Epidemic dropsy, its origins, and its spread in India and Bengal

After an individual is infected, the first symptoms of epidemic dropsy usually appear after five days. They include diarrhoea and vomiting, which lasts between a few days and a week. Other early symptoms can include fever, lower backache, and hair loss. Subsequently, between around nine days and two weeks, oedemas form at the lower end of the limbs. They are regularly painful, burning, and itching and become worse upon the infected individual standing. Other related symptoms can include anaemia, palpitations, and breathlessness. In extreme cases, it can cause heart failure, pneumonia, respiratory distress syndrome, and renal failure. The mortality of untreated infected people is usually around 5%.Footnote ¹³ The first confirmed epidemic occurred in Calcutta in 1877–80, although unconfirmed reports suggest it may have been present on a French ship carrying indentured migrants in 1863 and in northeastern India since the 1866 Orissa famine, which also affected Bengal.Footnote ¹⁴ Since then and up to the end of the twentieth century, outbreaks of epidemic dropsy were a regular occurrence in different parts of India, most recently around New Delhi in 1998.Footnote ¹⁵ Epidemics are also known to have occurred in Myanmar, Fiji, and Mauritius.Footnote ¹⁶

The cause of epidemic dropsy, which was confirmed in the 1940s, is the ingestion of edible oils that have been adulterated with Mexican poppy (Argemone Mexicana) oil.Footnote ¹⁷ Specifically, it is caused by two toxic alkaloids, sanguinarine and dihydro-sanguinarine, contained within Mexican poppy seeds, which are used to make oil. The Mexican poppy originates in Central America but it became naturalised throughout the tropics following European expansion across the Atlantic from the late fifteenth century. It was probably brought to the Indian subcontinent via Portuguese, Dutch, and/or English ships. By the late nineteenth century, it was prevalent throughout Bengal and across northern and southern India.Footnote ¹⁸ It grew (and continues to grow) wild and favours sandy and rocky soil, including on roadsides and in dried riverbeds. In the nineteenth century, there were debates within India’s colonial medical and botanical establishment about Mexican poppy oil’s uses, particularly in terms of its possible medicinal and/or narcotic qualities. Several sources also indicate its usage in local medicinal practice, and at least one source from Cuttack, Orissa noted that it was ‘mixed with mustard seed as an adulteration’.Footnote ¹⁹

Undoubtedly, this mixing of Mexican poppy and mustard seeds/oils led to the 1877–80 epidemic. Mustard oil was and remains the premier cooking oil in South Asian cuisine, especially in Bihar and Bengal. However, it is regularly difficult for both vendors and consumers to differentiate adulterated from unadulterated mustard oil. The seeds and oils of both plants are similar, with the seeds being small, round, and dark and the oils a clear yellow. Thus, accidental adulteration is possible, especially when mustard and Mexican poppies are grown in proximity.Footnote ²⁰ Mustard plants (Brassicaceae) are also less hardy to water stress than Mexican poppies, and they are susceptible to several diseases and pests. Therefore, the care and environmental conditions required to ensure a good crop may have incentivised producers and/or vendors to adulterate their mustard oil with the oil of the weed-like Mexican poppy, as the latter is regularly more available. Indeed, drought and its effects on agriculture during the 1866 Orissa famine may tentatively be suggested as a trigger for adulteration becoming more widespread. This hypothesis was made more possible for the late nineteenth century when the link between using Mexican poppy oil as an adulteration and epidemic dropsy was unknown.

The 1877 outbreak almost certainly originated in Bengal. This statement contradicts that of David Arnold, who suggested in 2010 that it first broke out in the Madras Presidency.Footnote ²¹ His argument was based on the 1886 report of Norman Chevers.Footnote ²² However, Chevers was one of several colonial medical officers in India who, until the 1930s, argued that epidemic dropsy was a form of beriberi. The latter disease is caused by thiamine (vitamin B1) deficiency, and its symptoms regularly include oedemas similar to that of epidemic dropsy. Beriberi was rife in the Madras Presidency in 1876–78 due to the famine conditions caused by disastrous colonial grain and labour policies in the context of severe El Niño-related drought.Footnote ²³ There is no known evidence (apart from that provided by colonial medical officers who conflated it with beriberi)Footnote ²⁴ that epidemic dropsy caused by ingestion of adulterated cooking oil broke out in the Madras Presidency in these years, even if drought and starvation probably limited the harvest of mustard seeds. Additionally, even in years of regular climate, Calcutta did not regularly import mustard oil from the Madras Presidency. Rather, Calcutta’s supply came from the ‘brown mustard’ (Brassica Juncea) grown in rural Bengal and Bihar, supplemented by other varieties and rapeseed oil brought via the East Indian Railway.Footnote ²⁵ Thus, Arnold’s suggestion based on some colonial medical officers’ reports that epidemic dropsy broke out first in the Madras Presidency is probably a result of misdiagnosing cases of beriberi and other famine-related diseases.

