Family Firms in Hyderabad: Gujarati, Goswami, and Marwari Patterns of Adoption, Marriage, and Inheritance

Scholars are looking again at banking and mercantile families in India's early modern history, responding to the challenge issued by Claude Markovits in the epilogue of his 2008 volume, Merchants, Traders, Entrepreneurs, to “return the merchant to South Asian history.” Some of the underlying assumptions and questions being asked are old and some are new. My own longstanding assumption, upon which this article relies, has been that bankers and merchants played multiple and important roles with respect to states in South Asia, and that their relations with non-kin officials and other political actors determined their success or failure and sometimes the success or failure of a state, most notably, the Mughal state. Questions are again being raised about “trust,” assumed to be a leading attribute of and asset to financial networks (especially in long-distance trade diasporas), and the notion so commonly put forward by scholars to explain the success of Hindu banking and mercantile communities. Recent work by Francesca Trivellato has found that membership in the Sephardic trade diaspora facilitated but did not guarantee trust or cooperation: the Sephardic merchants relied on non-Jewish as well as Jewish agents and networks of information, and evolving legal norms guided their business activities.


Consideration of legal norms alerts us to claims being made by Timur
Kuran and Anantdeep Singh about the advantages of Hindu law over Islamic law (as written in legal texts) for capital accumulation and the continuity of family firms in the South Asian and Middle Eastern economic worlds. Islamic laws of inheritance, they argue, divide and diminish family resources, partitioning at least two-thirds of an estate among children, spouses, parents, and sometimes more distant relatives. While Muslim women receive only half of what men in each category do, legally they should inherit. Hindu Mitakshara law, by contrast, helps consolidate and maintain resources within joint families; Hindu women do not inherit and joint-family enterprises pool capital and discourage partitions of an estate. 4 Other legal questions concern the interactions between customary or personal law and the evolution of British colonial law in much of South Asia. Sumit Guha, for example, discusses precolonial, eighteenth-century conceptions of rights and legal practices in the Maratha country, "a region of shifting political boundaries, often indefinite jurisdictions, in which several unwritten and written bodies of law and custom could be invoked or set aside as the case might be." 5 The present research pertains to Hyderabad, a princely or native state outside British Indian jurisdiction, where the state made no attempt to regulate customary practices with respect to marriage, adoption, and inheritance within each kindred, caste, or community. Legal pluralism prevailed in the native states even as British India codified colonial laws from the 1860s, and, even in British India, attempts to devise laws and impose them on religious and caste communities varied over time and by region. Lauren Benton suggests that Indian litigants helped move a plural legal order toward a state-centered legal order by seeking to "exploit jurisdictional complexity." Speculating that colonial conditions "intensified the fluidity of the legal order and enhanced the strategic importance of personal law," 6  and market-oriented colonial commercial law from the late nineteenth century; Birla refers to the "extensive negotiability" that characterized vernacular practices governing the symbolic capital of kinship, caste, and lineage, and the capital flows of market exchange and production, a situation that colonial laws sought to restrict. 7 In Hyderabad by the late nineteenth century, laws evolving in surrounding British India and the establishment of a British Residency in Hyderabad began to affect family firms by providing new legal arenas for cases involving transactions and property holding.
To consider these issues, I look closely at three Hindu and Jain banking and mercantile communities, focusing on adoption, marriage, and inheritance practices among Hindu and Jain family firms in Hyderabad, India. These firms came from three major financial communities: the Gujaratis, Goswamis, and Marwaris. Gujarati merchants and bankers are so named because their homeland is Gujarat, and they have many subdivisions, some of them Hindu (mostly Vaishnavite, worshippers of Vishnu) and some of them Jain. 8 Marwaris came originally from the Marwar region of Rajasthan. 9 The three major Marwari castes or subcastes are Agarwals, Maheshwaris, and Oswals, the first two predominantly Vaishnavite and the last predominantly Jain. 10 The Goswamis in Hyderabad were Shaivite (worshippers of Shiva) sanyasis, celibate mendicants by tradition, whose roots were in northern and central India. Not much has been written about these communities in Hyderabad, 11 although listings of firms by locality and biographies of individuals appear in various sources. 12 Historical study of Gujarati, Goswami, and Marwari banking firms in Hyderabad State prior to the state's 1948 incorporation into India suggests several lines of argument, the first concerning adoption. Before the demographic transition, Birla, Stages of Capital, starts with the 1860s and 1870s, but most of the colonial laws she discusses were enacted decades later. 10 The Agarwals are divided into eighteen named exogamous groups, or gotras, the Maheshwaris have seventy-two exogamous sections, or khamps, and the Oswals also have many exoga- law) to adulthood problematic. 13 Perhaps the most famous pre-modern Indian merchant's life history is that of Banarasidas, who wrote in a dialect of Hindi in Agra in 1641. He had nine children by his three successive wives, but none of the children survived him. 14 With this pattern in mind, I argue that Hindu and Jain encouragement of adoption, in contrast to Islamic constraints on adoption, gave an obvious advantage to Hindu and Jain family firms when it came to the continuity of their firms and financial networks. 15 Hindu law permits adoption even by widows and of adults as well as children, the purpose being to provide heirs to property and resources rather than to provide for children. 16 I will present detailed evidence of the prevalence of adoption in Gujarati, Goswami, and Marwari family firms, derived from orally constituted genealogies and also from Hindi family histories of Marwari entrepreneurial families in Hyderabad.
