute or madam as contagion that framed the fears over mobility within India, rather than the orchestrating maritime authority of the pimp. Rather than the mobility of Indian women who were either drawn into or willingly moved into prostitution, the source of governmental concern was the mobility of 'white' prostitutes. These concerns were not provoked by the possibility that they were enslaved (most encounters with the women dispelled this myth fairly sharply), that is, an issue of slavery and rights, but by the fact that they were white, that is, an issue of imperialism and race. The fate of European prostitutes who were estimated to live in India at any one time between the 1880s and 1914, and those from the Far East to a lesser extent, have been well researched. 124 In terms of numbers, the Government claimed in 1913 that there were 234 European prostitutes, and 320 Asiatic non-Indian prostitutes in the country, with 126 of the former in Bombay, and 50 in Calcutta. 125 Levine has studied the 'orientalist sociology' by which prostitutes were hierarchized and categorized, arguing that: 'Colonial definitions of sexuality served to map a geography of racial borders through a complex taxonomy or racial and ethnic distinction, in which these categories were always more than innocently descriptive'. 126 White prostitutes created a great deal of anxiety because they were: members of the ruling race, but of the subordinate sex; either financially independent or at risk of becoming economically enslaved to their pimp; and unconstrained by Victorian notions of sexual restraint. 127 For example, in the third edition of his sensational book, The Social Evil in Calcutta, published in 1886, R. Kerr denounced the phenomenon of white, Christian, prostitutes in the city. 128 These women were hierarchized by location and status. In 'class I' were women seduced and abandoned by men of position in society, which he numbered at about 15, and who lived in well ventilated houses with several furnished apartments. The lowest class, estimated at 365 women, dwelled in dark and dismal alleys, with narrow verandas and damp walls. It was the intermediate category of women, of which there were estimated to be 150, who flitted between men, that most roused Kerr's indignation. Often living in shared tenements that worked to degrade any I st class prostitute who found herself in their company, it was this intermediary level of vice that presented the greatest threat to society: 'She it is who inflicts upon it the greatest scandals and damage'. 129 Selecting her clients at the opera or theatre, the '. . .flaunting, extravagant quean' 130 would demand 'abject submission' from her lover until his wealth was drained, after which she would switch, '. . .even to his best friend. She is a white-washed sepulchre, fair to the eye, but full of rottenness'. This class of women could thus ruin Calcutta's men folk, both financially and biologically, but could also drag down its white women, from 'a spirituelle 131 Anonyma 132 to a slatternly, tipsy, foul mouthed drab 133 '. 134 One solution to the problem of European prostitutes was to stress the difference between east and west European women, and to try to deport the latter. As for the former, the police of Bombay city are well known to have 'herded' together European prostitutes in Kamathipura in the 1880s, 20 years before the introduction of segregationist powers for prostitutes more generally. 135 White prostitutes thus prompted a specific geography of segregation such that they could be placed under special surveillance as members, albeit differentiated and problematic, of the ruling race.
While governing officials were willing to accept white prostitutes in Bombay, this was part of a broader geography that accepted a degree of coastal cosmopolitanism, as defined against inland traditionalism. A parallel geography existed between the British and Princely states. The former were envisaged as modern and outward looking, whilst the latter were continually re-orientalized: first as the inheritors of oriental despotism, wedded to the ancient past; later as the lavish and ornamental representations of the exotic east. 136 But there also existed broader, sexualized geographies of the subcontinent within colonial discourse, which bear comparison to the gendered divisions which Chatterjee famously commented upon in Bengali nationalist discourse. Here, space was divided into the external, modern, industrialized and material realm, against the internal, traditional, Indian, spiritual realm. 137 One spatialization of this dichotomy was ghar (home) and bahir (world). At the continental level this dichotomy played out between the modernizing, globalized sites of India's major cities, especially those port cities exposed most regularly to external influences, and the rural heartland, the cherished homeland of Mother India. Male travellers, labourers and traders congregated at the port towns, while the inland territories ideally remained unpolluted, unchanged, and unmodern.