Nevertheless, the wider enviro-climatic conditions affecting much of the rest of colonial India increased the likelihood of an outbreak in Calcutta in 1877. Despite regular rainfall and agricultural production in Bengal in 1876–78, colonial responses to dearth in wider India, especially in the Madras and Bombay Presidencies, contributed to the price of staples doubling in Calcutta between 1875 and 1877.Footnote ²⁶ Additionally, epidemiological conditions in the city were generally adverse, with epidemics of cholera, malaria, and smallpox all breaking out between 1876 and 1880. According to colonial statistics, there were 7,253 deaths from cholera in Calcutta in 1877–80, as opposed to 1,693 in the four years preceding (and 958 of those were in the first six months of 1873). The 1877–80 cholera epidemic also spread to wider Bengal, especially along the Ganges and Brahmaputra rivers and the East Indian Railway.Footnote ²⁷ These economic, social, and epidemiological conditions heightened levels of desperation, especially in the poorer suburbs. Even with what was likely a regular harvest of mustard seeds in rural areas, they provided an ample incentive to adulterate mustard oil with Mexican poppy oil, as doing so extended the supply of cooking oil and reduced the need to purchase more at inflated prices.

After the initial outbreak in 1877, the epidemic made its appearance every year until 1880. Its prevalence, however, was not continuous. Rather, cases and deaths rose and declined according to the seasonal cycle. According to Kenneth McLeod, a colonial medical officer who was present in Calcutta at the time of the epidemic, it broke out every year in 1877–79 towards the beginning of the cooler, dry season (October–April). Its earliest recorded occurrence in any year was in 1878, on 4 September. Thereafter, cases almost ceased entirely at the onset of the warmer, monsoon season (May–September).Footnote ²⁸ Although McLeod did not make the connection, this pattern aligns with the agricultural season for mustard grown in Bengal and elsewhere in northeastern India.Footnote ²⁹ Mustard is sown in Bengal in September–October and is harvested in January–March, depending on the sub-region and strain.Footnote ³⁰ Thus, fresh mustard seeds and oil tended to be in their shortest supply towards the end of the southwest monsoon season. Some evidence from wider colonial Bengal suggests that the price of mustard oil could nearly double from the beginning of the monsoon season to the drier season.Footnote ³¹ Adulterating it with Mexican poppy oil by the latter time, especially in the context of higher prices and the desperation that characterised late-1870s Calcutta, could have been a strategy used by producers and/or vendors to try to keep prices down.

The spatial distribution of cases of epidemic dropsy also varied from year to year, except that it was most prevalent in Calcutta’s poorer suburban areas throughout the epidemic (see Figure 1). In 1877–78, it was largely confined to the southern suburbs in and around Garden Reach, on the eastern bank of the Hooghly River. Being the location of the port, this was a major transit area for goods and people travelling between the hinterland, the city, and the wider world. In 1878–79, it broke out in the same places and spread to neighbouring districts, namely Khidirpur, Bhowanipore, and Ballygunge. Finally, in 1879–80, it broke out across Calcutta’s southern suburbs, as well as (for the first time) in parts of the southeastern town area and the eastern suburbs of Beniapukur and Sealdah. It was also present in Shillong and some other locales in Assam in 1878–79 and 1879–80, as well as in Dhaka, eastern Bengal (present-day Bangladesh) in 1879–80.Footnote ³² These patterns suggest that the practice of adulterating mustard oil with Mexican poppy oil was limited to only a few suppliers/vendors in specific locales. However, it became more widespread as high prices took their toll over time. The end of the epidemic followed the re-establishment pre-famine food prices in 1879–80.Footnote ³³

Figure 1.

Map of Calcutta and suburbs with areas with reported infections highlighted according to year. Digitised map based on E.P Richards, ‘City of Calcutta’ (1913): https://curiosity.lib.harvard.edu/scanned-maps/catalog/44-990114901360203941 [accessed: 12 June 2024].

Notwithstanding these broad trends, it is difficult to assess the actual numbers of cases and deaths from epidemic dropsy in colonial Bengal and India during these years. As mentioned, many among the colonial medical establishment were slow to identify epidemic dropsy as a singular disease, conflating it with other diseases, especially beriberi, instead.Footnote ³⁴ According to McLeod, meanwhile, colonial medical practitioners in Calcutta regularly ascribed deaths from what he believed to be epidemic dropsy to ‘fever’, thus mixing statistics for epidemic dropsy with those of malaria (which was also epidemic in Calcutta at the time) and other diseases.Footnote ³⁵ McLeod also pointed to a small increase in the number of deaths attributed to dropsy (as opposed to epidemic dropsy), a generic disease that can be caused by phenomena, including malnutrition, other than consumption of adulterated cooking oil.Footnote ³⁶ Some of these cases may also have been beriberi, especially as some famine-affected individuals travelled from the Madras Presidency to Calcutta in search of food and/or wages.Footnote ³⁷ Thus, McLeod was probably correct in his assessment that statistics for ‘dropsy’ probably did not reflect the severity of epidemic dropsy in Calcutta in these years.Footnote ³⁸ There are thus no official tabulations of cases and deaths from epidemic dropsy in Calcutta (like there are for, for example, cholera), however flawed such official tabulations regularly were.

