Using historical newspaper accounts to reconstruct the impacts of eighteenth century hurricanes in Central America.

The survey of the historical literature and newspaper articles I obtained from the Burney Collection led me to believe that three storms might have been responsible for producing the AD 1780 overwash layer in Bacalar Lagoon. These occurred in late July and early August of 1784; late August of 1785; and late August and early September of 1787. The 1784 and 1787 storms seem the best fits in terms of storm tracks, intensities, and impacts. Sadly, most of the newspaper articles I obtained reported on their impacts in Belize, largely because British colonists evacuated the Mosquito Shore in early 1787 aspart of the enforcement of the 1783 Treaty of Versailles between Spain and Britain. Belize and the Mosquito Shore are, however, separatedby only about 240 miles. Given that the 1784 and 1787 storms followed east-west paths into the Gulf of Honduras, it seems likely these storms affected both regions.

The Well-Traveled but Poorly Documented Hurricane of 1784

The 1784 tropical cyclone is thought to have originated in the Lesser Antilles around 27 July. It moved westward across Jamaica to the Gulf of Honduras where it struck Belize. From there, it crossed the Yucatan Peninsula, entered the Gulf of Mexico, and turned to the northeast to make a second landfall on 5 August near Pensacola, Florida (Chenoweth 2006). This hurricane appears to have been noteworthy, but only sparse descriptions survive in the Burney Collection. The following brief quote describes its circuitous route and mentions that it affected the Gulf (Bay) of Honduras, especially the Spanish fort of Omoa on the Caribbean coast of Honduras:

"The late hurricane at Jamaica was not confined to the windward islands alone ... it was severe in both the Floridas, and in the bay of Honduras it was beyond example violent, and has done great damage to the Spaniards [sic] settlements.., at Omoa ... [where] it is mentioned, several buildings were blown down" (London Chronicle 1784: 371).

A second passage provides a general description of the 1784 hurricane season in the Gulf of Honduras, which certainly applies to Belizeand the Mosquito Shore given their proximity to each other. Althoughthis brief description does not specifically mention a storm, it leaves little doubt that the latter half of the 1784 hurricane season was particularly severe on the Caribbean coast of northern Central America:

"By letters from the Bay of Honduras we learn, that the settlers ... have experienced, for these five months past, the severest weatherever remembered in the memory of man; it having rained incessantly and overflowed considerable tracts of land ..." (Morning Chronicle andLondon Advertiser 1785).

Narratives of Loss and Recovery: Accounts of the 2 September 1787 Hurricane

Newspaper coverage of the 1784 hurricane that hit the Gulf of Honduras suggests it was an intense storm, but I only found some 857 printed words about it in the Burney Collection. By contrast, the 1787 hurricane generated at least 2,437 words, most of them in a single article from the General Evening Post, from which I draw extensively in this section. Word count is no clear reflection of storm intensity, but at the very least it is suggestive. The news accounts of the 1787 hurricane clearly indicate that its impact in the Gulf of Honduras was catastrophic. As with the 1784 storm, the 1787 hurricane formed in the Lesser Antilles, striking Dominica and Martinique before moving west across the Caribbean. It must have passed very near the Caribbean coast of Honduras because a ship named the Sally was reported to have foundered (i.e. sunk) in it soon after departing from Black River colony in the northwest corner of the Mosquito Shore. The following passage is an excerpt from a newspaper account of the event:

"The Sally, Capt. Smith, a new brig belonging to a Merchant at Jamaica, loaded with lumber from Black River for Jamaica, foundered in the [late] hurricane [of the Bay of Honduras[; she had just sailed [from Black River], and had not gained a league's difference" (Public Advertiser 1787).