My second argument is that marriage and inheritance practices among these patrilineal mercantile families were actually quite flexible, notably involving affines-relatives by marriage-as major players. Wives and their kin did play roles in mercantile family histories. 17 Flexible family strategies meant not only continuity for the family firms but also a potentially broad spatial range of financial networks. Gujarati and Marwari Hindu and Jain bankers and financiers practiced caste endogamy but gotra, or subgroup exogamy, so marriage networks could be and often were wide-ranging. In contrast, Indian Muslims in general encouraged and often practiced cousin marriage. This could result in a more limited range of options for marital networks, possibly leading again to an advantage for the Hindu bankers and merchants in terms of access to capital and other resources. (On the other hand, cousin marriage could offset the dispersion of resources caused by Islamic inheritance law: empirical data is needed here. 18 ) And although the Goswami sanyasi bankers did not marry, they did keep women and have children. There were no 13 India's mortality rate began falling in the 1920s, but India has not yet entered stage three of the demographic transition, when the fertility rate also falls significantly and stabilizes or decreases population growth. 14 Ramesh C. Goswami women before the twentieth century, but the men could choose (in Hyderabad) Rajput, Maharashtrian, or Telugu women, and they could adopt as chelas, or students, boys from a wide range of backgrounds. 19 Finally, I argue that trust is a mischaracterization or overstatement of the values and practices that helped maintain family firms and their financial networks; at least it seems a less-than-useful concept when analyzing the family firms in Hyderabad. Discussions and debates about trust focus chiefly on longdistance trade diasporas but also raise issues of kinship, caste, and religion that are relevant here, as well as the issue of the degree of merchant reliance on or autonomy from state power. 20 Claude Markovits writes about trust as a necessary ingredient to the successful conduct of business transactions but dismisses its basis in common religion, ethnicity, and/or caste; he also dismisses idealized "fictive kinship ties of 'brotherhood.'" While he sees trust as based on shared information about past behavior, Markovits does link it to power relationships within communities, communities based on locality yet also, he writes, on co-residence, kinship, and matrimonial alliances. 21 Others define business communities by shared dialects, account book scripts, or regional cultures, but shared caste and/or religion are always part of the discussion. 22 According to most scholars, trust is based on caste or kinship networks. 23 Kin or caste fellows care for family members in a trader's absence or apply negative sanctions to keep him in line. 24 Negative sanctions certainly depend upon membership in a kin-based or caste community. For example, asserting 19 In Mirzapur, Uttar Pradesh, the important northern depot for traders to the Deccan in the eighteenth and early nineteenth centuries, Alka Patel and I found that the Goswamis there kept Adivasi tribal women. Interview with Dr. Bhudev Pande, Mirzapur, 15 Aug. 2009. 20 Current debates center on the medieval or early modern period in North Africa and Europe. Avner Greif points out that a regional business culture could prevail despite differences of religion, showing that Maghribi Jewish merchants adopted Arabic language and culture and did not relate to European  23 Stewart Gordon, writing about family firms, and citing Bohras in western India in particular as largely self-governing and sharing both a faith community and internal marriage ties, states, "Distances were long, sums large and transport slow. All this demanded an extraordinary degree of trust among traders." Marathas, Marauders, and State Formation in Eighteenth-Century India (Delhi: Oxford University Press, 1994), 176-77. that caste-fellows were defending institutionalized trust when they tried to prevent a fellow long-distance trader in Moscow from converting to Christianity in the late seventeenth century, Scott Levi reasoned that the man's wife and other family members were being held accountable back in Multan, that their safety, even the children's access to marriage partners, depended on repayment of the loans taken to finance the man's trading venture to Russia. 25 Rather than illustrating positive aspects of trust, this stresses negative sanctions such as making transgressions public and denying opportunities for marriage.
However, although the Hindu and Jain family firms in Hyderabad to be discussed here were founded by migrants to the city and marriage partners were frequently brought from outside, the men heading the firms settled in Hyderabad and were not accountable to family members elsewhere. Thus entrepreneurial success on the part of Hyderabad's financial communities was more readily accounted for by close relationships with other political actors in Hyderabad, as I will show with case studies. Trivellato, reviewing scholarly work on trust, proposes instead the phrase "communitarian cosmopolitanism" to characterize the experience of Sephardic merchants who synthesized diverse traditions and mingled with non-Jews in a society of unequal and separate groups, and this conception is a useful one for considering diverse mercantile communities in the context of Hyderabad State. 26 moving inland from their trading ports. Still professing allegiance to Delhi, the Nizam and his successors established a separate dynasty and administration by the end of the eighteenth century, continuing Mughal administrative practices in a new context and with new sets of participants. 27 Banking firms, particularly what have been termed "great firms," 28 had been important participants in state-building in precolonial India, serving as state treasurers, minters of money, and revenue-collectors as well as maintaining long distance credit and trade networks. 29 The importance of these firms to state-building and maintenance is being increasingly confirmed, 30 and scholars are also looking hard at pre-modern merchant and trading networks in South Asia and beyond, reassessing their relationship to the development of capitalism. 31 Yet there is still no definitive history of banking in India that integrates pre-modern and colonial materials across the subcontinent (including the native or princely states). No authoritative, comprehensive history relates the "great firms" and diasporic trading networks to the agency houses and joint-stock banks that emerged under the East India Company, the British Raj, and postcolonial South Asia. The communities that participated in banking and trade differed by region within the subcontinent. Some trading networks bridged regions and countries. Diasporic communities of Shikarpuris and Hyderabadis from Sind and Khatris from the Punjab traded in Central Asia. 33 Gujaratis, both Muslim and Hindu, traded to eastern and southern Africa and Southeast Asia. 34 Tamil-speaking Chettiars and Muslims also traded in Southeast Asia, 35 and Afghan moneylenders moved down into India. Other bankers and merchants expanded primarily within the subcontinent. Major groups of Muslims, Hindus, and Jains moved out from Gujarat, Hindus and Jains extended out from Marwar and other parts of Rajasthan, and the Goswamis, sanyasis whose maths, or monasteries, were in northern and western India, also moved and became major traders between north India and the Deccan in the eighteenth century.