Even so, some measure of the spatial severity of the epidemic is still gleanable from colonial sources. Using police reports, McLeod suggested a very high death rate in and around Calcutta: about 20% in the town area and 44% in the suburbs. He also estimated a 19% death rate from his own investigations in nearby villages.Footnote ³⁹ The recorded death rate was much lower outside of Calcutta and its environs, with apparently only a total of three people succumbing in Shillong.Footnote ⁴⁰ This is probably reflective of two phenomena. First, the recorded death rate for Calcutta is probably inflated. Most cases, especially milder ones, were probably not reported or treated where either the police or McLeod could count them. Also, both counts probably included other diseases that were epidemic in Calcutta during the late 1870s, including malaria.Footnote ⁴¹ Second, despite these qualifications, the death rate in Calcutta was still probably higher than it was elsewhere in Bengal and northeastern India. Unsanitary conditions, the effects of other diseases on the immune system, and diminished diets caused by inflation probably reduced infected individuals’ capacities to recover, especially among the poor.Footnote ⁴² The specific socio-economic conditions that arose from colonial responses to El Niño-related drought heightened the chances of an outbreak of epidemic dropsy in Calcutta in 1877, and they may have increased the vulnerability of some suburban populations to death after infection.

The spread of epidemic dropsy to Mauritius in 1878–79

Epidemic dropsy broke out in Mauritius in 1878–79. Indentured labourers who had come to Mauritius from Bihar via Calcutta were recorded as being the most affected population. The system of indenture was instituted in 1834 following the abolition of slavery in the British Empire and lasted until the 1910s. Although often brutal and having many of the same characteristics of plantation slavery, indenture was partly distinguishable from slavery because it was cemented with a written labour contract.Footnote ⁴³ During the indenture period, just under 500,000 labourers arrived in Mauritius from colonial India and Bengal, largely to maintain and expand the monocrop sugar economy. In 1878–79, 5,216 indentured migrants made the journey.Footnote ⁴⁴ Of those, 3,092 departed from Calcutta (2,262 in 1878 and 830 in 1879).Footnote ⁴⁵ Disease has been a common theme of historical analysis in this context, notably with case studies focusing on the spread of cholera in 1819 and 1854–56 and malaria in 1864–67 from India.Footnote ⁴⁶ Such analyses refer to the spread of disease on ships and within Mauritius via human-to-human transmission and the spread of vectors, notably Anopheles mosquitos. This section furthers this historiography by examining another way in which disease spread between these two British colonies: the trans-shipment of food, specifically the trans-shipment of adulterated mustard oil.

Since the cholera and malaria epidemics of the 1850s and 1860s, the colonial medical establishments of India and Mauritius were highly conscious of the dangers of disease transmission via the indenture system. Arrival of infected migrants and infection of established labourers were bad for business. Thus, each colonial government instituted measures to limit the occurrence of trans-colonial epidemics. Recruits in India underwent a medical examination by the civil surgeon before the signing of the indenture contract and before embarkation.Footnote ⁴⁷ There were also quarantine stations on Mauritius’ outer islands, including at Gabriel Island and Flat Island, as well as at Pointe aux Canonniers, if cases of epidemic disease were identified aboard ship.Footnote ⁴⁸ Then, once ships arrived in Port Louis, the Depot Medical Officer vaccinated unprotected migrants from smallpox and made further checks for outbreaks of other diseases.Footnote ⁴⁹ It was only after these several rounds of checks that newly arrived indentured migrants were taken to sugar estates.Footnote ⁵⁰ The success of these measures was such that, in 1877–80, neither cholera nor smallpox broke out in Mauritius, despite them being epidemic in Calcutta at that time. Colonial Mauritius’ Depot Medical Officer reported ‘no epidemic sickness’ brought via new arrivals of indentured labourers in 1879.Footnote ⁵¹

Of course, this does not discount the possibility that at least some of the just over 3,000 arrivals from Calcutta in 1878–79 were infected with epidemic dropsy before they arrived in Mauritius. They boarded ships in Bhowanipore, one of the epicentres of the epidemic.Footnote ⁵² Thus, given the incubation period between infection and the manifestation of symptoms, some may have become infected in Calcutta and still passed the pre-embarkation medical examination. Moreover, meals for indentured migrants aboard ships destined for the sugar colonies, including Mauritius, were cooked with mustard oil or ghee.Footnote ⁵³ If the mustard oil was acquired near the ships’ departure point in 1878, then there is a reasonable chance that it was adulterated with Mexican poppy oil, which could have led to an outbreak on board. However, little evidence suggests that epidemic dropsy was prevalent on journeys between Calcutta and Mauritius in 1878–79. One informant to a colonial physician in Mauritius indicated the presence of ‘some people with enflamed legs’ aboard one ship, but this was an exception. Another report directly contradicted the hypothesis of large numbers of infections on board.Footnote ⁵⁴ Uncertainty in this context is enhanced by their rarely being medical officers aboard ships of indentured migrants.Footnote ⁵⁵ Moreover, at the same time as denying the presence of ‘epidemic sickness’ among new arrivals, the Depot Medical Officer also noted the presence of what we now know to be epidemic dropsy on sugar estates.Footnote ⁵⁶ This discrepancy in reporting suggests that the absence of ‘epidemic sickness’ in the Depot Medical Officer’s report cannot be attributed to him simply missing significant numbers of infected individuals as they passed through the port; he was clearly able to recognise an unusual disease. Thus, the number of infected individuals arriving in Mauritius was probably minimal, at least compared to the number of people who were eventually infected.Footnote ⁵⁷