After passing the Mosquito Shore, the hurricane continued along a westward or possibly west-northwestward track until making landfall in Belize and finally dissipating over southern Mexico. There is no evidence that the storm passed overland to regenerate in the Gulf of Mexico as had occurred with the 1784 storm. The following passage recounts the arrival of the hurricane in the vicinity of present-day Belize City. Based on this description of wind direction, the center of the storm appears to have been slightly north of the settlement when it made landfall:

"On Sunday, the 2d of September, 1787, between ... four and five in the morning, a gale of wind commenced from the N. N. W. At six o'clock, it came to blow at W. N. W. with great violence, attended with rain; the sky became dark and obscured, and carried with it a most threatening appearance. At eight o'clock it blew a most violent hurricane, attended with incessant rain ... about ten o'clock, the wind shifted to S. W. and blew if possible, with redoubled violence. At that instant the sea began to rise ... with such rapidity, that the banks ofthe river were in a few minutes overflowed, and shortly after the whole surface of the earth covered with water" (General Evening Post 1788).

This passage is important not only because it highlights the trackand timing of the 1787 storm at its moment of impact in Belize, but also because it includes a description of storm surge. Most of the storm accounts I found in the Burney Collection focused on wind, rain, loss of life, and property damage. Less common were descriptions of flooding. Observations of storm surge were even more rare. Storm surgeis a common feature of tropical storms and hurricanes, but its height varies according to the intensity of a storm and the timing of its landfall in relation to the tides. Furthermore, the destructive potential of storm surge depends on the physical terrain and distribution of human settlements in the area where a tropical cyclone makes landfall. The fact that the 1787 hurricane was notable for its surge is a good reason to believe this storm might have been responsible for the AD 1780 overwash layer in Bacalar Lagoon.

Newspaper accounts describe the 1784 storm as "beyond example violent," but provide few clues for the reader to determine what this actually means. In fact, most newspaper articles I found in the Burney Collection that document storms make common use of hyperbole, leaving the reader with the impression that each event was the worst disaster in the history of the place where it struck. Natural disasters make gripping news. This is as true today as it was in the eighteenth century. Without details to back them up, however, sensational claims are of limited use in scholarly reconstructions of historical hurricanes. The newspaper articles I found about the 1787 hurricane are notably different in this regard because they combine vivid imagery with relatively precise figures, making it possible to acquire a relatively clear picture of how intense the storm was, as can be seen in the following passage, which continues almost from where the previous one left off:

"After remaining in this uncertain and disagreeable situation until near one o'clock, the hurricane abated, and in a short time.., the weather cleared up; when it presented to the surviving inhabitants a most melancholy picture of misery and distress ... not less than fivehundred houses of different constructions having been blown down ...the dead bodies of those who had perished ... the carcases [sic] of ... live stock, lying in different parts of the bush .... Out of fifteen square-rigged vessels, sloops and schooners ... and other vesselsemployed in bringing wood from the different rivers... not a single one was to be seen; the whole having either sunk or been driven on shore, and many of their hands perished.... In this disconsolate situation the distressed inhabitants without any dry cloathing [sic], or necessary refreshment ... their bodies ... bruised by the blows they had received from the limbs of trees, logs of mahogany, and other pieces of wrecks floating about in the bush, betook themselves to the erecting a few temporary sheds, and by digging amongst the rubbish ... and what served to complete the measure of their misfortune, the plantations were all ... totally destroyed.... The number of mahogany treesblown down in the different rivers, and split to pieces, is incredible. The paths are so filled up with limbs and branches of trees, and other rubbish, that it will require a length of time to clear them. In short, the whole country, for upwards of twenty leagues along the sea-coast, and the same distance up the rivers, exhibits a scene of desolation and ruin" (General Evening Post 1788).

It is clear from the above passage the 1787 hurricane was catastrophic to the inhabitants of Belize. Its area of impact spanned twenty leagues along the coast and stretched another twenty leagues inland, encompassing the equivalent of 3,600 square miles (9,323 square kilometers). Such an extent conforms well to the impact zone of a relatively large and intense storm. The 1787 storm destroyed eleven seafaring ships and countless small vessels, and damaged numerous other watercraft. It destroyed about 500 houses and buildings, leaving Belize with few standing structures in its aftermath. The General Evening Post article and others reported damage from the hurricane totaled 30,000 [pounds sterling], which by contemporary exchange rates, is equal to about $60.8 million U.S. dollars. Financial damage from hurricanes today dwarf such an amount, but the 1787 storm must have been a painful loss for the colonists of Belize given that few would have received any compensation from insurance.

