THE YELLOW FEVER EPIDEMIC OF 1855 IN NORFOLK AND PORTSMOUTH:
An Analysis of Social Response and Climate.
By Lara Marie Hamilton
High Honors Graduate of the College of William and Mary
This thesis may not be reproduced in any format without the written permission
of both the author and the University Archives, The College of William & Mary.
Notes: Enlarge underlined items in browser for better viewing.
Page numbers appear bold and in brackets.
 Table of Contents
List of Tables Page 4 List of Figures Page 5 Chapter 1: Introduction Pages 6-8 Chapter 2: Research Design and Methods Pages 9-22 Chapter 3: Presentation of Findings Pages 23-57 Chapter 4: Relevance and Conclusions Pages 58-67 Bibliography Pages 68-71 Appendix Pages 72-73
 List of Figures
1. Map of Norfolk, Portsmouth, and Gosport, Virginia (Pollock 1886)
2. Map of Norfolk, Virginia in 1802 (Wertenbaker 1931)
3. Yellow Fever Transmission Cycle (Mattingly 1969:20)
4. Box Plot Depicting Monthly Average Temperature for July Through September in Fort Monroe and Crichton's Store
5. Box Plot Depicting Monthly Average Temperature for Fort Monroe Only
6. Bar Graph Depicting Mean Monthly Average Temperature for Fort Monroe (Without Sub-Groups)
7. Error Bar Depicting Mean Monthly Average Temperature for Fort Monroe (With Sub-Groups)
8. Bar Graph Depicting Monthly Diurnal Range for Fort Monroe
9. Scatter Plot Depicting Monthly Diurnal Range for Fort Monroe
10. Box Plot Depicting Monthly Precipitation for Fort Monroe and Crichton's Store
11. Bar Graph Depicting Mean Monthly Precipitation for Fort Monroe and Crichton's Store
12. Bar Graph Depicting Mean Monthly Precipitation of the Entire Year for Crichton's Store (With Sub-Groups)
13. Box Plot Depicting the Number of Precipitation Days for Fort Monroe and Crichton's Store
14. Box Plot Depicting the Number of Precipitation Days for Crichton's Store Only
15. Box Plot Depicting the Number of Frost Days for Crichton's Store Only
16. Bar Graph Depicting Mean Number of Frost Days for Crichton's Store Only
 List of Tables
1. Stage Characteristics of Social Response to Epidemics
2. Monthly Average Temperature Data from Fort Monroe and Crichton's Store
3. Monthly Diurnal Range Data from Fort Monroe and Crichton's Store
4. Monthly Precipitation (cm) Data from Fort Monroe and Crichton's Store
5. Number of Precipitation Days as Recorded in Fort Monroe and Crichton's Store
6. Number of Frost Days as Recorded in Fort Monroe and Crichton's Store
 Chapter One: Introduction
Imagine living in a well-known port city built on a reputation for being safe, clean, and healthy, and then in a matter of a few months experiencing an indescribable period of terror and death—people fleeing the city in fear for their lives, businesses and churches shutting down, poor immigrants being blamed for the cause and spread of disease, and a death toll rising every day. This was the scene for citizens of Norfolk and Portsmouth in the summer of 1855 when they faced a severe epidemic of yellow fever. Although it was not uncommon in the area, the citizens were uncertain of the causes of the disease. Many associated yellow fever outbreaks with poor sanitation, poverty (socioeconomic status), and foreigners (ethnicity) because the poor Irish immigrants in crowded, dirty tenements were usually affected by disease. They also knew that there was a connection between climate and fever but did not understand it well. People were most familiar with only the symptoms of the disease and the likelihood that death would follow.
This senior honors thesis examines the yellow fever epidemic of 1855 that devastated Norfolk and Portsmouth as a case study of the social response to yellow fever in a medical anthropology context, and climate conditions of 1855 in relation to the disease. Knowledge of the disease, prejudiced views held by the citizens, and preventative measures taken by the community are evaluated to explain the impact of the epidemic on the community's behavior. This paper presents a model explaining four stages of social response to epidemics. Primary sources were utilized, such as The Report of the Howard Association of Norfolk (Howard Association 1857), which describes the investigative process that the committee of physicians followed in attempts to identify the  cause of this devastating epidemic, and George Armstrong's (1856) History of the Ravages of the Yellow Fever in Norfolk, which is written as a series of letters describing the epidemic firsthand, as windows into the perspectives held by the citizens at the time of the epidemic. This case study also examines the correlation between environment, especially climate, and yellow fever. Weather conditions of 1855 are evaluated using inferential statistics to gauge whether climate was more conducive to the mosquito Aedes aegypti, the yellow fever vector, compared to other years. A warmer, wetter climate contributes to increased mosquito population and activity, which enhances the chances of an epidemic occurring with the presence of yellow fever in the area.
Medical anthropology is the study of how people in different cultural settings experience health and illness. Medical anthropologists apply anthropological theories and methods to the study of health, illness, medicine, and healing. The social response of the citizens of Norfolk and Portsmouth can be compared to other cases of outbreaks of infectious disease, using the four-stage model. With the prospect of global warming, the number of cases of infectious diseases like yellow fever and other mosquito-borne diseases are forecast to increase, especially threatening developing countries that are do not always have the latest medical technology or receive assistance from other governments. This paper can contribute to future preventive methods and education to prevent devastating effects on these peoples.
This thesis is divided into four chapters. After this introduction, Chapter Two describes the research design and methods used in obtaining and analyzing data. Chapter Three is the presentation of findings concerning the social response of the citizens of  Norfolk and Portsmouth to the yellow fever, and the statistical analysis of the climate of 1855. Chapter Four explains the relevance and conclusions of this paper.
 Chapter Two: Research Design and Methods
Norfolk and Portsmouth, the "Twin Cities by the Sea," were two important Virginia ports in 1855. Norfolk is located on the east bank of the Elizabeth River, opposite Portsmouth and Gosport, near Hampton Roads, Cape Henry, and the Atlantic Ocean (Figure 1). The cities are nearly surrounded by the waters of the river and marshy creeks. The climate of
Figure 1: Norfolk, Portsmouth, and Gosport (from Pollock 1886)
Norfolk and Portsmouth is relatively mild and agreeable throughout the year, although variable (Howard Association 1857:96). The nearness of the ocean affects the temperature, allowing mild winters. The summers are long and temperate with an average temperature of 78.7° F. Precipitation is well distributed throughout the year (Henry et al. 1959:1). The population of Norfolk in 1855 was 16,000 and that of Portsmouth about 10,000, with approximately a quarter of both cities being Black inhabitants (Howard Association 1857:97).
In 1855, Norfolk consisted of 185 clean, paved, and well-drained streets and lanes. As a result of its low elevation and minimal relief, Norfolk occasionally flooded in the 19th century. "Barry's Row," a poor Irish neighborhood, was built on man-made  ground to prevent it from being reclaimed by the river. The commercial part of the city, which consisted of forty warehouses and ten wholesalers and grocers, lay between Main Street and the river (Howard Association 1857:96). Norfolk merchants traded with New York, Philadelphia, Baltimore, Richmond, and Washington, D.C. The affluent citizens lived on Main Street east of Church Street or in the Northwest section of the city, whereas the impoverished crowded in tenements on West Main along Church Street, Bermuda, Hartshorn's Court, Holly Place, and in alleys off Water Street. The Black residents lived in the north half of town near Charlotte, Amelia, and Church streets (Parramore et al. 1994:189). These areas of Norfolk are depicted in Figure 2.