Bankers and merchants from these last three communities, of Gujaratis, Goswamis, and Marwaris, arrived in Hyderabad at different times, settled in certain localities, and established close relationships with particular officials and nobles at the Nizam's court. The communities' adoption, marriage, and inheritance practices varied somewhat, although for the non-sanyasi Hindus the Mitakshara system of Hindu law prevailed, so that all male members were coparceners of the joint family property. In the nineteenth century, the British resident in Hyderabad gave the localities under his jurisdiction some access to laws and precedents developing in British India, probably those from Bombay Presidency to the west, 36  , 762-803, discusses commercial banking (775, et passim) and "non-institutional" or indigenous finance and credit sectors (796, et passim), writing that not until 1929 did "unorganized finance" become a major concern of colonial policy (797). 33  36 Amrita Shodhan's pioneering work on religious groups and colonial law in Gujarat shows that before 1827 in Bombay Presidency the colonial legal system treated caste and religious groups as active polities, whose internal affairs were the concern of courts. After 1827, it treated them as textbased self-regulating communities, and after 1857, the Bengal judicial model was applied to all three Presidencies and attempts were made to unify codes and decisions. See her A Question of Community: Religious Groups and Colonial Law (Calcutta: SAMYA, 2001); and "Caste in the and the interplay between precolonial and colonial legal systems remain to be explored in Hyderabad State.
The tumultuous period of the late eighteenth and first half of the nineteenth century saw bankers becoming crucial players in Hyderabad state politics. The second Nizam, Nizam Ali Khan (1762-1803), moved the capital from Aurangabad to Hyderabad as the military aristocracy was succeeded by a Mughal-style bureaucracy based on the collection of land revenue. Nizam Ali Khan's successors, Sikander Jah (1803-1829) and Nasiruddaula (1829-1857), made agreements with the British resident establishing the Hyderabad Contingent, a military force that was, by 1813-1814, paid by the Nizam but led by British officers. The Nizam's inability to meet the payments put increasing financial pressure on the state as the East India Company became a creditor of the Nizam's government in 1813, making the Company a rival to the local banking firms that had been supplying loans. 37 Compelled to pay the Contingent but unable to use it as his own, the Nizam unsurprisingly built up his own forces, and support of these two military forces took a large part of Hyderabad's revenue.
When, in 1814, the payment of the Contingent Forces had to be guaranteed by the Nizam's government, the Diwan or Minister turned to the banking firm of William Palmer and Company, founded in 1810 by Captain William Palmer, Eurasian son of General William Palmer, and an officer in the Nizam's army. This firm, which I consider an indigenous one (see the case study of the Gujarati firm, below), advanced money to the Diwan Raja Chandu Lal at the rate of 2 1/2 lakhs rupees a month (a lakh is 100,000) and was assigned the revenues of certain districts (Berar) as security; the firm took interest at the rate of 25 percent. The firm continued to advance loans to the minister until its claim amounted to about a crore rupees, and when in 1820 it advanced a further 60 lakhs of rupees to Chandu Lal, this became the talk of India. A few months later Sir Charles Metcalfe became British resident in Hyderabad and protested to the governor general, citing a prohibition on British subjects making loans to native princes-but Palmer was Eurasian and not a British subject. 38 Metcalfe's outrage also clearly stemmed from the Palmers' immense popularity in Hyderabadi social circles, and his adamant opposition to Palmer and Company's key role in Hyderabad politics eventually led to inquiries in England and the firm's bankruptcy in 1824. 39 The British resident pressed for the back pay due to the Contingent, a debt that grew in the 1830s and 1840s and led to Hyderabad's financial crisis of the 1840s. Financial arrangements resorted to after the Palmer bankruptcy in 1824 failed to ameliorate the situation, and by 1850 the situation was drastic. Successive Diwans sought funds from bankers and mediated among bankers, revenue contractors, and mercenary troops, all of who became crucial to the state's finances. In the early nineteenth century, the Diwanship was held by the leading During much of the early period, effective control was in the hands of Raja Chandu Lal (a Hindu Punjabi Khatri), Peskhar or deputy Diwan (1804-1843) and also Diwan (1832-1843). Then Lala Bahadur (a Hindu North Indian Kayasth, daftardar, or land revenue record keeper, for the eastern half of the state), 40 played a key role similar to that played earlier by Chandu Lal from the late 1840s until 1853. The challenge was to secure loans from bankers to pay administrative officials and military troops, then to parcel out land from which the bankers and their forces could collect the debts the state could not repay, and finally to fend off the mercenary forces demanding payment for their services from nobles, bankers, and land revenue contractors. Just before the young Salar Jung became Diwan in 1853, the Company forced the Nizam's cession to it of the rich province of Berar, the strategic revenue-rich 38 I currently have under review a paper titled, "Palmer and Company: An Indian Banking Firm in Hyderabad State." 39 The governor general initially declined to interfere, but when the resident persisted and the claim of the firm was found to be a crore of rupees, Lord Hastings insisted on its liquidation. The Nizam signed a deed relinquishing in perpetuity his claim to seven lakhs of rupees payable annually as peshkush for the Northern Circars. When the governor general received this deed from Chandu Lal, he remitted a crore of rupees to Hyderabad to settle the claims of Palmer and Co. However, these actions were censured and prompted an enquiry in England, leading to the firm's bankruptcy in 1824: Peter Wood, "Vassal State in the Shadow