This linked to a wider, transnational network of sites which troubled the racial and sexualized hierarchies of the Indian administrative elite. For instance, in reflecting upon his career in the Indian military Sir George MacMunn commented upon the demoralizing experience of Indian migrants passing through places such as Shanghai and encountering various types of 'white scum'. MacMunn suggested that for many Indian men this was their first encounter with white women of the '. . .courtesan type and for the first time the Indian here realises that the woman of the white skin can be worse than the outcaste women of India. This does not improve the Indian outlook on life'. 138 Sohan Singh Josh later reflected upon this comment in the broader context of the Indian diaspora, showing that the 'demoralization' was of the British civilizing project, not of migrant labourers. 139 However, if the Indian migrants had passed out of Bombay, Calcutta, or Rangoon, there is every chance that they would already have experienced the phenomena of white prostitution. One of the main problems with this coastal cosmopolitanism was that three of the larger port cities were prestigious Presidency towns. The comments above on Ceylon display how ineffective laws and regulations were at disciplining these porous and complex sites. These ports were also the main emblems of the Raj and needed to reflect the achievements of the British Empire, hence the Secretary of State's heightened concerns that they should not be ravaged by 'contagious' (venereal) disease. But this emblematic status also brought with it the altered social state of urban modernity, including anonymity, spaces of experimentation, and opportunities to re-forge spatial relations and re-articulate race, gender, class and sexual identities. Although white prostitution was thus begrudgingly accepted in the major port towns (which also possessed the means to deport unacceptable displays of European sexuality), these developments could be destabilizing elsewhere so had to be regulated.
One example of this regulation was provided by the campaign of Reverend Joseph Reed in 1912. His complaints were concerned with the existence of white prostitutes in the Cantonments of Lucknow, who were said to openly flaunt and solicit their trade in this inland city of the United Provinces, and who should have been expelled. 140 Commenting on 19 June 1912, Quarter-Master General F. J. Aylmer dismissed Reed as a dangerous fanatic who never believed anything he was told on the subject, but stressed that he had the capacity to raise interest on the topic in the United Kingdom and should thus be treated with caution. As such, although expelling white prostitutes from cantonments might cause more harm than good, it was advised to meet Reed's wishes, which were also consonant with the interests of government, as Aylmer continued: We have, I submit, a special reason for excluding the foreign prostitute, which does not apply to natives. It is an 'Imperial' and very urgent reason. As long as a woman is white the native does not much consider whether she is English or Foreign. Her degradation constitutes in his eyes that of the ruling and paramount power in India. White prostitutes in India who openly ply their trade must do us a vast amount of harm. In an ordinary Cantonment their conduct is apparent to all. In a great town or city it is a mere incident. One of the reasons always quoted by the Government of India for their inability to follow the advice of what we may call the 'Ultra Purity Party' is that prostitutes in India belong to a regular caste and that the prostitute is born such and has no other means of livelihood. We cannot uphold this argument in the case of the white prostitute. 141 The geography of sexual scandals was here made clear: the cosmopolitan port towns cloaked the white prostitute with at least a little anonymity, or at least a more commonplace notoriety; in an ordinary, inland cantonment their scandalously racialized sexuality would be plain for all to see. It was thus agreed to be '. . .possible, practical and legal. . .' to exclude white prostitutes from cantonments, force them to the great sea ports, and deport them as undesirable aliens should they continue to be of trouble. The problem which had been laboured over so intensely in the previous ten years concerning native prostitutes, that they might '. . .loaf about soliciting. . .', was not thought to apply to white prostitutes, '. . .who would not be able to live such a precarious existence and could easily be detected if they did'. 142 Investigations by the Home Department showed that the existence of these women in Lucknow was part of a wider and more troubling trend that had seen white prostitutes penetrating the Gangetic heartland of India. It was shown that they had previously been settled in Rawalpindi and Lahore, but were ejected from those towns due to the scandal that their presence had caused. 