Figure 2: Map of Norfolk, Virginia in 1802 (from Wertenbaker 1931)
Portsmouth is southwest of Norfolk on the opposite side of the Elizabeth River. Gosport is south of Portsmouth, separated from it by a marshy creek (Figure 1). It had a population of five or six hundred, mostly Irish laborers and their families. Its main street, Water Street, runs from the creek to the navy yard. In 1855, both Portsmouth's and  Gosport's streets were mostly unpaved and poorly drained. However, on the Water Street front, where many workmen lived, the surface was clean and well drained. At the time, Portsmouth had little commerce, so few warehouses or ships were at her wharves (Howard Association 1857:97).
On the sixth of June in 1855, the steamer Benjamin Franklin from St. Thomas in the Virgin Islands pulled into port at Norfolk for repairs. It was known that yellow fever was prevalent in St. Thomas when she left, so she was quarantined. The captain and crew, however, were allowed to visit the city. When it was reported by the Norfolk health officer, Dr. Gordon, on the 18th of June that there were no cases of yellow fever on board, the ship was given permission to go up to Page and Allen's Yard at Gosport near Portsmouth. On the 5th of July, a man from Richmond who had boarded the ship to provide assistance was stricken with the disease and died four days later (Howard Association 1857:102-103). This caused panic in Portsmouth and the Town Council ordered the ship to be sent out to quarantine the next day, but the disease had already begun to spread (Pollock 1886:157).
By the middle of July, yellow fever cases were reported in Norfolk. Barry's Row, the Irish section of Norfolk nearest the port, reported the first cases (Howard Association 1857:107). The Board of Health barricaded the streets of Barry's Row and removed all of the inhabitants. Many citizens fled the city in fear. However, other cities including New York, Richmond, and Petersburg, which at first had given asylum, refused to receive any others. Places like the Eastern Shore, Mathews County, and Fredericksburg in Virginia continued to welcome refugees (Tucker 1972:73). On the ninth of August, the entire block of Barry's Row was burned to the ground in attempts to control the  disease. Nevertheless, the fever continued to spread throughout the cities and beyond. Hundreds of people were contracting the disease and few were surviving. The wealthy areas of Norfolk suffered equally with the less sanitary, poorer areas (Howard Association 1857:109).
Authorities in both Norfolk and Portsmouth took all possible measures to stop the progress of the yellow fever. For example, the Howard Association, a local charity, transformed Norfolk's largest hotel into a hospital to care for the victims (Trask 1996:48). In Portsmouth, the government gave up the naval hospital for care of the sick. The Relief Association was also established to minister to those suffering from the fever. Due to accounts of the epidemic in the national papers, many non-local nurses and doctors provided medical assistance and other people across the country contributed funds and supplies. Provision Stores were opened to issue food and other necessities to those in need (Pollock 1886:159).
The number of deaths continued to climb so much that coffins became a luxury and most were buried in boxes or blankets in common graves. In both Norfolk and Portsmouth, all businesses were suspended and the city functioned mainly as a hospital (Tucker 1972:74). At the height of the outbreak, during the first week of September, there were at least eighty deaths a day in Norfolk and thirty a day in Portsmouth (Burton 1877:21, Goldfield 1973:38). As a result, thousands of more timid citizens left town with their families with the intention of returning after the first frost in October when the epidemic was expected to end. After nearly four months of devastation, the epidemic ended on October 26th, leaving over one third of the populations in both cities dead. The  yellow fever destroyed families and left many children orphaned, for almost no survivor of the epidemic did not experience the loss of a loved one (Burton 1877:23).
Many years before 1855, Norfolk had a reputation of being a sickly place. Once it was paved and drained in the 1820s, health conditions improved and the mortality rate decreased. The common diseases of each season during the 19th century were catarrh (acute influenza) and pneumonia in the winter, and dysentery and cholera infantum (an intestinal disease similar to cholera that affects infants) in the summer (Merriam-Webster 1998). At the end of the summer, Norfolk typically experienced mild, rarely dangerous intermittent fevers. October and November were the healthiest months of the year, whereas June and July had the greatest mortality. The most thickly settled parts of the cities such as Main and Water streets were the healthiest, and before the 1855 epidemic they were usually exempt from yellow fever (Howard Association 1857:97).
During the nineteenth century, people in much of the United States were familiar with yellow fever. There had been occurrences in Philadelphia, New York, and New Orleans. Yellow fever outbreaks had occurred in Norfolk and Portsmouth before in 1795, 1802, 1821, and 1826 but none were as destructive as 1855 (Tucker 1972:72). By the middle of the century, yellow fever had become the South's most feared illness (Trask 1996:46). The cause of yellow fever was unknown at the time. After the 1855 epidemic, investigators evaluated several potential causes which included infected air carried by the wind, poor sanitary conditions, foul bilge water, the crew on board the steamer Benjamin Franklin, and the weather (Howard Association 1857:106,108). Many people believed that a lack of sanitation played a role in spreading the disease. In fact, the leaders of Norfolk delayed announcing the presence of yellow fever in the city  because they did not want to ruin its clean, urban image (Goldfield 1973:37). The belief of uncleanliness as a cause of disease contributed to the prejudiced views towards poor immigrants, who lived in less sanitary areas of the city. Since Portsmouth and Norfolk were important port cities in Virginia, domestic and international trading ships often visited, bringing with them not only goods but disease as well. As a result, quarantines were often placed on ships coming into port to prevent the spread of illness.
People of the nineteenth century knew there was a connection between yellow fever and the weather. For example, the citizens of Norfolk and Portsmouth who left during the outbreak knew that the yellow fever would be gone after the first frost and that it would be safe to return. The citizens were aware of the presence of mosquitoes but never made any connection between them, the climate, and yellow fever. Some believed that yellow fever was a contagious disease. However, the investigators of the 1855 epidemic concluded that there was no reason to suspect that the disease was contagious because it did not spread beyond Norfolk and Portsmouth, despite inflicted citizens who fled and later died of yellow fever (Howard Association 157:110).
During the nineteenth century, many had the prejudiced belief that yellow fever would only infect those who were unsanitary and of lower status (Howard Association 1857:108). They associated the disease with Irish immigrants of Barry's Row because they were seriously afflicted with the disease (Trask 1996:47). However, in sermons during the 1855 epidemic, ministers informed citizens that yellow fever was indiscriminate (Burrows 1855, Cummins 1855, Handy 1855). New immigrants, established families, and children all fell victim to the fever. White citizens noticed during the course of the epidemic that although there were an equal number of yellow  fever cases among Blacks and whites, very few Black inhabitants died from the disease (Howard Association 1857:110). Some believed that the slaves would take advantage of the situation and rebel against their owners (McDaid n.d.:5).
Statement of Problem
This paper examines the yellow fever epidemic of 1855 in Norfolk and Portsmouth in the context of medical anthropology by focusing on two main issues—social responses to epidemics and the role climate plays in the spread of infectious diseases. This case study first evaluates several factors that influenced the behavior of citizens of Norfolk and Portsmouth, including their knowledge of yellow fever, their prejudiced views, and their perceptions of a correlation between environment and disease. It presents a four-stage model of social response to epidemics which can be tested against other case studies. The second part of the paper investigates the role of climate by examining weather conditions during the 1855 epidemic and comparing them to those in surrounding years. Various inferential statistical tests like t-tests and analysis of variance tests were performed to evaluate several hypotheses. These hypotheses and tests attempt to determine the amount of influence climate has on the spread of infectious diseases. Studies have shown that warm-wet climates, like those during El Niño, are more conducive to mosquito-borne disease (Diaz and McCabe 1999). This study evaluates the correlation between El Niño/Southern Oscillation (ENSO) and yellow fever as suggested by Diaz and McCabe (1999).
