What was the cause of the epidemic in Savannah in 1733?

On 17 November 1732 Englishman James Edward Oglethorpe (1696–1785) set sail on the Anne ¹ from Gravesend on the River Thames with 114 colonists (Ref. 1, pp. 295–98) with a Charter from King George II to establish an English colony in Georgia. Oglethorpe intended the colony as a second chance for those who had been in debtors prison, while King George also sought this colony as a buffer for the English colonies in the Carolinas against the Spanish colony in Florida. The crossing took eight weeks and the colonists landed in Charles Town, South Carolina on 13 January 1733. The colony was established at Savannah, Georgia on 12 February 1733 (Figure 1). (Charles Town, now Charleston, remains a sister city to Savannah.)

Figure 1.

Arrival of English and Jewish settlers in Savannah, and the dates of 1732–1733 influenza pandemic in Europe.

On 6 April 1733 William Cox, the colonists’ physician, died. In early July, a febrile illness ² struck the new colony taking the lives of three people in the early part of July. On 11 July 1733 a ship that had departed from London carrying 41 Jews (mostly Portuguese Jews with two German Jewish families) who had been living in London unexpectedly arrived in Savannah.

Among the arrivals was a Portuguese Jewish physician Samuel Nunes. Eleven more of the English colonists died in July. But Oglethorpe credited Dr Nunes with alleviating the epidemic and saving the colony ³ : ‘He proceeded by cold Baths, cooling Drinks and other cooling Applications. Since which the Sick have wonderfully recovered.’ While initially being reluctant to allow the Jewish settlers to join the colony, the actions of Nunes caused Oglethorpe to allow them to join the new colony. This initiated an era of inclusion for the colony – e.g. a group of Lutherans were allowed to join the colony in 1734. Oglethorpe was also against slavery, though, unfortunately, apparently on the grounds that having slaves made others not have to work themselves, rather than on moral grounds. Tolerance and inclusiveness did not extend to Catholics, at least partly out of fear for perfidy with Spain.

What caused the epidemic in Savannah in 1733? We have not seen this question studied previously. Here using records of Oglethorpe's colonists and the ship carrying the Jewish settlers, we will study this question.

Ages and dates of death are known for most of the English colonists, as well as family relationships among them and occupations. ¹

We found and collated the following sources as data for the Jewish settlers: (1) a diary of deaths, births, marriages and arrivals and departures from the colony compiled by settler Benjamin Sheftall and completed by his son Levi Sheftall; ⁴ (2) English records that corroborate Sheftall's diary entries in terms of individuals who were alive after July 1733 – e.g. reports from 1735 that some Jewish settlers attended Church of England services in Savannah^5,6; and (3) records^5,7 of Jewish settlers granted land in December 1733 or later.

We analysed the data for factors such as age, sex and occupation of individuals who died in July of 1733 compared with individuals who survived the epidemic. We also looked at the time course of deaths prior, during and after July 1733.

In Supplementary Table 1, data for the original colonists who set sail on the Anne are shown. Data for the Jewish settlers are shown in Supplementary Table 2. Statistics for death rates for various groups of the English colonists are given in Table 1. First, the kill rate of the epidemic was extraordinarily high (13%).

Table 1.

Calculated death rate for men, women, children and adults.

Date rate	Percentages
Epidemic	12.8
Men	12.1
Women	14.0
Children (age 0–16 years)	5.4
Adults (age 17 years and above)	16.7

Figure 2 shows the number of deaths of these colonists by month for the years 1732–1735. July 1733 was indeed the most deadly month, by far, for the early colonists. Fourteen colonists out of 109 alive as of 30 June 1733 died in July for an extremely high mortality rate of 12.8%. The distribution of deaths during July 1733 is shown in Figure 3. There were nearly four times as many deaths in the last two-thirds of the month, than in first third. The epidemic then suddenly ended.

Figure 2.

Number of deaths of English colonist by month.

Figure 3.

Number of deaths in June, July and August 1733.