Given the hurricane's catastrophic impacts in Belize, it is unfortunate that I was able to find no documentation of its effects in eastern Honduras, aside from the report describing the loss of the Sally that I quoted earlier. This lack of evidence, however, is most likelythe result of diplomatic maneuverings between Britain and Spain. In the 1783 Treaty of Versailles, Britain agreed to abandon its claims to the Mosquito Shore in exchange for Spain recognizing Belize as a British colony (Naylor 1989). Implementation of this agreement was delayed until 1787 when British inhabitants of the Mosquito Shore were finally evacuated toBelize, Jamaica, and other locations. I found numerous references to this evacuation in newspaper accounts leading up to the hurricane of 2 September 1787. The following quotes, printed in April and October of that year, indicate the final evacuation of the Mosquito Shore was completed less than a month before the hurricane arrived:

"Kingston, Jan. 13. Tuesday morning sailed from Port Royal for Musquito [sic] Shore, his Majesty's ship the Camilla, the Swan sloop of war, and the London Transport, having detachments from the 10th and 19th regiments on board, under the command of Major Nesbitt, for the purpose of facilitating the evacuation of that country" (The World and Fashionable Advertiser 1787).

"Kingston, Sept. 8. His Majesty's ships Camilla and Cygnet from the Musquito [sic] Shore, but last from Truxillo, arrived at Port Royalon Thursday last [30 August]. Prior to their departure, the evacuation was entirely completed, and the Shore left in possession of the Spaniards" (General Advertiser 1787).

Many of the former inhabitants of the Mosquito Shore ended up in Belize and it is clear from the newspaper reports that some of them arrived only a short time before the hurricane. The following excerpt from the General Evening Post article describes how one former resident of the Mosquito Shore, as well as his family and slaves, died in Belize during the storm:

"Besides those abovementioned, Mr. John Pitt, a most respectable member of the community, lately arrived from the Mosquito Shore with his family, was overset in his schooner, with which he had come from Rowley's Bight the evening before, and perished. With them, most of the negroes on board, who composed part of the vessel's crews, were entirely lost; there was a number of other persons, whose names are not mentioned, but supposed in the whole to be about one hundred and upwards" (General Evening Post 1788).

With just over 100 human fatalities, the hurricane that struck Belize on 2 September 1787 pales in comparison to many other storms, butits impacts were severe enough for it to be ranked the 149th deadliest Atlantic Basin tropical cyclone by Rappaport and Fernandez-Partagas (1995). It is clear from newspaper reports of the time that the months that followed the hurricane were difficult ones for the survivors. As the following passage illustrates, government assistance and charity in the aftermath of the storm played as critical a role in helping the colonists rebuild their lives and livelihoods as humanitarian aid does today:

"Since this melancholy event the superintendent has paid four months provisions to a part of the Mosquito Shore inhabitants ... The other inhabitants have been likewise relieved from the apprehensions of famine, by the arrival of three or four vessels from Europe and Jamaica, which have brought them a seasonable supply of provisions and cloathing [sic], with some other necessaries for family use, which they were much in want of. They are now employed in erecting themselves temporary huts, until they are enabled, by a supply of lumber, to build themselves more comfortable habitations" (General Evening Post 1788).