Understanding the interactions between disease and culture is an important topic in medical anthropology because it is "a productive way of understanding humanity"  (Johnson and Sargent 1990:187). The medical view of a disease can differ from the way people think of it and respond to it, and the latter is what interests medical anthropologists (Mascie-Taylor 1993:94). Societies respond to disease in various ways because the interpretations of it vary from culture to culture. Behavior in illness is strongly influenced by social expectations about disease based on the "social construction of illness" for that culture (Mascie-Taylor 1993:94). People's views depend on "various kinds of social learning and persuasion" which create cultural representations people use to cope with disease (Mascie-Taylor 1993:94). Even within a society, individuals of different ages, genders, socioeconomic groups, and ethnicities are affected differently and respond differently to disease depending on their perspectives (De Salle 1999:155). It is important to understand the way humans respond to disease as anthropologists because it is often in accord with how they perceive or define the situation, reflecting their religious beliefs, knowledge, and culture. People draw on what is familiar to them, so their perceptions are often connected to assumptions based on personal or cultural experience, as opposed to medical science (Knutson 1965:159-162).
Disease must be analyzed and understood in a human context with respect to ecology and culture. The study of disease and human behavior in an ecological perspective is important because diseases are "context dependent" (Johnson and Sargent 1990:208). Cultural practices can directly alter the relationship between the environment, hosts, and agents of disease and, thus, human behavior plays a significant role in the etiology of infectious diseases. Culture contributes to patterns of disease and death because it may shape behaviors that predispose individuals to certain diseases. Through culture, people change the nature of their environment often in ways that affect their  health. According to the archaeological record, environmental change caused by humans can have both positive and negative effects on disease rates (Johnson and Sargent 1990:187). Human behavior such as international trade and travel, large-scale irrigation, deforestation, mining, and misuse of insecticide contribute to the emergence of vector-borne disease (De Salle 1999:61). In contemporary society, with the prospect of global warming, the incidence of mosquito-borne infectious diseases like yellow fever is forecast to increase, threatening developing cultures that are do not have the latest medical technology or receive assistance from other countries (Epstein 2000:50, Shute 2001:50). This paper can contribute to the awareness of the need for preventive methods and education to prevent devastating effects worldwide.
There are four reasons medical anthropologists pay attention to infectious disease (Inhorn and Brown 1997:53-54). First, infectious diseases act as important agents of selection for both behavioral and biological characteristics of species. Second, human actions that affect ecology significantly influence the distribution of infectious diseases. Third, infectious diseases are a leading cause of suffering and deaths in societies frequently studied by anthropologists. Finally, by learning about disease, medical anthropologists can be more effective in improving control programs and health care.
Reconstructing infectious disease history allows people to "draw the best inferences possible from past experience; for this, history can be a valuable guide" since we cannot foretell the future (Inhorn and Brown 1997:18). It also demonstrates the potential danger of uncontrolled infectious disease. By studying these reconstructions, people can learn about the biological history of the disease's evolution, the cultural  history of human response to epidemics, and the history of public health measures (Inhorn and Brown 1997:18).
Medical anthropologists and historians have identified different responses to outbreaks of disease, ranging from panic to resignation and hysteria to heroism. Some have identified specific stages of social response. Humphreys (1992) suggests that the basic stages of epidemics are: 1) first cases appear, 2) physicians and government officials attempt to keep it secret, 3) the disease spreads, 4) people flee in panic, 5) preventative measures are taken but it is too late, 6) hundreds or thousands die, and 7) examples of heroism are identified (Humphreys 1992:11). Rosenberg (1989) recognizes just four stages of response. In the first stage, people acknowledge the disease and that it threatens social order, risking panic and social dissolution. The second stage consists of efforts to manage and understand the randomness of death, and attempt to explain the death toll. Rosenberg's third stage occurs when social response is negotiated using various policies and rituals. The final stage is the process of the epidemic slowing and society realizing the fragility of order, limits of science, and the essential balance between humans and the environment (Doka 1997:24).
Doka (1997) describes society's understanding of the disease as influencing behavior through his five stages of denial, avoidance, scapegoating, transcendence, and art and literature. In the stage of denial, people believe that the disease will not affect their city, class, or group. Economic and commercial interests push city leaders towards denial as a means to protect the city's urban image. A society turns to the avoidance stage when denial is no longer possible because the disease is upon them. This stage is reflected in their actions through flight, policies such as quarantines, and avoidance of  those infected. The third stage of scapegoating is a result of the failure to scientifically explain the cause of the disease. Transcendence is the next stage that reflects the religious interpretation of the disease. Finally, art and literature, such as "Ring Around the Rosie" and the Pied Piper, act as an "emotional counterattack" to epidemic death by allowing society to recapture "a sense of control over disease and thereby lessen the unmitigated horror and devastation of death" (Doka 1997:16-20).
Studies recognize how epidemics tear apart societies by causing panic, dread, prejudice, and fear. Smelser (1963) states that panic results during an epidemic when a general feeling of unease and ambiguity exists in society and is confirmed by the spread of disease. Panic usually causes mobilization for flight (Doka 1997:54-55). Dread occurs in a society when medical science is uncertain as to how the disease is transmitted and every person and act is seen as a potential source of danger. If the population does not trust the medical science, doctors' theories will not make a difference on social response. Dread typically leads to irrational panicked reactions (Doka 1997:53). Epidemics also instigate religious, racial, class, and political conflicts because people openly display their prejudice and fear. Punitive policies are usually implemented to expel or isolate those suspected of spreading disease and reflect society's response to political, social, and moral interpretations of the epidemic rather than a medical one. Fear shapes the public health policies and medical treatment. As fear abates, the disease is viewed more as a medical issue and people respond more rationally. Fear more than anything else shapes social response (Doka 1997:57).
A combination of documentary and climate research was conducted to address the problems described in the preceding section. The majority of research was conducted last summer and fall while working at the William and Mary Center for Archaeological Research under a National Science Foundation grant to reconstruct the historical climate of the southeastern United States. Related analysis was performed in conjunction with a Statistics of Anthropology course during the Fall 2000 semester.
This paper draws on information from primary, nineteenth century sources to provide details of the 1855 epidemic and climate. The College of William and Mary's Swem Library, the Virginia Historical Society in Richmond, the University of Virginia's Alderman Library, the Colonial Williamsburg Foundation Library, and the Mariners' Museum in Newport News all hold such sources.
Next, several types of non-traditional "proxy" sources dealing with climate and yellow fever including diaries, farm journals, sermons, and newspaper articles were surveyed. Each source was evaluated based on whether or not it contained useful data on yellow fever and climate, if it was written prior to 1870, consistency of records, if it was a firsthand or published account, etc. The sources with climate data were also ranked according to frequency of recordings, whether it had instrumental data or relative observations, and the proximity to Norfolk and Portsmouth. After narrowing the topic to the specific aspects of climate and social responses of the 1855 yellow fever epidemic, information was extracted from the best sources. Descriptions of the epidemic that  included events, social responses, and influence and observations of climate were noted. For example, The Report of the Howard Association of Norfolk (Howard Association 1857) describes the investigative process that the committee of doctors followed in attempts to identify the cause of this devastating epidemic. This paper analyses this report and other primary sources as windows into the perspectives held by the citizens at the time of the epidemic. Then, the collected data was analyzed and organized into specific stages of social responses to epidemics, which will be further discussed in Chapter Three.