H Y D E R A B A D ' S M U S L I M A N D H I N D U B A N K E R S
While the absence or near-absence of Muslim bankers is often attributed to the Islamic prohibition on the taking of interest, 42 Muslims were bankers and traders in the subcontinent and in Hyderabad. One explanation offered for Muslim under-representation in the management of large Indian firms stresses not just Islamic inheritance laws but enforcement of them under British rule, 43 but this would not explain the scarcity of Muslim bankers in Hyderabad, a state outside British jurisdiction. Muslim bankers were present and active in Hyderabad in the late eighteenth and early nineteenth centuries, but they did not establish family firms that lasted. 44 Afghan or Pathan moneylenders were sometimes mentioned as loaning money to state officials and nobles but were not prominent as bankers. 45 Khoja and Bohra Muslims, from formerly Hindu caste communities that continued to follow Hindu law and so were important as merchants and bankers in western India, 46 were not leading bankers in Hyderabad, although there were Bohra families settled in the old city locality of Husaini Alam.
Tantalizingly, it seems that the original "Panch Bhai" (five brothers) bankers of Hyderabad in the late eighteenth century were Muslims, although later these roles were filled by various Hindus. These Muslim Panch Bhai  [1911][1912][1913]. 45 In the 1820s, for example, an anonymous defender of Palmer and Company stated that the Nizam's government credit was so low that the Minister was unable to secure funds from "the native bankers" and had to turn to "Gosains or Patans": Anonymous, "Further (Ashurkhanas are Shia religious buildings housing alams, standards symbolizing the martyrdom of the Prophet's grandson Hussain in 680 CE that are taken out in procession during the annual Muharram commemorations of that martyrdom.) Located in Karwan (the bankers' locality) and registered as a wakf in 1940 by its Muslim mutawalli, or manager, who lived adjacent to it, this ashurkhana had the Nizam's name substituted as manager in 1958. 48 Whether these two sites were connected to the same family or families is unclear.
Then there was Makhdum Seth, known throughout India as the most famous of Hyderabad's Panch Bhai bankers early in the nineteenth century, and foremost among the bankers and grain merchants living in Begum Bazar. His son, Syed Ahmed, took an interest in military affairs; he dressed as an Afghan and maintained many Afghan, Arab, and other troops as well as seven horses and two elephants, hoping to be appointed a jamadar. Failing in this goal and in business, he disappeared from the banking scene, his kothi (residence/office) taken over by Pestonjee, a Parsi banker from British India, who by the early 1840s was paying Syed Ahmed 100 rupees a month rent for it. 49 In 1846, Syed Ahmed made a final attempt to collect debts he thought due him from nobles, traders, and bankers: he held a five-day party for young relatives of leading bankers and officials and then locked them up and held them hostage, demanding money for their release. The attempt ended badly for Syed Ahmed, since the Nizam sent troops to join with those of the families of the imprisoned youths and they were freed. Then the Diwan Siraj ul Mulk imprisoned not only Syed Ahmed but also eleven of his relatives, including-unusually, the source notes-his mother, wife, and sister. 50  sometimes defied caste or community stereotypes. For example, one Hindu Marwari, Raja Shambu Pershad Agarwal, in fact a military man, had converted to Islam after having founded a Hindu temple. 51 The Gulzar-i Asafia, a Persian history, describes Hyderabad's evolving banking communities and their localities in the early nineteenth century and highlights non-Muslims. It lists ten bankers, five in Karwan (all Gujaratis), four in Begum Bazar (three Goswamis and a Marwari), and one in Chaderghat, or Residency Bazar (a Parsi): "Side by side in Begum Bazar are the houses of the Marwaris, Gosains, Komatis, Afghans, and other financiers and traders … and in Karwan Sahu, the Gujaratis reside.… The bankers are millionaires, lending millions of rupees to the state and financing the land revenue contractors.… [One of them, a Gosain] associates with the Afghan military leaders and is fond of fighting.… The bankers also loan to the nobles and to the Nizam himself; they have access to the Court." 52 The case studies that follow highlight Gujarati, Goswami, and Marwari adoption, marriage, and inheritance patterns in Hyderabad and provide empirical details relevant to the assumptions and questions raised in the introduction. They also highlight a few instances where women played important roles in the family firms-we are interested in practice rather than legal norms. Given the provisions of Hindu Mitakshara law, Claude Markovits, for example, found a surprising incidence of women recognized as heirs in 118 succession cases among Shikarpuri Sindhi merchant court cases in the 1890s. 53 The first case study is of the Gujarati firm of Benkati Das Bhagwan Das, dating from the 1750s. This firm's fortunes were connected to the well-known Eurasian firm of Palmer and Company, discussed above. The second is of Goswami firms, in particular that of Umrao Gir/Narsingh Girji/Dhanraj Girji, dating from 1801, and the third is of the Marwari Agarwal firm of Mahanand Ram Puran Mal Ganeriwala, dating from 1802. These three firms, like some others from the late eighteenth and early nineteenth centuries, are all represented by descendants of one sort or another in Hyderabad today. And that is my point: their longevity. of this family functioned as kotwal, or mediator, for all Gujaratis in Karwan. Karwan was dominated by Gujaratis, although others were also settled there. 