There was no significant difference in death rates between men and women (Table 1). However, there was a very high preferential mortality rate among young and middle-aged adults compared with children (Table 1, Figure 4). Indeed, only 2/37 children alive as of 1 July 1733 died during the month (5.4%), while six times as many adults died (12 of 72 adults [16.7%]). In contrast, from the sailing of the Anne through June 1733 three of the five individuals who died were children.

Figure 4.

Age distribution of English colonists dying and surviving through July, 1733. Chart shows the mean ages (range) of those children and adults who died and survived through July 1733.

Of the 41 Jewish settlers who survived the crossing to land in Savannah in 11 July 1733 we have evidence of 29 being alive months or longer after July 1733. And unlike the English colonists, there is no evidence for mortality among the Jewish settlers in July 1733. Now, of course, absence of evidence is not evidence of absence, though from existing data it would seem that mortality among the Jewish settlers in July 1733 was zero or very low. Indeed, Sheftall ⁴ does further record the death of a baby during the Jewish settlers’ crossing of the Atlantic. There thus appears to have been a very significant difference in mortality between the young adult English colonists and Jewish settlers.

What febrile illness could have caused this high death rate among the English colonists in July 1733? It would be helpful to have a good case description of the disease; however, we have only a brief one from Oglethorpe. ² ‘But the illness being once frequent became contagious. It appeared chiefly in burning feavers or else bloody fluexes attended by convulsions and other terrible symptoms.’ The natural history of the disease that caused the epidemic is not described; however, we do know that it was a disease that had an extraordinarily high kill rate in a population that prior and subsequent to July 1733 did not suffer any other such events.

One possibility to consider as the cause is yellow fever. Yellow fever is caused by a flavivirus that is transmitted by the bite of female mosquito, Aedes aegypti. ⁸ Yellow fever epidemics were not uncommon in the summer months in port cities in North America at that time. Indeed, there was a major yellow fever epidemic in Charles Town, New York and Philadelphia in 1732, one in Norfolk, Virginia in 1737 and another one in Charles Town in 1739. ⁹ But no yellow fever epidemic was known for North America in 1733. ⁹ There was significant traffic of individuals between Charles Town and Savannah in the spring and summer of 1733. Typically about 85% of individuals infected with yellow fever have a mild disease course. Fifteen percent can progress to a much more severe phase of the disease with mortality for this stage being as high as 30–50%. So the mortality rate of 13% (14/109) for the English colonists is not unprecedented for yellow fever, but at the far upper range. Indeed, no yellow fever epidemic in the United States is known to have had such a high mortality rate. ⁹

Yellow fever can cause gastrointestinal bleeding, but due to coagulopathy from liver damage there is also often bleeding in the mouth and eyes, and these signs are not mentioned by Oglethorpe. Further, in the severe or fatal form yellow fever typically causes significant jaundice, hence its moniker, and this too was not noted by Oglethorpe. The discrepancy in mortality rates between adults and children is not inconsistent with yellow fever: whether due to some immune state in childhood or lack of exposure to the mosquito vector from not working or living near seaport docks, mortality in yellow fever epidemics is lower in children than adults.

The time course of the Savannah epidemic is not particularly consistent with yellow fever. The incubation period for yellow fever is 3–6 days. This is followed usually by a 3–4 day mild self-limiting illness and then return to health with lifelong immunity against the virus. In a minority of cases the initial illness is followed by a day or two remission period and then a toxic phase which can lead to death in 10–14 days. So the time from infection to death is about 21 days. The epidemic in Savannah only lasted one month and so infection would have taken place from early June 1733 to early July. Yellow fever epidemics did not usually ⁹ cease until the fall as the drop in temperature removed the mosquito vector. There is no evidence of a sudden significant drop in temperature in early July 1733. Another possibility is that Dr Nunes instituted measures to prevent yellow fever infection for both English and Jewish colonists. First, Oglethorpe does not describe such measures. Further, this would be unprecedented prevention from yellow fever infection. Indeed, during a very severe epidemic of yellow fever in Philadelphia in 1793 no less than President George Washington left the city, reportedly to protect his wife, ¹⁰ as there was thought to be no way to prevent or treat yellow fever other than evacuation from an epidemic area. Finally, Oglethorpe appears to be describing not just a febrile epidemic but one that was contagious from person to person which yellow fever is not.