After examining these non-traditional "proxy" sources, the World Wide Web was searched for magazine articles and books written on medical anthropology, global warming, infectious disease, and epidemics. Web pages of various universities, which offer medical anthropology courses, were visited to find useful sources. The World Health Organization and Center for Disease Control web pages were consulted to find current information that dealt with the issues of epidemics of infectious disease and global warming.
Statistical Analysis of Climate's Role
Most of the climate research was conducted during last summer and fall while working at William and Mary's Center for Archaeological Research. In order to reconstruct the climate of 1855, the methods stated above were followed. Several sources were evaluated based on criteria that included proximity to Norfolk and Portsmouth, consistency of instrumental data, and supporting detail of weather. The two data sources selected for this case study were the weather records from Fort Monroe and  Crichton's Store because both met the necessary criteria (Smithsonian Institute 1853-1861, United States Army 1824-1892).
Next, both descriptive and inferential statistics were used to analyze climate's role in the 1855 yellow fever epidemic. Descriptive statistics provide a means of characterizing central tendencies, variation, and trends of variables through charts, graphs, and exploratory data analysis. Inferential statistics consists of systematic procedures for predicting unknown parameters from samples, which represent the population being evaluated. The weather data for this paper was explored using SPSS, a statistics computer program, to examine the changes in climate over time and to see if the year 1855 was anomalous and more conducive to the yellow fever mosquito, Aedes aegypti, compared to its surrounding years. Knowing that the female Aedes aegypti mosquito needed specific weather conditions to not only survive but to also breed and be highly active, the monthly average temperature, diurnal range, monthly precipitation, number of precipitation days, and number of frost days were examined. Inferential statistical tests, such as t-tests and analysis of variance tests, were used to evaluate the following five hypotheses:
1) 1855 was significantly warmer than its surrounding years.
2) 1855 had a significantly narrower diurnal range.
3) There was a significantly greater amount of rainfall in 1855.
4) There were significantly more rain days in 1855.
5) There were significantly less frost days in 1855.
Finally, the data was interpreted and applied the analysis of climate's role not only in the 1855 yellow fever epidemic but also to present and forecasted outbreaks.
 Chapter Three: Presentation of Findings
This chapter discusses social responses and weather conditions during the Norfolk and Portsmouth yellow fever epidemic of 1855. The first section describes the four stages citizens experienced during the outbreak in both a historical and cultural context. The section describing weather conditions compares the climate of 1855 to other years. Inferential statistics were applied to evaluate conditions of the epidemic to see if they were anomalous and more conducive to the Aedes aegypti mosquito.
Social Response to Disease in Norfolk and Portsmouth
As reviewed in the preceding chapter, medical anthropology is concerned with cultural responses to epidemics. Medical anthropologists have identified different responses to outbreaks of disease. The purpose of this paper is to explore the generalizations of medical anthropologists about social stages of response in the context of this case. Specifically, the cultural sequence of events that surrounded this outbreak of yellow fever are determined and evaluated.
Although medical anthropologists have identified up to seven different stages of social response to epidemics, this paper recognizes only four stages in the Norfolk and Portsmouth yellow fever outbreak: 1) initial response, 2) initial spread, 3) climax, and 4) recovery. In the first stage, the initial response to yellow fever is examined and compared in both cities. It begins with the arrival of the Benjamin Franklin in Portsmouth and the appearance of the fever in Barry's Row in Norfolk. The second stage describes the citizens' response to the initial spread of yellow fever throughout the cities.  The spread of the fever beyond Barry's Row marks the beginning of this stage. The climax of the epidemic is examined in the third stage. This stage begins with the northeasterly storm during the first week of September which caused a dramatic increase in yellow fever cases. The final stage deals with recovery after the epidemic ended with the first frost in October.
These four stages are the most pragmatically defined stages of social responses to disease because they are general and widely applicable. Seven characteristics are distinguished within each stage, describing the varying degrees of expression relating to the factors of the spread of disease, number of cases and deaths, social concerns and behavior, and impact on the city (Table 1). In effect, they also incorporate the stages identified by medical anthropologists and historians (Doka 1997, Humphreys 1992, Rosenberg 1989, Smelser 1969). Some overlap of characteristics occurs between stages, especially between Stage Two (Initial Spread) and Stage Three (Climax).
Stage One: Initial Response (June 4 - July 30, 1855)
Stage One described the initial social responses to disease (Table 1). Norfolk and Portsmouth experienced Stage One at different times during the 1855 epidemic. The first yellow fever cases appeared in Portsmouth with the arrival of the steamer Benjamin Franklin on June 6th, whereas Norfolk did not encounter the fever firsthand until the end of July. This section evaluates the experiences of the "Twin Cities by the Sea" by comparing and contrasting their initial responses to yellow fever.
During Stage One, disease was restricted to a limited area. When the Benjamin Franklin came into the port at Portsmouth on June 6th, it was quarantined as a precaution
 Table 1: Stage Characteristics of Social Response to Epidemics
 because it came from St. Thomas where yellow fever was present. Her captain denied that any cases existed on board when interrogated by health officials. After a few days, the steamer was allowed to leave the quarantine ground and its crew was permitted to enter the city (Pollock 1886:157). On the 30th of June, the first cases of yellow fever appeared in the poor areas of Gosport in the vicinity of Page and Allen's Shipyard (Howard Association 1857:105). It gradually spread to the Irish Row communities along the Water Street front of Portsmouth. At first, the citizens expected the epidemic to be confined to Gosport and the Irish communities in Portsmouth, but on July 29th cases appeared across the river in Norfolk's Barry's Row.
There was minimal apprehension in both Portsmouth and Norfolk during this earliest stage. Portsmouth locals were not anxious until the death of a young man who assisted the Benjamin Franklin on June 8th. The Town Council immediately ordered the return of the steamer to quarantine ground, but the yellow fever had already begun to spread (Pollock 1886:157). Even then, many families stayed in Portsmouth, believing that the epidemic was a mild one. However, some Gosport families, where the disease was spreading, crossed the river into Norfolk to escape the fever. Believing the Elizabeth River could act as a barrier, Norfolk residents hoped the fever would not spread (Wertenbaker 1931:210).
When the yellow fever was recognized in Norfolk on July 30th, citizens were both doubtful and alarmed. The attempts of political leaders to protect Norfolk's urban image by not acknowledging the existence of the epidemic caused people to question the existence of yellow fever in the area. The city newspapers did not even make note of it. For example, Hugh Grigsby of the Southern Argus did not write a report on the outbreak  for six weeks. Dr. Upshur, the physician who treated the first case of the fever, waited almost two weeks before he shared with the public that he believed he was treating a victim of the fever. Some people blamed him for not telling the public sooner so that precautions could have been made. Others blamed him for being an alarmist and ruining the business of the city (Armstrong 1856:7).
Denial was a distinguishing characteristic of initial response to disease. In Norfolk and Portsmouth, locals thought that they would not be affected by the fever because their cities were healthy, clean, and, thus, immune to disease. The citizens also felt that the presence of yellow fever could not be attributed to climate or sanitary conditions since other cities with the same conditions were unaffected (Burton 1877:20). Because people associated yellow fever with filth and refuse, it was not surprising to many when the fever struck in 1855 that it first infected the poor areas of town. Most believed that the disease would remain in Gosport, the Irish Rows of Portsmouth, and Barry's Row in Norfolk.