Thanks to public assistance and to their own initiative, the survivors of the hurricane of 2 September 1787 apparently were able to recover in the months that followed the storm. The final passage comes from the last newspaper report I found in the Burney Collection about the storm. Published in March 1788, but based on correspondence dated from January of that year, we learn that life was well on its way back to normal for the colonists of Belize, and presumably other settlements in the Gulf of Honduras that had been struck by the hurricane. Although the writer of the letter is identified as a "British Settlerat Honduras," it is important to recognize that prior to the twentieth century, Belize was commonly referred to as Honduras or as British Honduras in many historical documents. The reference to St. George's Quay (or Cay) clearly locates the correspondent near the present-daysite of Belize City:

"A letter received from a British Settler at Honduras, dated St. George's Quay, Jan. 30, has the following paragraph "Every thing is going on prosperously, the Settlers are rapidly recovering from their losses by the late dreadful hurricane and inundation ..." (Public Advertiser 1788). Conclusions

Tropical cyclones are dangerous natural hazards that seasonally affect warm, maritime regions of the world. Most are relatively weak and bring little more than heavy winds and rain. Catastrophic hurricanes, on the other hand, pose grave threats to human life and property, and their impacts can be felt years later. Population growth and urbanization in the southern United States, Gulf of Mexico, and CaribbeanSea have brought ever larger numbers of people into areas exposed totropical cyclones. Although eastern Honduras and Nicaragua remain among the most sparsely settled areas of Caribbean Central America, demographic growth has occurred there as well and local residents have become increasingly connected to the larger world (Cochran 2008). By virtue of their greater numbers and relative poverty compared to otherareas of the isthmus, they are also becoming increasingly vulnerable to tropical cyclones and other hazards.

My research helps us to better understand the contemporary challenge of tropical cyclone mitigation by looking to the past for insight into the long-term frequencies of catastrophic storms and their historical impacts in Caribbean Central America. My first major archival research effort yielded some important clues about eighteenth century storms in the region. I found accounts of several events, for example, that might have produced the AD 1780s overwash layer in Bacalar Lagoon. The September 1787 storm is particularly promising in this regard, but the descriptions I found focus almost entirely on Belize rather than eastern Honduras. Over the next several years, I hope to conduct additional archival research in Britain, Spain, and Central America to locate more accounts of this storm and learn more about other storms that struck the region in the past.

The newspaper accounts I found in the British Library also provide insight into tropical cyclones as contemporary hazards. Coverage of the September 1787 storm is particularly helpful in understanding the human dimensions of these extreme weather events. The eighteenth century residents of Honduras and Belize were fewer in number and more isolated than their counterparts today and they had no modern means of forecasting tropical cyclones. Despite these differences, I was struck by how similar historical accounts were to contemporary news coverage of storms. Descriptions of the 1787 event vividly portrayed the winds, rain, storm surge, and destruction that came with the storm. Later accounts emphasized the struggles of local residents to rebuild their lives and the efforts of local administrators to help them. The last reference to the storm, written almost six months after it made landfall, provides us with a few brief lines telling us that life was returning to normal. No two tropical cyclones are identical, but thethemes emphasized in these accounts of apocalyptic destruction and death followed by a gradual return to everyday life seem to be near-universal features of storm narratives, whether they recount an event of the past or one of today. Such discoveries make historical study of tropical cyclones a particularly rewarding field of inquiry for me. Embarking on archival research is quite different from my previous work in the steamy tropical lowlands of eastern Honduras. It is an adventure, nonetheless, and I see it as an exciting opportunity to contribute new knowledge and insights into hurricane mitigation in Central America.

Hurricanes in the area, mid 1700's thru mid 1800's

16-21 September 1749 Dominica to Roatan and Bay of Honduras

* 27 July to 5 August 1784 Lesser Antilles to Gulf of Honduras and Gulf of Mexico (8)

*23-31 August 1785 Northern Lesser Antilles to Jamaica and Belize (4)

*29 August--2 September 1787 Dominica to Gulf of Honduras and Belize (11)

*19-23 September 1787 Cuba and Jamaica to Gulf of Honduras and Belize (1)

6-10 October 1802 Western Caribbean to Gulf of Honduras

5-1 June 1.812 Leeward Islands to Belize

20-26 October 1817 Barbados to Nicaragua

10-17 August 1831 South of Barbados to Yucatan Peninsula

Note: Asterisks refer to storms for which I obtained news coverage in the Burney Collection. The numbers in parentheses correspond to the number of articles obtained for each storm.

Focus on Geography