59 This Gujarati family firm all by itself illustrates many issues of interest here: adoption, flexible use of kinship ties, mistrust or rivalry among relatives, and, in oral accounts of the firm's affairs, leading roles for women. Oral interviews trace the family from Ahmedabad, Gujarat to Aurangabad and Hyderabad, with marital connections in Varanasi, Burhanpur, and other places. Benkati Das (grandson of Govardhan Das of the garden) was forced from the Aurangabad branch of the firm when it was looted and his father Jagmohan Das died. He and an older brother, unnamed and untraced, moved to Hyderabad's Karwan locality, but Benkati Das went on to Calcutta and supplied goods to Europeans. The story goes that a mutiny took place there in the late eighteenth century, and Benkati Das helped some Europeans escape by ship and gave them provisions. 60  Das' descendants who represent this famous firm in Hyderabad today. 64 They preserve not only an original partnership agreement with Benkati Das' signature in Hindi script, but also, in their puja room, William Palmer's silver sugar bowl and creamer, which bear his name and the date 1817. 65 Genealogies provided to me in 1983 for the Benkati Das family and six or seven other families relate leading Gujarati family firms to each other (Hari Das Kishen Das, Lachmi Das Lachman Das, and later permutations of these). These genealogies show the family firms responding to mortality and age structure, passing leadership to sons, cousins, nephews, or relatives through affines to carry on entrepreneurial activities. Whether or not actual adoptions were involved is not always known (only one was explicitly mentioned in interviews), but the flexible family strategies are very clear. 66 Benkati Das of this Govardhan Das/Bhagwan Das family, establishing himself in Hyderabad in the early nineteenth century, brought his sister's two sons from Burhanpur. 67 , 194). 65 The Nizam rewarded Palmer for opposing a plan for British administration of the state by giving him a monthly allowance for his own support and the education of his two sons in England: "Palmer Histories," India Office Library, Mss., 7-8. In the mid-twentieth century, Palmer's unmarried granddaughter Mabel was still getting a pension of 100 rupees left for her by him: email from Bruce Cox of Melbourne, Australia, a great-great-grandson of Palmer, 2 Feb. 2008. 66 B. R. Agarwala wrote about his own Marwari community that not only lineal descendants but also sons-in-law, brothers-in-law, and maternal relatives formed kindreds that commanded capital and industries: "Caste in a Mobile Commercial Community," Sociological Bulletin 4 (Sept. 1955): 138-45, 141. 67 K. Krishnaswamy Mudiraj, Pictorial Hyderabad (Hyderabad, 1929(Hyderabad, , 1934, states that this family went from Modhera or Morera, Gujarat, to Delhi, and then to Hyderabad: vol. 2, 497. young sons to start over on their own, and his sister's sons Kishen Das and Hari Das assumed leadership roles before Benkati Das' young sons matured. Hari Das headed the community in Karwan in the 1830s, when it was divided between the residents of the upper and lower bazaars and there were often fights, especially on the occasion of Holi: "Injuries and even deaths were common, and after each year's battles people prepared for the next year. But Hari Das Seth called the inciters from the two sides to his house and gave them shawls; he forbade fighting and got trusted people as mediators. The people were thus saved and protected, and this wonderful Seth, devoted to the welfare of the masses, would soon get to see the Nizam. Oral accounts give considerable agency to the women in these families, and indicate their high status by calling several of them Raja (e.g., Chand Raja) and identifying many wives as sisters of so and so, showing the importance of brothers-in-law. When Lachmi Das and Lachman Das were kidnapped in 1853 (by the military? a Sikh mutiny?), their mother Gangabai was the one who paid the ransom. 71 All of these twists and turns certainly suggest that trust, so often easily attributed to Hindu family firms, needs to be examined more closely. Many of the men had more than one wife, sometimes at the same time, suggesting that multiple wives as well as adoptions served the strategic end of securing living sons. The material also suggests that both patrilineal and affinal kin competed at times but also continued each other's claims when appropriate.
The British Residency, or Sultan Bazar locality, attracted increasing numbers of bankers from Karwan and Begum Bazar and the old city from the 1820s since the British Residency area offered them protection from the Nizam's desperate requests for more money. 72 A competing British Indian legal system was developing there, one that enjoyed substantial military backing from nearby Secunderabad and in British India. The Benkati Das cluster of closely related Gujaratis in the banking business shifted residences and temples (their Gokulnath temple and the private Bhagwan Das family Giriraj temple) from Karwan to Sultan Bazar in about 1900. Also in the early twentieth century, part of the family property went under the Court of Wards administration because, although Raja Bhagwan Das' son Vallabh Das was living, he was a minor at the time of his father's death. We will see this pattern in the Hindi family histories cited below of sons inheriting while still minors, and this reminds us that a possible source for tracing Muslim banking families would be the Court of Wards records.