During the earliest stage, few preventive measures were implemented. In Portsmouth, this was in part a result of the Benjamin Franklin captain's denial that cases of yellow fever existed on board. However, after the first death caused by yellow fever, the Benjamin Franklin and other incoming ships were placed under stricter quarantine (Wertenbaker 1931:210). In the 1850s, Norfolk was extremely concerned with its urban image because it wanted to compete with the trading cities of the north (Goldfield 1973:34). As a result, city officials attempted to squelch news of the spread of yellow fever to Norfolk and did not implement many obvious preventative measures. The Board of Health, however, quarantined vessels on July 26th because it did not want a repetition  of the fatal summer of 1852 when several hundred local cases were reported and 50 to 100 citizens died (Parramore et al. 1994:176). On July 30th, the Board of Health conceded that an epidemic existed. Barry's Row was vacated, and ailing tenants and their families were exiled to a hastily established hospital outside the city limits (McDaid n.d.:3). A 24-foot high board wall was also erected around the neighborhood in hopes of restricting the yellow fever (Parramore et al. 1994:177).
In Stage One, city politicians were more concerned with the city's reputation than with the actual epidemic. Pestilence damaged the coveted successful and healthy image that officials desired. In the pre-Civil War South, urban image was vital to "promote commerce, attract industry, and improve facilities for transportation" (Goldfield 1973:34). Urban leaders often hesitated to recognize the beginning of an epidemic because they did not want newspapers of competing cities to spoil their image by calling the city "sickly," which was the most devastating charge a city could suffer (Goldfield 1973:37).
During epidemics, closely held prejudices often become more apparent (Doka 1997:20). Norfolk citizens anticipated that yellow fever would infect Barry's Row as a result of their belief that the poor Irish immigrants who populated the area were less sanitary (Armstrong 1856:12). Barry's Row was regarded as a "perfect cess-pool of pestilence" with its row of shanties and inhabitants living in a degraded state (Parramore etal. 1994:177).
Much of this prejudice in Norfolk and Portsmouth was the result of a new political alliance, Know-Nothingism, which was founded on opposition to American Roman Catholicism and liberal immigration laws. The Beacon newspaper, Whigs,  shipyard workers, and other members of the community supported it. Know-Nothingism won control of city government in 1854 and retained most offices in 1855. Although Norfolk traditionally maintained a tolerance for religious and national minorities, Know-Nothingism thrived and resentment toward the city's Irish Catholics and other immigrant workers grew. The Beacon described immigrants as "scum of other countries who would rather beg than work, and rather steal than beg" (Parramore et al. 1994:180).
Although most white Southerners assumed that Blacks were immune to the fever, there was still some prejudice held towards them. Some people feared that the slaves, who seemed to possess a genetic immunity to yellow fever, would rise up against the weakened citizens of Norfolk and rebel (McDaid n.d.:5). However, for the most part, the Black population was virtually ignored compared to the poor white segment of the population, which was at the mercy of the upper classes (McDaid n.d.:5).
The 1855 yellow fever epidemic in Norfolk and Portsmouth began with the seven characteristics described for Stage One (Initial Responses). The fever was restricted to the poor areas of both cities; as a result, there were not many cases or deaths. Most citizens were not apprehensive at this stage of the epidemic because they believed the disease would not affect them. Few preventative measures were implemented by city authorities since they were either unaware of the disease or did not recognize it as a result of their concern for the city's reputation. Finally, the prejudiced views of the locals were evident by their attitude towards the spread of the fever and their treatment towards those affected.
 Stage Two: Initial Spread (August 1 - August 31, 1855)
Stage Two described the citizens' response to the initial spread of disease beyond Barry's Row (Table 1). Inhabitants were more panicked and attempt to escape the epidemic through flight. As the cities began to feel the effects of the epidemic on community services, non-local volunteers tended to the sick and contributions were received from abroad.
In Stage Two, more people were severely affected by the epidemic as a result of the disease's spread. By August, not only the poor and overcrowded were afflicted, leading families were as well, including Norfolk's Mayor Woodis (Parramore et al. 1994:179). No yellow fever had occurred in the wealthy and aristocratic section of Norfolk prior to the Benjamin Franklin's arrival (Burton 1877:20). In earlier epidemics, "persons living in that district (i.e., Woodside's Lane, Upton Street, and South of Main Street) had just to move to the North of Main Street and they were as safe from the fever as they would have been several miles off (Wertenbaker 1931:208).
The yellow fever's spread at first was mild and slow but became rapid and malignant. Citizens in Norfolk were dying at a rate of 60, 70, or even 80 persons per day, and in Portsmouth the rate 20 to 30 per day. As is expected for Stage Two, people were seized with terror with new yellow fever cases. The epidemic gripped a community that was totally unprepared, paralyzing the citizens with panic (Pollock 1886:159). When the disease was beyond the control of the health authorities, citizens fled in all directions "from the frightful scenes of disease, wretchedness, and woe—amazed and horror struck at the savages of the unsparing agent of destruction" (Burton 1877:20). Many decided to leave their business rather than face the danger of death. Trains and steamers were  completely filled (Wertenbaker 1931:211). At the beginning of August, Portsmouth's streets did not have the same bustle and activity because three-quarters of the population had fled. By the middle of August, half of Norfolk's population had left the city (Armstrong 1856:13-14). By the end of the month, only one-third of the white population was left in Norfolk (Armstrong 1856:18).
Most of the fear stemmed from the fact that so little was known about the causes of disease and epidemics, or effective preventative measures and treatment. It was instinctive for people to associate disease with refuse and bogs but none could guess the precise cause or a way to combat it (Parramore et al. 1994:179). Authorities and many others believed in contagion theory and thought that yellow fever could spread relatively easily from city to city, and that quarantine of incoming goods and people from infected areas limited and prevented the spread. Many cities also attempted to prevent the spread by prohibiting their inhabitants from entering affected areas (Oldstone 1998:53). At the same time, people possessed the idea that the yellow fever was not transmitted person-to-person but by contact with infected houses or decaying matter (Wertenbaker 1931:210). Many believed in this miasmatic theory which stated that disease originated from the decay of vegetable and animal matter through the action of heat and moisture (Harper's New Monthly 1857:196-197, Trask 1996:52). As a result, people assumed that damp cellars, the proximity of marshes, unsanitary and crowded living conditions, the presence of pig pens in town, and the pumping of bilge water near wharves were all factors (Wertenbaker 1931:206). Only elements of the germ theory were acknowledged in the 1850s. Doctors tended to cure symptoms and not disease by prescribing diuretics,  purgatives, and blistering agents to restore the body's "harmony" which caused more harm than good (Trask 1996:50).
The citizens of Norfolk and Portsmouth knew from past epidemics to respect yellow fever. Some preventative measures taken before 1855 included strictly enforcing quarantines, filling shallow and stagnant "slips," tearing down unsanitary houses, and keeping street drains cleared (Wertenbaker 1931:209). In a 1795 Norfolk epidemic, people fled the city at first approach of the epidemic, which saved hundreds of lives (Wertenbaker 1931:209).