Instead, the spread of epidemic dropsy in Mauritius can be attributed to the nature of the island colony’s food supply during the late nineteenth century. The centrality of sugar production to Mauritius’ colonial enterprise disincentivised domestic food production, making the island reliant on food imports, including cooking oil. Thus, ships of indentured migrants from Indian ports were also laden with food, which was sold to Port Louis’ shops and wholesale to the depot.Footnote ⁵⁸ Indeed, this dependency on imports was beginning to grow at the time of epidemic dropsy’s outbreak owing to what is known as the grand morcellement from about 1875. While it had smaller-scale antecedents in previous decades that had given rise to small cane farming, the grand morcellement signified the dividing and selling of large sugar estate lands to several smaller landholders, especially to former indentured labourers who had fulfilled the terms of their contract and had adopted Mauritius as their home.Footnote ⁵⁹ Such small landholders had to focus on sugar production to make their enterprises profitable, meaning that small tracts on the edges of large sugar estates that may once have been put towards domestic food production were increasingly given over to sugar.Footnote ⁶⁰ This is despite references to the presence of mustard on the island.Footnote ⁶¹ Economic conditions encouraged (former) indentured labourers to purchase the ‘cheapest articles’ from estate shops.Footnote ⁶² Dependency on food imports from India in the context of the latter’s socio-economic challenges in the late 1870s increased the chances of Mauritius’ exposure to adulterated mustard oil.Footnote ⁶³

The hypothesis that most infections of epidemic dropsy occurred in Mauritius rather than pre-embarkation or aboard ship is also supported by a critical examination of the numbers of infections reported by the island colony’s medical establishment. Andrew Davidson, a colonial medical officer, suggested that between 80,000 and 100,000 people, or a quarter of the entire island’s population, were infected at some point,Footnote ⁶⁴ although this is probably an overestimation. McLeod later estimated 20–30,000 infections.Footnote ⁶⁵ Additionally, 729 deaths were registered in Mauritius as ‘acute anaemic dropsy’ – which we now know to be epidemic dropsy – in 1879.Footnote ⁶⁶ If this number is generally acceptable and the mortality rate was around 5% (as has been the case with most outbreaks of epidemic dropsy since the late nineteenth century), then this would put the number of infections at under 15,000. The addition of some numbers from 1878 and accounting for infections not observed by the colonial medical establishment as well as for misdiagnoses of other diseases would still produce a figure closer to the lower end of McLeod’s estimate.Footnote ⁶⁷ Even so, these numbers dwarf the figure of just over 3,000 new arrivals from Calcutta in 1878–79. This strongly indicates that most of those infected in Mauritius had been present in the island colony since before the disease broke out.

Further, colonial medical reports indicate that the disease first broke out in Moka, on the Central Plateau.Footnote ⁶⁸ Since the 1860s, when drought and disease pushed many planters from littoral zones, this was part of the island colony’s core sugar-producing area.Footnote ⁶⁹ One medical officer identified a ‘peculiar’ form of dropsy in November 1878 in this locale, while also suggesting that it may have been present as early as September or October.Footnote ⁷⁰ This broadly aligns with the chronology of the outbreak in Calcutta in that year, with the first official report in the latter also dating from September, possibly suggesting that they were linked to the same batch of adulterated mustard oil.Footnote ⁷¹ The disease then spread, by December 1878, to nearby Plaines Wilhem and Port Louis, the colonial (and current) capital on the western coast. Subsequently, from January 1879, infections enveloped most of the island’s littoral regions, including Savanne, Pamplemousses, Flacq, Riviere du Rempart, and Black River. The epidemic reached its greatest extent in February, when it also appeared in Grand Port, on the southeast of the island, before it receded from April and disappeared entirely by July.Footnote ⁷² Thus, the travel of infection probably went from inland to the coast, rather than the other way round (see Figure 2).

Figure 2.

Reported cases of beriberi, acute dropsy, acute anaemic dropsy, anasarca, and other diseases’ mortality from November 1878 to February 1880. Data taken from: Francis Lovell, ‘Report on Acute Anaemic Dropsy in Mauritius’, Indian Medical Gazette, 16, 12 (1881), 343–4 [Return C and Table D].

Colonial medical sources indicate that indentured labourers who originally boarded ships in Calcutta suffered the highest rate of infection throughout the epidemic. Those who travelled from Madras were apparently also severely affected, though not to the same degree.Footnote ⁷³ By contrast, it only spread among the wider populace in Port Louis, a cosmopolitan centre where a wider demographic would inadvertently have had access to the supply of adulterated mustard oil.Footnote ⁷⁴ Reports from colonial hospitals also indicate that more men were infected than women (see Figure 3). Nevertheless, it is necessary to be careful with such reports. There is a possibility, for example, that colonial medical officers who drew parallels between the symptoms of the novel disease in Mauritius and Calcutta may have overestimated infections amongst indentured labourers from Calcutta to emphasise their scientific case. Certainly, some believed (wrongly) for a time that new arrivals from Calcutta spread it to others already based in Mauritius via human-to-human transmission.Footnote ⁷⁵ Nevertheless, even with these qualifications, this rough outline about the extent to which it affected different groups may be broadly acceptable because of what is known about the indenture system in the late 1870s.