G O S WA M I R A J A S : U M R A O G I R / N A R S I N G G I R J I / D H A N R A J G I R J I
The second case study is of Goswami Rajas, as leading bankers from the Gosain or Goswami sanyasi community were called in Hyderabad, and particularly of one leading Goswami math or establishment. Bernard Cohn suggested in a 1964 article that Gosains might have an advantage over Hindu family firms because their inheritances were not partible-one chela got most of the assets. 73 What this seems to have meant, at least in Hyderabad, was that contests among chelas for the resources of the mathdars, or heads of maths, were serious affairs. In Hyderabad, there were sixteen or eighteen Goswami maths located east of Karwan in Begum Bazar, Shah Inayat Ganj, and down along the northern bank of the Musi river. These were loosely grouped into at least two factions, with contested rankings and internal divisions as well. Like other bankers, the Goswami Rajas began by the mid-nineteenth century to engage with the British resident and British Indian notions of Hindu law, but they continued to rely on mercenary troops to guard themselves and their interests.
Contemporary accounts of a mid-nineteenth-century inheritance dispute between two Goswami chelas vividly illustrate the political context in which Goswamis and other bankers operated. This dispute was between two chelas, the elder one, Baldevgir, claiming to be sole inheritor of his deceased guru's property, but challenged by the younger one, Bijigir. The sources here, Englishlanguage newspapers of the mid-nineteenth century, reflect some British terminology and understandings of the community and the issues involved. These two Gosains quarreled about "a patrimony," and in the fall of 1849 their armed retainers battled in the midst of Begum Bazar, leaving two killed and six wounded. Bijigir had employed a party of Rohilla mercenaries and promised to pay them from his "share" once it was secured. The Rohillas wounded Baldevgir's "adopted son," or chela, and carried him off as a prisoner. Then the Rohillas demanded their pay, besieging Bijigir as well as Baldevgir and driving both Goswamis to seek protection in the Residency Bazar. A few months later, in May or June of 1850, the Rohillas held both Baldevgir's chela and Bijigir as hostages for their pay, but the Nizam's government paid some money and got them released. The Minister (Diwan) then confined Baldevgir and Bijigir in his own palace, so the Rohillas besieged that palace and demanded that the Gosains be turned over to them. Failing to achieve this, they seized a "person of note," namely Jai Gopal, the son of the leading Marwari Agarwal banker Govind Ram, and held him for a ransom equal to the pay they were owed by Bijigir. Jai Gopal was seized within two hundred yards of the Residency Bazar where he resided, leading to calls for the resident to intervene. Finally, Jai Gopal was released in exchange for Bijigir and Baldevgir, but the government sent troops against the Rohillas right in Begum Bazar, assisted by the armed guards long retained by these rich Goswami bankers. 74 This exciting and ongoing contest shows the growing engagement 74 These accounts are in Ali, Hyderabad Affairs, V: Madras Spectator, 3 Oct. 1849, 646-67; Englishman, 16 Oct. 1849, 647; Englishman, 26 Oct. 1849, 649; Englishman, 24 June 1850, 664. The two were still fighting each other in 1853, as the "old quarrel" caused fresh bloodshed: Bijigir sent two hundred Sikhs to attack Baldevgir's house in Begum Bazar, which was defended by ten Arab retainers. When one of the Arabs was killed, one hundred Arab reinforcements showed up and drove the Sikhs away, wounding and capturing Bijigir in the battle. The journalist opined that the government, having awarded the property to Baldevgir for "a consideration," might now give Bijigir a share for a similar payment: Englishman, 18 May 1853, 746. of both the Nizam's government and the British resident with questions of inheritance and jurisdiction in Hyderabad's urban localities. 75 Both governments sought to maintain law and order and to retain the allegiance of important bankers in the city, whether they resided in Begum Bazar or the Residency Bazar.