Another possibility is a gastrointestinal illness such as Shigella. Such an infection would not be inconsistent with the fever and reported bloody stools. However, one would expect that children would suffer mortality at a similar rate to adults. ¹¹ In addition, the incubation period for Shigella is usually 1–3 days (may range from 12–96 h and up to 1 week for S. dysenteriae); since the Jewish settlers arrived only one-third of the way through July 1733, if this were the cause of the epidemic, we would have expected mortality from them as well.

An intriguing possibility as a cause of the epidemic is pandemic influenza. 1732–1733 was a year of pandemic influenza. As shown in Figure 1, interestingly, the pandemic flu did not come to England until late December 1732 – after the English colonists had left for North America on the Anne. So the English colonists would not have had immunity to the pandemic virus. (The influenza pandemic of 1732–1733 reached America, though the dates it did are not known as well as they are for Europe. ¹² )

Conversely, the Jewish settlers most likely were in England during the time when the pandemic passed through the country and they would then have had immunity to pandemic flu in July 1733. The date when the Jewish settlers left England is not known, but much can be surmised from what is ⁵ : In the 1720s and 1730s the London Jewish community worked to improve the lot of its less fortunate members. A propos the current topic the Trustees of the nascent Georgia colony were approached about allowing Jewish settlers – at no expense to the Trustees. But records ¹³ of the debates of the Trustees from 17 January 1733 and 31 January 1733 show that the trustees were against this. On 17 January 1733, ‘… We agreed that no Jews should be sent…’ It would seem that the Jews in London took this ‘no’ as a ‘must go’. And the thinking ⁵ is that the Jews would be settlers trying to sail from England to Savannah in middle or late January 1733. However, the ship immediately sustained damage necessitating repair before finally leaving English soil ¹⁴ : ‘When the ship first started, she sustained some serious injury in the river Thames, and was compelled to land her passengers and undergo repairs. After this was accomplished a re-embarkation of the passengers took place, and the ship set sail for the “NEW WORLD.” The passage was a disagreeable and boisterous one: gale succeeded gale, and the ship came near being wrecked off the coast of North Carolina, and was forced to seek safety in “New Inlet,” where she was necessarily detained for some weeks. She again set sail, and arrived and landed her passengers in Savannah on the 11th day of July 1733.’ Even assuming one month repair time in North Carolina and a long two months Atlantic crossing time, the Jewish settlers were in England through January 1733, and most surely February as well.

Thus, the English colonists, but not the Jewish ones, would have been susceptible to pandemic flu in July of 1733. This would explain the gross difference in mortality rates in that month between those two groups. Also, pandemic influenza, while causing significant mortality in individuals of all ages, tends to have increased mortality in young adults rather than non-infant paediatric patients. (Seasonal flu, by contrast, only typically causes mortality among infants and the very old.) Indeed, data collected during the pandemic of 1732–1733 showed a much increased mortality in young adults compared with children. ¹² The explanation for this is not fully understood, but a recent paper suggests that part of the explanation may be due to anti-flu antibodies that have been built up in young adults that are ineffective in neutralizing a novel pandemic strain, but bind complement and lead to lung and other organ destruction. ¹⁵ However, in the case of the 1733 influenza pandemic there was a prior pandemic in 1729, but not one before that until back as far as 1580, ¹² so it is not clear for how long influenza was circulating prior to 1733. Fever is typical of influenza. Diarrhea was reported as a symptom of the 1732–1733 pandemic, ¹⁶ though not bloody stools. However, in a vulnerable population such as colonists in a new and foreign land, secondary bacterial infections or nutritional concerns could have accounted for bloody stools if that was a significant symptom. ‘Anguished fits’ were also reported as a symptom of the 1732–1733 pandemic. ¹⁶

We hope that further study of the records of the English and Jewish settlers to the Savannah colony and other colonists and visitors to Savannah at that time may be helpful in further elucidating the cause of the epidemic in Savannah of 1733. We also hope that better understanding of past epidemics may be helpful in planning for or treatment of future epidemics.