Figure 3.

Return of acute dropsy cases from January to March 1879. Data: Lovell, ‘Report on Acute Anaemic Dropsy’, (1881), 346 [Return].

Most indentured labourers who boarded ships in Calcutta originated in Bihar. They arrived in the port city via the Ganges and Brahmaputra rivers and the East Indian Railway. Since the end of the 1850s, the annual number of new arrivals had decreased precipitously. The peak figure of 44,397 Indian migrants, of which 23,180 were from Calcutta, in 1859, for example, dwarfs those of the late 1870s.Footnote ⁷⁶ By this time, rather than being replaced with new arrivals, indentured labourers increasingly renewed their contracts or purchased small tracts of land on which to grow their own sugar.Footnote ⁷⁷ The relatively few new arrivals, meanwhile, were largely recruited by ‘returnees’ – those who had finished their contracts in Mauritius and were employed to return to Bihar to recruit more to follow in their footsteps. Returnees were encouraged to recruit people from their kin group, as well as entire families. This process was officially sanctioned by the Mauritian colonial government and was more effective in the hinterland of Calcutta than around Madras and Bombay. It was viewed as more cost-effective than using local contractors, and it promised to provide a longer-term solution to the labour question, as whole families might be more likely to stay in Mauritius after their initial contract than individuals, who may have wished to return to their kin network in Bihar.Footnote ⁷⁸

The result was that, by the late 1870s, the labour camps that served the sugar estates and were populated by indentured labourers who boarded ship in Calcutta were fairly insular social places. Richard B. Allen refers to a ‘social structural stability’ within the camps. This social structure was governed by the concepts of family and village, both of which were influenced by prior connections to Bihar.Footnote ⁷⁹ The camps of labourers comprised close-knit communities of around 300 individuals.Footnote ⁸⁰ These conditions provided a fertile environment for epidemic dropsy to spread amongst particular demographic groups, especially those who originated in Bihar because of the relative success of the ‘returnee’ system from that locale. Further, the high proportion of reported male versus female infections suggests that most were infected via consumption of mustard oil in rations provided by sugar estates to plantation workers, who were mostly male.Footnote ⁸¹ However, lower visibility of female-dominated domestic spaces to colonial observers may mean that several infections of women were overlooked.Footnote ⁸² Harsh working conditions on the plantations may additionally have made male sugar workers more vulnerable to the disease’s more severe symptoms, thus increasing their chances of mortality and that they would be noted in colonial medical documents.Footnote ⁸³ In any case, the social structure of labour camps within the context of the wider indenture system may have made labourers who originally boarded ships in Calcutta more vulnerable than other groups to infection with epidemic dropsy in the late 1870s.

The spread of epidemic dropsy to and within Mauritius in 1878–79, therefore, was contingent on a range of local and global factors, both in the short and long terms. Heightened reliance on food imports from India made the island colony more vulnerable to exposure to certain food-borne diseases, such as epidemic dropsy, especially in the context of drought, famine, and inflation in India and Bengal during the late 1870s. This history stands in contrast to other types of trans-oceanic disease transmission to Mauritius at this time, such as that of cholera and smallpox via human-to-human contact, whose longer-term influence had contributed to the implementation of several preventative measures. Further, the social and organisational structures that underpinned sugar production and the indenture system by the late 1870s may have contributed to certain groups in Mauritius – namely indentured labourers who originally boarded ships in Calcutta – being more susceptible to infection with and mortality from epidemic dropsy. Vulnerability to disease and its effects in Mauritius were in some ways built into the imperial structures that linked the island colony to India and Bengal during the nineteenth century.

Colonial medical understandings and responses

The outbreak of epidemic dropsy in the late 1870s elicited a significant reaction from the colonial medical establishments of India, Bengal, and Mauritius. Speaking at the third meeting of the Calcutta Medical Society in March 1880, Dr. E.W. Chambers, a medical officer based in Calcutta, said, ‘Nothing [had] been of more interest to the [medical] profession within the [previous] few months.’Footnote ⁸⁴ Much of this interest was theorised and discussed in the Indian Medical Gazette, a monthly medical journal founded in 1866 as a forum for colonial medical practitioners to exchange and improve knowledge about diseases in India. Within its pages, there are a multitude of reports on what we now know to be epidemic dropsy from Calcutta, Shillong, Dhaka, and Mauritius dated between 1878 and 1882. McCleod then published the culmination of the knowledge gained from the 1877–80 epidemic, and the debates therein, in 1893.Footnote ⁸⁵ Close analysis of the publications in this context sheds significant light on the nature and circulation of trans-oceanic, intra-imperial medical knowledge in the late nineteenth century. It also shows how a much smaller medical establishment in Mauritius came to leave an outsized imprint on colonial medical knowledge production in India. This history builds on recent contributions to medical history in the colonial Indian Ocean World, which emphasise the multiple spatial and contextual origins of ‘modern’ medical science.Footnote ⁸⁶