Goswami gurus and chelas were using both customary and British Indian law to claim successions and inheritances by the early twentieth century, and a gradual shift from sanyasi to Hindu family law took place in Hyderabad as a Goswami caste was constituted there in that period. One lineage or banking firm among the Hyderabadi Goswamis, that of Umraogir/Narsing Girji/ Dhanraj Girji, illustrates these changing community practices. Moving from Gandheri in northern India to Pune under the early Peshwas, and then to Hyderabad in 1801, the line of Dhanraj Girji and his older brother Pratap Girji testifies to the community's transition in the early twentieth century from guru-chela lineages to biological lineages, from mathdaris, or ascetics to grihasthas, or householders. Shejras, or genealogies, for this and other Goswami maths trace guru-chela relationships over time, but dates are rarely given and it is difficult to calculate the rate of turnover, of mortality, for these mathdars. Three shejras with nine-to-thirteen generations, and some indication of dates, suggest that gurus presided over their maths for about fifteen years. Conflict and deaths saw the lineages dividing again and again. In the late nineteenth century, Hyderabad's Goswamis were grouped into two dhadas, or factions: those of Umrao Gir/Narsing Girji/Dhanraj Girji and Narpat Girji/Lal Girji/ Birbhan Girji. 76 These sanyasis were supposed to be celibate and unmarried, but women were present as "keeps" or "garlanded wives," not as "real" wives. The custom was for the mathdars or gurus to appoint each other's real sons as chelas to each other, providing for both biological and "spiritual" heirs as best they could. Women associated with the Goswamis sued for inheritance, too, much as they did in British India. 77 The garlanded wives generally lost in Hyderabad, as the mathdaris gave way to the grihasthas, the latter drawing "real" Goswami women from northern and central India where the trend to householdership had taken hold earlier. Again the issue of trust comes up, as my Goswami informants traced the guru-chela lineages on shejras against the real biological lineages on other shejras and discussed past inheritance conflicts. Choices had to be made, and Dhanraj Girji became the head of his math (and one of the richest men in Hyderabad and Bombay in the twentieth century) only because his elder brother Pratap Girji married according to Hindu law early in the century. Dhanraj Girji had garlanded wives and children but did not marry them under Hindu law, so Pratap Girji lost his claim to the math because succession was still decided according to sanyasi customary law. Moving to Hindu law was preferable, I was told, because "the marriage system was better than the shishya [chela] system; there were fewer disputes under Hindu law." Fewer disputes perhaps, but there were still disputes among Goswami family members. For the manja requisites sent to the court in 1839: ibid., 217-18. Puran Mal sent his gifts on 20 August, the day of the mehndi ceremony, and Kishen Das, Lachmi Das, and Jaganath Das, Gujaratis of Karwan, sent theirs the next day, when the bridegroom left for the bride's residence. 80 A jagir is a conditional or unconditional land assignment from which the holder collects the land revenue. An inam is a hereditary, rent-free land grant. earlier period, older than the temple and sharing a stepwell with it. 81 An 1894 newspaper article about a controversy over who should manage the mosque says the Nizam himself attended the laying of the temple's foundation, but the Agarwal Hindi history makes no mention of this. 82 The family lived within the temple compound, and two impressive samadhis, or tombs, of Mahanand Ram and Puran Mal were erected on the temple grounds. 83 In this Marwari genealogy, I found only the one legal adoption (and it worked against the adopted son, who later tried to claim the jagirs but had been adopted out of the "natural" line of descent). 84 However, the men often had more than one wife-two or three each-the women coming from Hyderabad's old city localities of Char Minar and High Court and also from Nagpur, Ajmere, Calcutta, and Gaya Bihar. Again, having multiple wives was another way of securing heirs and increasing one's marital and financial networks. One can make the obvious point that this option was even more available to Muslims, if they did practice polygyny.
While the Hindi history of this Ganeriwal family mentions no women, family oral tradition credits women from the Mali (gardener) community as having helped connect Puran Mal to the Nizam directly. The story is that a Mali woman and her descendants worked for the Ganeriwal family, and the Nizam had a "keep" who was the daughter of a Mali woman working for the Ganeriwals. The latter came to visit her mother in the Ganeriwal home and reported back to the Nizam that she had seen her grandfather (Puran Mal or perhaps Mahanand Ram) and that he was "good with money." Therefore, the Nizam appointed Puran Mal Treasurer in Berar, and Puran Mal's son Prem Sukh Das collected revenue there in the midst of Nizam/Maratha conflicts. This case again suggests that wives and women could have significant roles, 81 Alka Patel and I found the same situation at the Jham Singh Deval in Hyderabad, where an earlier mosque on the grounds was hidden behind a new and ugly rathkhana (carriage house) and disavowed by temple authorities in January 2009. 82 "Sitarambagh Temple Case," Evening Mail, 17 Apr. 1894, in the Andhra Pradesh State Archives Clipping Collection. "A respectable Hyderabad merchant" had been preserving the mosque and appointing Muslim managers for some seventy years, but then an old moulvi claimed it and the Hyderabad government got involved, leading to petitions submitted to the Nizam, the resident, and the viceroy by leading merchants. For more details (but no resolution), see also Bombay Gazette, 12 Apr. 1894; Deccan Budget, 13 Apr. 1894; and Hyderabadee, 30 Apr. 1894, all in the Andhra Pradesh State Archives Clipping Collection. 83 Family members today continue to maintain property, including temples in Rajasthan, in Lakshmangadh and Pushkar. 84 Since the jagirs were in Berar, ceded to the British to support the Hyderabad Contingent in 1853, the cases went through the Court of the Resident in Hyderabad in 1896: installment 9, list 1, serial number 26, D2/d1, Religious Affairs, Endowment Department, "Sitaram temple complaint of mismanagement," Andhra Pradesh State Archives. Also in the Archives, see installment 31, list 9, serial number 663, L5/b945, settled in 1933 by the Commissioner of the Berar Division of the (British) Central Provinces; and the 1307 Fasli file titled "Sita Ram Bagh," in the Amur-i-Mazhabi files.
serving to relate bankers to potential partners or clients and, in this case, to the Nizam himself. mention no adoptions, and the other eight record sixteen. Of the forty-two Maheshwari family histories, of from one to eleven generations, twenty mention no adoptions, and twenty-two mention a total of thirty-five. Of the thirty-two Agarwal family histories (including a few Digamber Jains), ranging from two to nine generations, there were thirty-seven adoptions in nineteen families. In sum, forty-nine of these ninety Marwari families migrating to the Nizam's state give histories that include adoptions, eighty-nine in all. In some families there were as many as four or five adoptions, most of them being a second or younger son of a younger brother given to an older brother or cousin brother.