During the summer of 1855, locals used these experiences to their advantage to fight the disease and its spread. In Portsmouth, officials organized the Relief Association to attend to the needs of the sick and suffering, and to manage the expenditure of contributions from the rest of the country. Provision Stores were opened to issue food and other necessities to the public. The government also gave up a naval hospital for the care of victims of the fever (Pollock 1886:159). In Norfolk, tar fires were set in streets to combat the deadly air but had no effect. The sanitary committee inspected streets and lots for places that might breed disease, such as rain barrels, food tins, and fountains, and ordered physicians to report new cases by sunset (Parramore et al. 1994:177). On August 9th, Barry's Row was burned in front of 2,000 to 8,000 approving, frightened onlookers in attempts to stop the spread and rid Norfolk of any remaining infection (Armstrong 1856:16, McDaid n.d.:4, Parramore et al. 1994:178, Wertenbaker 1931:210).
By this second stage, cities surrounding Norfolk and Portsmouth also found it necessary to implement preventative measures. At first, the surrounding communities allowed refugees to enter but as the epidemic worsened, cities like New York, Richmond,  and Petersburg, began closing their doors, forcing the citizens of Norfolk and Portsmouth to become not only refugees but fugitives as well. Quarantines, such as these, were suggested more by fear than coherent medical theory (Doka 1997:57). Norfolk's "slalom" into hysteria worsened as a result of restrictive measures and fines enforced by the places where people sought refuge (Parramore et al. 1994:178). Desperate to escape, people illegally caught trains and ships to cities with quarantines, or sought refuge in the surrounding country in private homes, cabins, schoolhouses, churches, tents, and barns (Parramore et al. 1994:179). The Eastern Shore, Mathews County, and Fredericksburg in Virginia continued to offer asylum (Tucker 1972:73). At the end of July, all major Virginia cities and other important port cities also issued interdicts against trade with Norfolk (Goldfield 1973:38).
During Stage Two, there was also a modest impact on basic community services. In both cities during the epidemic, the yellow fever put an end to the order of everyday life. Stores and hotels closed. Yellow fever halted the publication of city newspapers (Goldfield 1973:38). The government gradually collapsed as people fled or contracted the fever. The quarantines also disrupted the inbound and outbound flow of goods and people (Trask 1996:52). The post office moved with other businesses onto Main Street, while the hospital moved to Lambert's Point where there was "purer air" (Armstrong 1856:18). By the end of August, most business in Norfolk was suspended and the city became one great hospital (Tucker 1972:74, Wertenbaker 1931:211).
Another characteristic of Stage Two was assistance received from abroad. When accounts of yellow fever appeared in the nation's papers, volunteers from across the country came to nurse victims, donated supplies, and collected money (Tucker 1972: 73).
 Physicians and nurses volunteered to minister to the sick and dying when local staff became exhausted. Many of these heroes knowingly risked their lives by exposing themselves to the fever for the first time (Trask 1996:48). Some of the doctors had experience with the fever and did not consider it contagious (Armstrong 1856:17). Clergymen from various churches remained faithful to their mission and also helped the sick (Pollock 1886:159).
In summary, Stage Two (Initial Spread) marked the period when there was a gradual loss of community cohesiveness as people continue to act irrationally, trying to escape disease and death. Areas of Norfolk and Portsmouth that had not been previously affected by the fever were now coping with cases. A significant increase in the number of cases and deaths caused people to flee the city in panic. The epidemic's spread also terrified other cities which started to quarantine refugees from Portsmouth and Norfolk. Quarantines and the flight of citizens affected basic community services (i.e. businesses and government), and several nurses and physicians from across the country volunteered their services to combat the disease.
Stage Three: Climax (September 1 - October 25, 1855)
Stage Three corresponded with the climax of the epidemic of 1855 (Table 1). This stage was marked by a severe impact on community services as the number of cases and deaths drastically rise with the unstoppable spread of disease. In the case of the Norfolk and Portsmouth epidemic, a northeasterly storm during the first week of September marked the beginning of Stage Three.
 In Stage Three, the epidemic affected the majority of the city; it is no longer restricted to a limited area. In September of 1855, the yellow fever followed this pattern. The first week of September was one of the most terrible times of the epidemic for the yellow fever was at "its most appalling fury" (Burton 1877:21). It continued to spread from the poor sections of Norfolk and Portsmouth into areas that were previously unaffected by disease. Yellow fever was beginning to be seen as more than an "immigrant's disease" or one that only affected the impoverished because rich and poor alike were becoming victims. Armstrong (1856) described how another "chill northeasterly" storm caused the yellow fever to spread into the "heart of the city" (Armstrong 1856:20). Over half of this section of the city had never experienced yellow fever firsthand (Armstrong 1856:20).
The most alarming characteristic of Stage Three was the dramatic increase of cases and deaths that occurred as a result of the increased spread of the epidemic. The number of cases more than doubled by the end of the first week of September. There were 300 to 400 cases of yellow fever at the beginning of the week and by the end, that number grew to 1200 to 1500 cases. Entire households were sick (Armstrong 1856:20). As a result of the numerous cases of yellow fever, the Howard Association took over a hotel in the center of Norfolk (Armstrong 1856:23). The fifth day after the "chill northeasterly" storm was a terrible day, for most yellow fever deaths occur on the fifth or seventh day (Armstrong 1856:23). People of every class were "falling like withered leaves shaken by the winds" (Burton 1877:21). Over 400 citizens were buried in that one week. Forrest wrote that "no pen can adequately portray the horrors of that dark period,  which as brief as it was, has sufficed to produce an age of misery and woe" (Burton 1877:21).
At Stage Three, people were unable to escape and must face the disease. This caused despair for many. When the yellow fever raged in the wealthy and aristocratic section of Norfolk, it was too late to flee (Wertenbaker 1931:214). By this point, most families were affected by the fever either with death or illness and did not want to leave any family members behind. Many of the city's inhabitants did not have the resources to flee the fever. From the beginning of the epidemic, slaves, free Blacks, and working class whites had little means of escape (Trask 1996:46). It was a time of intense excitement and consternation (Burton 1877:21). People were "overcome by the threatening calamity" and many thought the pestilence would wipe them out (Armstrong 1856:27-28). People saw yellow fever as similar to a mole "work[ing] beneath the surface and beyond the range of human sight," causing fear and dread throughout the city (Armstrong 1856:43).
A deserted appearance of the city was another characteristic of Stage Three. At this stage of the epidemic in 1855, the city and cemetery switched characters (Armstrong 1856:29). "The city was wrapped in gloom. . . [with] no sign of wholesome animation" (Wertenbaker 1931:214). Most people had fled and those that remained were tending to the sick. Norfolk resembled a ghost town with its wharves, streets, and businesses completely deserted (Goldfield 1973:38). The post office, which was usually a cheerful and social place even during Stages One and Two, had changed for there was no crowd (Armstrong 1856:33). The harbor was desolate with only the J.E. Coffee periodically coming into port to get mail and bring coffins (Armstrong 1856:30, Tucker 1972:74).
 In Stage Three, there was a severe impact to the services of the city. The yellow fever brought the local economy and functions to a stand still. Doctors were too busy and feeling powerless because they could do little to cure or comfort their patients (Armstrong 1856:21,37, McDaid n.d.:3). As a result of the numerous deaths, proper funeral services could not be held. Black inhabitants were pressed into service as grave diggers, hearse drivers, and even as nurses because they seemed to have little fear of the disease (McDaid n.d.:6). There was also a shortage of coffins even though it had been the most frequently imported item during the epidemic. Many victims were buried in boxes, blankets, or in communal graves (Armstrong 1856:29). In Norfolk, the city government was hardly functioning as Mayor Woodis died from the fever, the acting mayor was sick, and the other officials in the city were also dying (Armstrong 1856:31).