As is well known, colonial medical knowledge was constrained by several scientific and institutional factors in the late nineteenth century. Miasma theory, which attributed the spread of disease to noxious air, still predominated, and germ theory remained a minority position, especially amongst colonial medical practitioners. Interpretations of miasma theory were compatible with racialised tropes rooted in environmental determinism.Footnote ⁸⁷ The reports on epidemic dropsy contained in the Indian Medical Gazette are filled with postulations about the cause of the disease, and they regularly explore race and climate as possible triggers.Footnote ⁸⁸ Some reports do, however, mention the possible role of specific foods, but even these were couched behind an inherent racism. For example, Andrew Davidson, a colonial physician based in Mauritius, argued that diets based on rice ‘predisposed’ people to some of epidemic dropsy’s symptoms, specifically anaemia.Footnote ⁸⁹ Such discourse was rooted in a colonial belief system that associated rice cultures with negative racial stereotypes.Footnote ⁹⁰ None of these reports examined the possibility that a specific ingredient in the common diets of the island became adulterated or compromised at a particular time. This only came later, notably with an outbreak of epidemic dropsy in Fiji via mustard oil imported from India in the 1920s.Footnote ⁹¹

Understandings of epidemic dropsy in the late 1870s were also rooted in the specific contexts within which the respective medical establishments of India and Mauritius worked during the late 1870s. This meant partially understanding the outbreak in terms of the diseases or challenges that they were already familiar with or were still investigating. In India, one early report suggested that what we now know to be epidemic dropsy was a ‘malarious fever’, while also linking it to malnutrition.Footnote ⁹² But the overriding paradigm through which they sought to understand the outbreak was through identifying it with beriberi.Footnote ⁹³ The first report in the Indian Medical Gazette to give the disease a name, which is dated to January 1879, used the title, ‘Acute Dropsy (Beriberi?)’, despite earlier reports suggesting it was a ‘new disease’.Footnote ⁹⁴ Others titled their papers similarly soon after, and in April–May 1880, one prominent medical practitioner discussed the outbreak in Calcutta simply under the heading ‘Beri-beri’ while drawing direct comparisons with the outbreak of said disease in the Madras Presidency.Footnote ⁹⁵ The influence of this line of thinking was such that by May 1881, it had become the prevailing opinion among the colonial medical establishment in India and Bengal that what we now know was epidemic dropsy was ‘the same disease as has been described by observers in Madras and Ceylon under the term Beri-beri [sic]’.Footnote ⁹⁶

The pervasiveness of the conflation of epidemic dropsy with beriberi was partly rooted in struggles within India’s colonial medical establishment to understand beriberi itself. Indeed, around this time, India’s colonial medical officers were also conflating beriberi with other diseases, such as kala-azar (visceral leishmaniasis).Footnote ⁹⁷ Broadly speaking, until the 1870s, they understood the disease as a distinctly Indian problem, despite outbreaks and accompanying medical literatures in and from other parts of the tropical world.Footnote ⁹⁸ The outbreak of epidemic dropsy in Calcutta in 1877–80 and their linking it to beriberi in the Madras Presidency, meanwhile, allowed them to classify beriberi as an ‘epidemic fever’. Their doing so served two purposes. First, it enabled them to demonise and criminalise internal migrants as ‘fugitives’, especially supposedly infected individuals from Madras who some claimed spread it to Calcutta.Footnote ⁹⁹ Second, according to Arnold, it ‘seemed to give their investigations into beriberi a unique insight and an international mission that, in the competitive world of tropical medicine, might put British India on a par with, even ahead of, investigations elsewhere’.Footnote ¹⁰⁰ The idea that beriberi could be epidemic had the potential to be regarded as an important medical discovery, with consequences for the medical profession across the tropical world. The conflation of epidemic dropsy with beriberi in India owed much to political circumstances within India and to the colony’s medical establishment’s position in the production of global medical science.

The influence of the beriberi paradigm for understanding the epidemic in Calcutta was such that it affected early reporting in India about the outbreak in Mauritius in 1878–79. The first editorial of India-based medical practitioners in this context, dated August 1880, was entitled ‘Beri-beri in the Mauritius’.Footnote ¹⁰¹ The article sought to answer two questions: First, whether the disease was imported into Mauritius from Calcutta, and second, if it was beriberi. In response to the first question, the authors were cautious, noting that the ways they understood the evidence (based on comparable symptoms, demographics of those infected, and the epicentres of the epidemic) were largely circumstantial. Such caution contrasted, of course, with the concurrent rhetoric perpetuated by some medical officers that linked the epidemic in Calcutta to the migration of ‘fugitives’ from the Madras Presidency, which was based on similarly limited evidence.Footnote ¹⁰² In this context, there may have been an initial reluctance to acknowledge that new epidemics originating in India, such as the one under review, could travel internationally, as Michael Christopher Low has observed in relation to cholera around the same time. Acknowledging international communicability had the capacity to provoke new measures, such as enhanced quarantine restrictions, that might limit British India’s mercantile fleet, which was integral to its prosperity. Such prosperity was of much higher priority than public health to British India, including to its medical establishment.Footnote ¹⁰³