143 The label 'scandal' recurs in this discussion of the women: the Home Department wrote to the Chief Secretary to the Government of the United Provinces in 23 July 1912 explaining that the Commander in Chief's attention had previously been drawn to the women of ill-fame in Rawalpindi due to the '. . .scandal from the presence and conduct of European women;' the Adjutant General was said to be '. . .anxious to nip the scandal in the bud. . .' at Lucknow. Great interest was taken in the way by which the Punjab Government had expelled the white prostitutes without using the Foreigner's Act (III of 1864), as usually happened in the port towns, but had used a combination of the municipal act and cantonment code. On 1 July 1912 the Judge Advocate General's Department examined the laws that had been used and suggested that the Home Department consult the Rawalpindi and Lahore authorities, both civil and military, to ascertain how they 'got rid' of these women and asking for hints as to judicious procedure, such as dealing with objections that these laws were only enforced against white women. It would then be possible to suggest that Lahore use the same methods and, if successful, they could then be used in the United Provinces and elsewhere. 144 This solution to the problem was particularly appealing to the Home Department because there were still concerns about the scalar politics of regulating prostitution, and an unwillingness to promote an official policy from the centre. Commenting from the Foreign Department, W. Booth-Gravely argued on 24 September 1913 that the central government should proceed tentatively and not at once issue instructions to local governments: The subject bristles with difficulties-to mention only one, there is the ticklish matter of justifying a policy of comparative non-interference with indigenous or Oriental prostitutes while proceeding vigorously against foreigners. Whatever is done should therefore be done with caution, and so as to give the least opportunity for criticism which though not necessarily ingenuous or intelligent may none the less be effective and difficult to meet. 145 The proposed medium of suggestion was that of a 'demi-official' letter to local governments suggesting action to be taken and pointing out the cantonments in which foreign prostitutes were living. As a result, the Quarter-Master General circulated a confidential letter on 20 October 1913 to all general officers in both British India and the Princely States. It stated that the Commander in Chief, on both political and moral grounds, wished to rid military stations of foreign prostitutes. This was because natives would not distinguish between white Europeans and British women, but also because white women would prove a greater temptation for British soldiers. As such, it was felt justifiable to distinguish between native and European prostitutes. It was made explicitly clear that no extra measures be introduced against native prostitutes, but European women would be ejected using the existing Cantonment Code, which allowed for the brothels to be closed down. Addressing the problem of regulating expelled prostitutes, the government listed the powers available to neighbouring municipalities, including the sections used by the Punjab Government against the original women in Rawalpindi and Lahore, and newer acts such as the Bengal Disorderly Houses provisions from 1906.
What this incident highlights is the ongoing way in which Indian states were networked together through correspondence between sites of judicial experimentation. This involved pursuing white prostitutes around central north India, seeking out a legal mechanism suitable to banish them from the heartland and back down towards the more obviously modern spaces of the coast. This geographical distinction would become evident again during the Madge Bill debate of 1913, which would also mark a watershed between advocacy of segregation and abolition.
On 14 November 1912 the Government of India forwarded to the Secretary of State in London two separate private Bills introduced into the Imperial Legislative Council. 146 Both had the object of the 'further protection of women and children' so were submitted together. The Bills were also circulated to the provinces of British India for comment, and a summary of these debates and the conclusions the central government had reached were sent to London on 24 July 1913. Though it was decided that neither of the Bills would become law, the debate suggests a discursive tipping point had been reached by 1913, which the scandals of the interwar years would send over the edge.