Speculation as to the cause and origin of the epidemic was a feature in Stage Three. In 1855, there were two main theories of the origin of yellow fever, it was either local or imported. Proponents of the local origin theory, or miasmatic theory, believed that the combination of heat, rainfall, humidity, swamp gases, and filth gave birth to yellow fever. They also thought that the disease spread by inanimate carriers like clothing and luggage. These theorists saw a massive citywide cleanup as the solution. Others argued that the fever was imported because they believed their city was a healthy living environment. They believed quarantines were the answer (Trask 1996:52).
The citizens of Norfolk and Portsmouth also recognized the influence of the weather on the epidemic. A few people, including Armstrong, believed that the yellow fever spread most rapidly in dry, hot weather and during seasons of drought. Armstrong (1856) was confused by the weather of 1855, which was unusually "seasonable, not as  much extremely hot weather as is usual," and yet the fever was spreading slowly on dry, warm days and rapidly during "chill northeasterly" storms that were frequent throughout the epidemic. The citizens of Norfolk and Portsmouth were conscious of the influence the storm during the first week of September had on the dramatic increase in cases and deaths caused by yellow fever (Armstrong 1856:20-21). Other people believed that warm weather affected the spread of disease. For instance, Grace Whittle noted in her diary that the disease appeared to gain strength as the intense heat increased (McDaid n.d.:4).
Stage Three (Climax) marked the period when there was a total breakdown of the community with the loss of control over the epidemic, as in the yellow fever epidemic of 1855. Most sections of the city were infected by yellow fever, including those of the wealthy and aristocracy. With the first week of September came another northeasterly storm, which was seen by some as the cause of the dramatic increase in cases and deaths. The inability of the citizens to escape caused great despair. Norfolk and Portsmouth appeared deserted as a result of the majority of the citizens having fled or been affected by the fever. The large number of cases and deaths and the quarantines had a severe impact on services, supplies, and everyday life. Citizens also began to speculate about the cause of the epidemic, trying to decide if it was of local or imported origin.
Stage Four: Recovery (October 26 - December 31, 1855)
Stage Four described the recovery of the cities after the epidemic (Table 1). These communities were in a state of complete social disruption and attempted to regain control  and recover from the devastation of the epidemic. This final stage began with the first frost in October which brought an end to the yellow fever.
The most distinguishing characteristic of Stage Four was the end of the epidemic when no new cases occurred. After the first frost in October, Norfolk "lay suffering, stunned, still unable to grasp the meaning of the fearful calamity" (Wertenbaker 1931:214). In nearly four months, there were between 8,000 and 10,000 yellow fever cases—nearly the entire population that remained in the city. Very few citizens escaped its effects with the exception of those who had contracted the disease before. Black residents were as susceptible to yellow fever as whites but they handled the disease much better and were able to recover. About 2,000 people or one-fourth of the entire population remaining in the city "perished by its ravages," a loss from which the city had not wholly recovered by 1886 (Howard Association 1857:110, Pollock 1886:161). The last fever fatality was on November 10, 1855 (Pollock 1856:163).
In Stage Four, many who fled during the epidemic returned to the cities. Half of the refugees came back to Norfolk and Portsmouth after the first frost in October because they knew the fever would not continue to spread (Armstrong 1856:46). However, a considerable number decided not to return (Parramore et al. 1994:191).
A systematic search for a cause was initiated in Stage Four. After the summer of 1855, a committee of physicians was organized by Norfolk's government "to investigate the cause and origin of the yellow fever of 1855" (Howard Association 1857:95). The investigation took place over a period of nearly eighteen months (Goldfield 1973:39). They searched for an answer to the question, "is the yellow fever of local origin, or an imported disease," in hopes of saving the prosperity of Portsmouth and Norfolk (Howard  Association 1857:96). They compared the climate of healthy years in Norfolk to the year of the epidemic, seeking to identify potential meteorological and imported causes. However, Norfolk's unhealthy image was not improved by the physicians' report of yellow fever being an "imported disease" (Parramore et al. 1994:192).
The idea that mosquitoes caused yellow fever was advanced by a few people but never proven with experimental evidence until Walter Reed's study at the beginning of the 20th century (Oldstone 1998:61). In The Manhattaner in New Orleans: or Phrases of "Crescent City" Life (1851), Abraham Oakey Hall, a lawyer, journalist and future politician, recognized the major components of epidemics as fresh water, warm temperatures, mosquitoes, a confined population, trade with the West Indies, and the fever itself, without understanding the connections between them (Trask 1996:49). Armstrong dared to express the opinion that yellow fever was not contagious and almost pinned the mosquito as the cause (Wertenbaker 1931:215). Armstrong, for example, stated in September that the "plague-fly" (mosquitoes) was collecting around the houses and that the crisis of the epidemic was approaching (Armstrong 1856:30). He wrote that he first noticed the "plague-fly" on August 31st. He also observed that they were in great numbers during the first week of September and had almost disappeared by September 13th (Armstrong 1856:47). Black inhabitants also believed that the "plague-fly" spread the disease by consuming the "morbific" matter that caused yellow fever (Armstrong 1856:46).
After the end of the epidemic in Stage Four, the cities dealt with significant social disruptions. The yellow fever epidemic of 1855 shattered Norfolk during its "Reform Age Renaissance" (Parramore et al. 1994:190). It was a blow to public morale,  leadership, labor supply, and business relations. People blamed the Norfolk's Know-Nothing political leaders because they had given less attention to the yellow fever in its critical early stages as a result of their prejudiced belief that it was an "immigrant's disease" (Parramore et al. 1994:191). Some saw the outbreak as a curse as a result of the prejudice and hatred caused by Know-Nothingism (Parramore et al. 1994:180). The fever destroyed the finances of the Norfolk and Petersburg Railroad and brought the economy of the city to a standstill (Goldfield 1973:38). There were also a number of orphans and people who lost their jobs and homes (those who lived in Barry's Row), which was a long-term burden to both cities (Parramore et al. 1994:192).
During the final stage, Norfolk attempted to reconstruct its good reputation. Diseases like yellow fever reinforced "Dixie's image as a poor, disease-riddled region" and they were important factors that contributed to economic stagnation (Trask 1996:54). For a while after the epidemic, it was questionable if Norfolk would recover (Parramore et al. 1994:191). In 1859, the advancement of Norfolk was reported by the Southern Argus as being "slow, too slow." It also blamed the plague for "melting away the population like snow" and shaking the self-confidence of the city (Goldfield 1973:40). The 1855 yellow fever epidemic was a catalyst for public health and social reforms in Norfolk and Portsmouth. An orphanage was established from the monetary donations (Pollock 1886:163). Despite attempts to reestablish its urban image, Norfolk was unable to share in the successes of other port cities. It never did enjoy the success of the tobacco trade as much as Richmond and other Virginia cities (Goldfield 1973:40).
Stage Four was also characterized by honoring the heroes of the epidemic. Monuments of heroes of the plague and prayers of thanksgiving reminded citizens of  their collective tragedy (Goldfield 1973:39). For example, a monument was erected to memorialize all the volunteers from across the country who sacrificed their lives tending to the yellow fever victims in Norfolk and Portsmouth. Twenty-six out of forty-five volunteer physicians died (Wertenbaker 1931:215).