In response to the second question, about whether the outbreak in Mauritius was beriberi, the authors were equivocal. Apart from the problem of possibly having to acknowledge the disease, which they had up to that point understood as a distinctly Indian problem, could travel from India to other regions, their response was complicated by the observations of colonial medical officers in Mauritius itself. In this context, they drew on a report by Francis Lovell, colonial Mauritius’ Chief Medical Officer. The report was dated 28 April 1880 and was briefly summarised in the same edition of the Indian Medical Gazette as the editorial.Footnote ¹⁰⁴ It was not published in full, however, until December 1881–January 1882, a delay that may be indicative of the controversy arising from it.Footnote ¹⁰⁵ Rather than in terms of beriberi, Lovell’s report referred to the disease as ‘acute anaemic dropsy’, arguing that there were significant differences between the two diseases. It also used the term ‘epidemic dropsy’ in text, which is probably the first written usage of the term.Footnote ¹⁰⁶ At the same time, while challenging the beriberi thesis, Lovell shed doubt on the veracity of investigations into beriberi in India as a whole by stating that the disease had only been ‘loosely defined’.Footnote ¹⁰⁷ Despite perpetuating the ‘beriberi paradigm’ in headlines, doubts stemming from geo-political factors and interpretations from Mauritius started to creep into some of the colonial Indian medical establishment’s discourse by mid-1880.

It is probably unsurprising, therefore, that medical knowledge about epidemic dropsy developed along vastly different lines in Mauritius than in India. As the April 1880 editorial in the Indian Medical Gazette stated, ‘The Mauritius cases [were] under more close and systematic medical observation than those which were seen in Calcutta.’Footnote ¹⁰⁸ Colonial physicians in Mauritius interviewed patients and tracked hospital admissions, and they attempted to tabulate numbers of deaths from the disease at district and (sometimes) estate levels (see Figure 4).Footnote ¹⁰⁹ Further, Lovell used regular reports on the transmission of – and mortality from – a range of diseases that bore similar symptoms to what we now know was epidemic dropsy, such as anasarca and generic dropsies and oedemas, across several years (see Figure 5). Beriberi, not being present in Mauritius at this time, was not one of them. This enabled him to dismiss early suggestions that epidemic dropsy was related to malaria or was a form of anasarca, and to ignore the Depot Medical Officer, who relied on information from India to refer to the epidemic as beriberi.Footnote ¹¹⁰ Instead, he was able to use colonial physicians’ in-country fieldwork to strongly suggest that what he called ‘acute anaemic dropsy’ arrived in Mauritius in September–October 1878, despite the first formal observation only being written in November of that year.Footnote ¹¹¹ The comparatively small colonial medical establishment in Mauritius mobilised its resources in ways that suggested scientific rigour, enabling a challenge to mainstream colonial medical thinking in India.

Figure 4.

Reported cases of ‘acute anaemic dropsy’ on estates in Savanne from 17 December 1878 to 1 February 1879. The tables detail the daily progression of epidemic dropsy in six sugar estates and the gender and age distribution. The tables do not contain the precise year, but such has been inferred from descriptions and previous tables. Data: Lovell, ‘Report on Acute Anaemic Dropsy,’ (1881), 346–7 [Table E].

Figure 5.

Reported cases of anasarca, oedema, and dropsy cases from June to December 1877–1878. While Dr Clarenc learned of a peculiar disease distinct from ‘dropsy, anasarca and oedema’ with symptoms similar to beriberi in November 1878, Lovell reinforced the theory that the disease was present before, and he compiled both the 1877 and 1878 tables to show such trends. The December peak in 1878 was attributed mainly to the district of Moka, with 45 cases as the disease became ‘general’ there. Table A makes mention of ‘Total Mortality’ in the final row, but this was probably a typographical mistake, as there is no mention of mortality in the associated paragraph of Lovell’s report, and the corresponding Table B does not make mention of mortality. Lovell, ‘Report on Acute Anaemic Dropsy’, (1881), 342–3 [Tables A and B]. Map based on: A. Descubes, ‘Map of the Island of Mauritius’ (1880): https://nla.gov.au/nla.obj-231272435/view [Accessed: 27 July 2023].

Mauritius’ colonial medical establishment’s capacities in this context were rooted in the island colony’s labour and economic structures. As Yoshina Hurgobin has argued, ‘the colonial state viewed the worker’s body as a crucial link to the regime of production processes since workers’ health directly determined his or her production.’Footnote ¹¹² This was no truer than in Mauritius, where the colonial economy was dependent on indentured workers’ production of sugar and therefore also on workers’ health. Maintaining workers’ health in this context, however, was complicated by the colonial apparatus’ unwillingness to invest capital in measures that would enhance sanitary conditions, improve workers’ diets, or provide adequate hospitals.Footnote ¹¹³ Indeed, the existence of estate hospitals in Mauritius since the 1830s has largely been understood in terms of colonial control rather than health – indentured labourers were only admitted once their condition was desperate.Footnote ¹¹⁴ The death rate at said hospitals was consequently shockingly high.Footnote ¹¹⁵ Thus, the colonial medical establishment saw scientific investigations that identified diseases as the more cost-effective strategy for understanding, and therefore limiting, the spread of epidemics. Medical science was understood as fundamental to Mauritius’ sugar economy.