Other relevant features of these histories concern age at marriage, age of becoming head of a firm, and divisions within family firms. Sometimes mentions are made that a man married at or before puberty (perhaps because this was unusual), and of boys taking on business responsibilities at the ages of nine to twelve (again, perhaps since it was unusual). 89 The histories routinely specify the year of division for family firms, usually between or among brothers. No reasons are given for these divisions, although they seem to be associated with the number of surviving males in the family and with expansion into new territories or enterprises. Scholars sometimes state or imply that trust required keeping family firms undivided across generations, reading divisions among brothers or other descendants as negative indications of a family firm's healthiness. 90 But in these texts such divisions were not viewed as negative events or breaches of trust. This raises questions about whether non-partibility was necessary for the continuity and prosperity of family firms.
Only eight women appeared in these family histories. In one Oswal family, a nineteenth-century widow, whose son was only two at her husband's death, carried on the business until she herself died. 91 Women are mentioned in five Maheshwari family histories. In the first case, "the ladies of the family" went to Khandesh and worked successfully to raise money to free their ancestor, whom the ruler Mewat had imprisoned after the firm had been looted. In another case, the mother of a leading banker was famous for her charity, reportedly rivaling Hyderabad's Diwan Chandu Lal in her openness and generosity. In the other cases, a nineteenth-century widow saved the firm by carrying on the business while her son studied, and two early-twentieth-century educated women became community and political activists. 92 In one Agarwal family, a nineteenth-century widow raised her young adopted son with the advice of the firm's accountant. In the same family, a second wife, widowed in the early twentieth century at age twenty-nine, took over the business and ran it efficiently; she directed factories and cotton mills and managed the accounts and books, while educating her three daughters and marrying them well. 93 Most of these Oswal, Maheshwari, and Agarwal family histories feature, for early-twentieth-century members, activities linked to community welfare and even Indian nationalism, and educated women were beginning to participate in those activities.
C O N C L U S I O N Patterns of adoption, marriage, and inheritance among the predominantly Hindu and Jain banking communities in Hyderabad State show family histories interacting with changing political and legal regimes. Recent work by Kuran and Singh relates normative Islamic and Hindu inheritance laws 94 to the relative continuity of Muslim and Hindu family firms. While the possibilities of such links are promising (particularly with respect to adoption, not considered by Kuran and Singh), much more study is needed. Their work rests on several assumptions: first, that both Hindu and Muslim financial firms actually followed Hindu and Islamic inheritance law when it came to succession; second, that the common Muslim practice of cousin marriages did not effectively counteract the divisive effects of Islamic inheritance law; and third, that the striking success of the Khoja and Bohra Muslim communities was due chiefly to the continued adherence of these endogamous communities (once Hindu castes) to Hindu law with respect to inheritance. 95 Further, while the Hyderabad evidence testifies to the frequency of adoption in Hindu and Jain financial communities, it also shows that multiple wives and the flexible incorporation of at least some affines into family businesses helped supply male heirs to the family firms. For Muslims, multiple wives would be even more probable, and while this would have supplied heirs, as adoption did for Hindus and Jains in the days before the demographic transition, it would also have increased the partibility of estates according to Islamic law. Conflict between or among family members (or chelas in the case of Goswamis) contributed to some divisions of Hindu and Jain firms, but divisions also occurred because of the numbers of surviving males and/or the expansion of the business into new territories or enterprises. Finally, oral and written evidence revealed a small number of women achieving public recognition for one reason or another in the Hindu and Jain communities; research on women in Muslim mercantile communities is needed.
Scholars continue to compare European and Asian economic histories, 96 and family firms are central to these histories. Claude Markovits argues that Asian merchants were "some kind of capitalists" and writes that the South Asian merchants were able "to maintain significant areas of independent international operations throughout the period of European economic and political domination in Asia." 97 Jack Goody contends that in both South Asia and Europe kinship groups and extended domestic units often played critical roles in commercial and industrial activities both before and after the development of capitalism. 98 Francesca Trivellato in her work on the early modern Sephardic Jewish trade diaspora relates Jewish and Christian marriage, dowry, and inheritance laws and practices to evolving legal norms, 99 and frequent marriages between close relatives are credited with insuring the continuity of the famous Rothschilds family banking firm. 100 As scholars look again at South Asian banking and mercantile families and firms, the material presented here highlights the significance of marriage, adoption, and inheritance practices within communities, and whether or not such practices were regulated by the state in any way. Trivellato's phrase "communitarian cosmopolitanism" well describes the operation of Hyderabad's banking and mercantile firms, people of diverse traditions engaging with each other in a plural society organized around the Nizam's court in precolonial times. Hyderabad's bankers and merchants relied upon state power and responded to the constraints and opportunities offered by the state and its officials. 101 By the late nineteenth century, however, new legal arenas and practices offered by the British Residency and British India beyond the state's boundaries competed with those available through the Nizam's administration. 102 Indian litigants in colonial India took advantage of new resources and participated in shaping new legal regimes rather than being dominated by them, in processes that were complex and multi-layered. 103 How mercantile families and communities were actually constituted, transmitted their resources, and participated in changing legal regimes still needs far more historical study, in both British India and leading native states like Hyderabad.