The first Bill was submitted by the Honourable Manicekji B. Dadabhoy, a Parsi advocate and industrialist from Nagpur. 147 It was extensive in its aspirations and made various members of the Home Department uneasy due to its intervention into the religious and social life of the country and its ambition to legislate on sexual subjects. It proposed legislation that would address: the religious dedication of girls to temples, or the removal of girls from British India (to other countries or Princely States) for such dedication; procuration and defilement of girls for immoral purposes; sexual intercourse with a girl aged between 12 and 16 with consent; the possession, adoption or guardianship of girls by persons of immoral or undesirable character; the possession of a girl under 16, who is not her daughter, by a prostitute; mock marriages; giving away a child to concubinage under the age of 16; and transfer of wives. The majority of these objects were agreed to be served by existing clauses of the Indian Penal Code or the Criminal Procedure Code, although some amendments were suggested on the basis of Dadabhoy's legislation.
It was, however, the 'White Slave Traffic Bill' of Mr W. C. Madge that attracted the most attention, with its two objects of suppressing the importation of foreign women for prostitution and the punishment of importers, and those who profited thereby. The form of the Bill was damned quite comprehensively by the Home Department, who criticized its proposed procedure for deportation, the opportunities it would open for blackmail, and the failure to consider the fate of imported women. It was agreed, however, that the Bill raised three objects that prompted an interesting debate amongst the provincial commentators. The potential objects were: the total suppression of non-Indian prostitution; the exclusion of pimps and procurers; and the closure of houses of ill-fame.
The covering letter sent to London in July 1913 contained the statistics, referred to earlier, that India contained 234 European prostitutes and 320 Asiatic non-Indian prostitutes. While it was admitted that small 'colonies' of the women had 'drifted up' for short periods to places like Agra, Chakrata and Rawalpindi, they were mostly a seaport town phenomenon. They were believed to be almost entirely non-British '. . . as, so far it has been found possible to enforce an unwritten law that this class is not permitted'. It was also stressed that these women were not victims of the white slave traffic, moving mostly by free will and mostly being experienced prostitutes from Europe.
Turning to the three potential objects of the bill, the total suppression of non-Indian prostitution in seaport towns was not believed to be achievable. It was believed that the women could easily come back to India in another guise, and that removing them would give an impetus to prostitution in the domicile community of Europeans, or cause European men to use Indian brothels more frequently, both of which could do great harm. 'Up-country', however, was a different matter. The spread of non-Indian 'loose women' was a recent phenomenon and it was decided to address local governments on the topic.
Tackling pimps was more difficult, because they stayed in the country for only brief periods, and often 'imported' women who were travelling of their own free will. As the Home Secretary, H. Wheeler, summarized the internal debate on the matter in his notes on file: 'Put briefly, we cannot ordinarily hope to obtain the evidence requisite to support prosecution against these men in Court; consequently we can only take a wide executive power against them as undesirables and kick them out of the country under that'. It was proposed to achieve this by strengthening the Foreigners Act (III of 1864). Along with the measures against the movement of foreign prostitutes, the above tightening of measures against pimps were in line with movements internationally towards a focus on 'trafficking' that would fully emerge in India in the 1920s. However, like much international opinion at this time, there was still a belief in segregation. Viceroy Hardinge's letter to the Secretary of State of July 1913 referred to reports from Bombay and insisted that the system of segregation had its advantages: It localises the evil and greater demoralisation might ensue if these women were scattered throughout the town. So long as they are not detained by force or ill-treated they are probably better off than if left to find separate shelters for themselves, and the managers of the house exercise a certain restraint over them.
As such, while no major legislation resulted from the two Bills, on 2 August 1913 a letter was sent from the Secretary of the Home Department to the heads of 'up-country' provinces where white prostitutes had established themselves in 'small colonies' stating that: The scandal caused by the presence of these unfortunates up-country is more conspicuous than in the large sea-ports; they are in smaller towns and among a less cosmopolitan population, with the result that they must attract more attention.