This self-imposed mandate for Mauritius’ colonial medical establishment, which linked workers’ production and the health of the economy to advancements in medical science, was especially strong in the late 1870s. Although epidemics of cholera that spread from ships of indentured labourers had taken a massive toll in the 1810s and 1850s, the arrival of malaria in the mid-1860s provided longer-term challenges. Unlike cholera, which could be periodically eradicated through quarantine measures, malaria continued to spread – colonial medical science had yet to link the disease to Anopheles mosquitos, instead relying largely on miasma theory.Footnote ¹¹⁶ Moreover, it was particularly virulent during seasons or years of above-average rainfall, as damper hydrological conditions are conducive to mosquito reproduction and expansion of their territory, especially on sugar plantations.Footnote ¹¹⁷ The transmission and death rate spiked in 1867 and 1877, both of which were years of above-average summer rainfall triggered by positive ENSO and (in 1877) IOD anomalies.Footnote ¹¹⁸ Thus, epidemic dropsy broke out in 1878–79 during a period of heightened concern among Mauritius’ medical establishment about the effects and spread of disease among the island colony’s indentured labourers.

The cumulative effects of disease on indentured labourers in Mauritius during the late 1870s led the British Government to establish a Sanitary Commission in 1879.Footnote ¹¹⁹ Its mandate was specifically geared towards understanding and arresting the spiralling effects of malaria. However, the 1878–79 outbreak of epidemic dropsy also featured. Charles Meldrum, a Mauritius-based meteorologist and astronomer, took a holistic, global, and long-term approach, associating heightened disease transmission to climatic anomalies, sunspots, and deforestation.Footnote ¹²⁰ He did so by identifying several ‘years of greatest mortality’ at about 11-year intervals in Mauritius across the nineteenth century and by linking them to similar temporal patterns in Ceylon (Sri Lanka), Australia, India, Gibraltar, Malta, Europe, and the Caribbean.Footnote ¹²¹ In this context, he identified 1877–80 as global ‘years of greatest mortality’, with 1879 being the most severe in Mauritius, coinciding with the height of epidemic dropsy’s spread.Footnote ¹²² Indeed, with 729 recorded deaths, he cited ‘béri-béri [sic.] or acute dropsy’ as the driver of heightened mortality in 1879.Footnote ¹²³ In so doing, he acknowledged the limits of contemporary colonial knowledge about what we now know to be epidemic dropsy in both India and Mauritius, while also inserting the disease into a global conversation about epidemics and death in the tropical world.

These developments were not lost on members of the colonial medical establishment in India. Although we now know that Meldrum’s ideas, which linked disease transmission to climate, environment, and their effects on miasmas, were misguided, they – and the broader Mauritian context they were situated in – contributed to a reframing of how some India-based medical officers understood the disease. For example, some theories that positioned epidemic dropsy as a ‘new disease’, which originated in late 1878, were given renewed vigour. One leading figure in this context was McLeod, who in May 1881 was the first to title a published article, ‘epidemic dropsy’, despite, at the time, suggesting questioningly that the disease could be a form of beriberi.Footnote ¹²⁴ In so doing, he used a term first deployed in Lovell’s report to describe the disease in Mauritius.Footnote ¹²⁵ Twelve years later, however, he built further on Lovell’s work by claiming that beriberi was ill-defined and arguing that the term would later ‘disappear from medical nomenclature’ as the ‘true pathology and causation’ of diseases to which this name had been attributed were discovered.Footnote ¹²⁶ Thus, in drawing on the medical literature from Mauritius, he challenged, increasingly directly, the majority position in colonial India’s medical establishment that understood epidemic dropsy in relation to beriberi.Footnote ¹²⁷ His interpretation of Lovell and others’ reports from Mauritius allowed him to distinguish between the Calcutta epidemic he witnessed and previous outbreaks of what was termed beriberi in other locales in India and Sri Lanka, including in the Madras Presidency in the late 1870s.

The core of the Indian colonial medical establishment was not to be cowed, however. Instead, they, like Meldrum in Mauritius, increasingly sought to interpret the 1877–80 epidemic through a global lens. They did so, however, to enhance the perceived importance of beriberi to medical scientific investigations and the role of colonial India’s medical establishment therein. In so doing, medical officers, such as Chevers, drew on medical reports in Mauritius to emphasise beriberi’s potentially epidemic characteristics and the possible international political, economic, and social ramifications that an outbreak could have. This marked a departure from the time of the epidemic itself, when the evidence suggests that acknowledging the possible epidemiological connection between India and Mauritius may have been seen by some as problematic.Footnote ¹²⁸ By contrast, Chevers wrote, ‘an analysis of the characteristics of the epidemic disease which first appeared in Calcutta in 1877, and Mauritius in 1878, enables us to declare… that acute beriberi is an exanthematous fever.’Footnote ¹²⁹ This use of evidence from Mauritius helped some in the colonial medical establishment in India to emphasise the importance of understanding beriberi as an infectious disease. Thus, according to Arnold, while before Chevers’ discussion of the 1877–80 epidemic ‘medical observers in India had… regarded beriberi as a local disease of secondary importance’, they understood it in global terms afterwards.Footnote ¹³⁰ Colonial medical practitioners in India used evidence from Mauritius to both discredit and emphasise the 1877–80 epidemic’s relevance to an understanding of beriberi.