Rather than legislating on such a potentially controversial matter, however, the central government once again delegated responsibility and suggested courses of action. It commented on the use of the Foreigners Act to expel white prostitutes from Calcutta and Bombay, and then attempted to stimulate action: In the absence of information as to the local conditions, the Government of India do not wish to direct this course, but they trust it will receive the earnest consideration of the local Government, which doubtless is at one with the Imperial Government in desiring to nip in the bud this incipient tendency of foreign women of loose morals to find their way up-country.
While obviously backing segregation, delegation to local government opened policy up to local pressures when the turn against segregation began in earnest. Although segregationism had its defendants up to 1917, by the end of the war and the exceptional conditions that it presented, the stage was set for a new round of judicial experimentalism that would rely on a Bombay-Burma axis for the model upon which the suppression, rather than segregation, of prostitution was pursued in the 1920s.

Conclusion: stimulation, segregation, scandal, suppression
By the time of the First World War, attitudes in India appear to have reached a tipping point regarding the toleration of prostitution, trafficking, and segregated brothels. Pauline Rule has described the attention paid to the violence meted out to young wives, which was said to lead many to desert their husbands only to find themselves captive in Calcutta's brothels. 148 In 1918 two cases came to light that exposed the dangers of the brothel system: one of an Inspector General offering protection to a brothel that abducted young girls; the other of a 14 year old girl abducted from a festival and forced into prostitution. When the captors of the latter were sentenced to seven years, great public feeling was aroused, with applause and demonstrations inside and outside the court.
The two great segregationist sites of Bombay and Rangoon likewise fell into scandal during the years of the First World War. Ashwini Tambe has sensitively described the fall-out in 1917 of the murder of a young woman confined to a brothel in Bombay's segregated area. 149 In Rangoon the protest focused on the effect of locating the segregated area within the heart of the civic district, which was said to be threatening the health and morality of the city. 150 These scandals discredited the segregation system for a variety of reasons. As suggested above, the status of the brothel was central. From a site of visibility, knowledge and security, it had become a site of biological and social risk, invisibility and confusion. It became associated with sodomy, trafficking, and the scandalous sexualization of 'up-country' India, while failing to reduce rates of venereal disease. But these failings became a trans-local scandal because of shifts in what Sinha terms the national and imperial social formation. The state was increasingly seen as the barrier, rather than the road, to social progress. Late-colonial nationalists and reformers demanded that the state encourage health throughout the Indian population, prove its claims to be a civilizing force, and allow Indian selfreform regarding child marriage, temple prostitution, and women's uplift. The segregationist system clearly demonstrated the state's commitment to a military, enclavist mentality, even if the protection of its troops and soldiers now took place in municipalities and cities.
In exploring the emergence of the segregation system, this paper has brought to light features of the colonial state that should prove of wider interest. A system of dyarchy had been in operation for prostitutes long before 1919, through which provincial authorities instituted their own form of local self-government in terms of health, policing policy and social affairs. But this was within a system whereby the central government denied, in public, any association with 'organized vice', but influenced policy through patronage and persuasion in private. This informal imperial combination of '. . .official absence with effective presence' is more often noted between states than within them. 151 But this absent presence can clearly be seen in the Government of India's turn of the century endorsement of the segregation of prostitution. Though segregation did not survive in India for long following the First World War, the system for dealing with prostitution demonstrates continuity across the Montague-Chelmsford reforms.
The central state's response, after the 1919 Government of India Act, was to insist that dyarchy made prostitution a local scale problem, to be dealt with by provincial legislation. The Suppression of Immoral Traffic Acts that spread throughout India in the 1930s were in response to the first Act in Calcutta of 1923, which itself was based on abolitionist legislation passed in Burma (1921) and Bombay (1923). However, the state continued to encourage and shape this legislation, with the aim of reducing venereal disease in the military and convincing the League of Nations it was doing something to reduce trafficking. Though reworked to encompass the scalar terrain of dyarchy and provincial self-government, the state continued to influence post-segregation and post-scandal urban sexual landscapes through the powers of suggestion, stimulation, and desire.