The yellow fever outbreak in Norfolk and Portsmouth displayed the typical characteristics of the final recovery stage. The end of the epidemic was marked by no new cases or deaths. Many refugees returned after the first frost in October when they knew it would be safe. Norfolk and Portsmouth were socially disrupted after the fever because they experienced high death tolls and a shortage of goods throughout the epidemic. Norfolk made fruitful attempts to reestablish its urban image but continued to lag behind other port cities. The epidemic caused many to realize the limits of science which instigated research and investigations of the disease. The heroes of the epidemic were memorialized.
In summary, the social responses of the citizens of Norfolk and Portsmouth during the epidemic of 1855 can be categorized into four significant stages—initial response, initial spread, climax, and recovery. Each stage (except Stage Four) shows an increase in areal spread, number of cases and deaths, emotional stress, and general disruption of the city. Some characteristics, such as the impact on basic community services and the deserted appearance of the town, overlap between stages. The various stages demonstrate how society interprets and understands disease through the citizens' behavior and emotions.
 Climate Data
This section of the chapter focuses on weather conditions during the yellow fever epidemic in Norfolk and Portsmouth. It examines the climate of 1855 through the use of descriptive and inferential statistics and compares the conditions to other years to see if they were anomalous and more conducive to the Aedes aegypti mosquito and yellow fever.
The female Aedes aegypti mosquito is responsible for the spread of yellow fever. It is a domestic mosquito that lives close to humans, depending on them for blood meals and breeding in puddles or containers of water near their homes (Cooper and Kiple 1993:1101). The surface of calm and relatively clean, fresh water in rain barrels, roof gutters, bilge water, and food tins in the nineteenth century served as breeding tanks (Trask 1996:48). Aedes aegypti thrives in a fairly closely packed population because her range is very short. Rainfall is a prerequisite for urban yellow fever because A. aegypti can only survive a few days without water and requires it for breeding. Hot weather is also required because A. aegypti will not bite when the temperature is below 62° F (Cooper and Kiple 1993:1101). The optimum temperature for activity, breeding, and feeding is from 81° F to 88 ° F. At a temperature of 75 ° F, the mosquito feeds an average of every four days. Both high humidity and moderately high temperature favorably affect the duration of life and activity of A. aegypti. Temperatures below 32° F and above 106° F kill A. aegypti (Diaz and McCabe 1999:21).
Aedes aegypti acts as a vector in the transmission cycle (Figure 3). Through blood feedings, it carries the disease from one person to the next. This exchange must occur during the first three to six days of infection in yellow fever victims because that is  when the disease is still present in the blood. After the virus has entered the mosquito, it must incubate for a four day to two week period before the mosquito can infect another human being (Diaz and McCabe 1999:21). After this period of incubation, the mosquito will remain infective for the remainder of its life, which is typically one to two months (Cooper and Kiple 1993:1101).
Figure 3: The yellow fever transmission cycle (from Mattingly 1969:20)
Knowing that Aedes aegypti requires specific weather conditions to survive, breed, and be highly active, the monthly average temperature, diurnal range, monthly precipitation, number of precipitation days, and number of frost days are examined to see if the weather of 1855 was anomalous compared to its surrounding years. The hypotheses, as stated in Chapter Two, attempt to answer the question: did 1855 experience the ideal weather for Aedes aegypti!
1) 1855 was significantly warmer than its surrounding years. In higher temperatures, Aedes aegypti has a high amount of activity, breeding and biting more. The high number of mosquitoes in Norfolk and Portsmouth would have increased the possibility of being exposed to yellow fever.
 2) 1855 had a significantly narrower diurnal range. The diurnal range is the difference between the minimum and maximum temperatures in a day. If the difference is small, then there is less change in temperature. In the case of 1855, if the diurnal range was narrow, there would have been less cooling, which is conducive to Aedes aegypti.
3) There was a significantly greater amount of rainfall in 1855. A large amount of precipitation leads to higher humidity, which increases the activity of Aedes aegypti. Rain also provides areas with still water for mosquito breeding.
4) There were significantly more rain days in 1855. With a higher frequency of rain, humidity remains at a higher level. If there are less dry periods, then breeding areas are available for Aedes aegypti.
5) There were significantly less frost days in 1855. With less frost days, the growing season would be longer and the temperature would be warmer. There also would not be any drastic drops in temperature that would kill the mosquitoes in the area.
The two data sources used in this research were the weather records from Fort Monroe and Crichton's Store (Smithsonian Institute 1853-1861, United States Army 1824-1892). Each data source had records of daily temperature, amount of precipitation, and precipitation days. Fort Monroe is located near Norfolk and the data was complete for July through September for 1850 to 1860. These three summer months are crucial when studying the yellow fever epidemic of 1855 because they coincide with the middle of the epidemic. The ten-year span is also necessary to identify potential differences over time. Crichton's Store is located in Brunswick County in southern Virginia near the North Carolina border. The data for the years 1854 to 1857 was used to look at the variations in the variables. The recordings in this source are not complete, especially for temperature readings. Periodically for 1854 and 1855, temperature was only recorded twice instead of three times daily while a few days were even blank. These records became more consistent over time.
 First, the monthly average temperature for July through September in both Crichton's Store and Fort Monroe were examined to see if the temperature in 1855 was significantly higher compared to other years (Table 2). There were not an equal number of cases for all years with both sources (Figure 4), so the Fort Monroe data was examined separately in a box plot to display the central tendency, outliers, and extreme cases of each individual variable. It illustrated that there was an almost cyclical change in temperature over eleven years with 1854 and 1855 sharing the highest arc (Figure 5).
The Fort Monroe data was divided into subgroups (1850-1853, 1854-1855, 1856-1860), based on the exploratory data analysis (EDA) results indicating the mean average temperatures clustered in three groups. These sub-groups are shown in Figure 7 below.
 Based on this EDA, a one-way ANOVA test and a one-tailed, one sample t-test were run. ANOVA compares the means for three or more samples to determine if at least one is different from the population, whereas a one-tailed, one sample t-test compares one sample mean to the population mean. The hypotheses were the following:
Null hypothesis: The monthly average temperatures for all three samples come from the same population and are essentially equal.
Alternate hypothesis: At least one of the samples is from a different population.
One-tailed, One Sample T-test
Null hypothesis: The mean average temperature of the 1854-5 sample is equal to the population mean.
Alternate hypothesis: The sample mean is not equal to the population mean.
A one-way analysis of variance indicated that the mean monthly average temperature differed significantly between the populations, or subgroups (p = 0.029, F = 3.99, m = 12 ri2 = 6, n3 = 30). Post hoc tests including the Scheffe and Bonferroni tests (Norusis 1999) indicated that there is a statistically significant difference between 1850-3 and 1854-5.
Based on the results of ANOVA, a t-test was run to see if 1854-5 was significantly different from the surrounding years. A one-tailed, one sample t-test indicated that there is a statistically significant difference between the mean monthly average temperature for the 1854-5 sample and the mean monthly average temperature for the population (t = 2.74, df= 5, p = 0.021, n = 6). Based on Drennan's (1996:163) scalar approach to significance testing, it is "fairly unlikely" that the null hypothesis is true, or that there is no difference between the 1854-5 mean and the population mean of average temperatures for the entire eleven year interval. Thus, the conclusion is that the monthly average temperature for 1854-5 was significantly higher than the surrounding years.
 Next, the change in diurnal range over time was evaluated by examining the Fort Monroe data, as it provided information on this variable for a relatively long period of time (Table 3). The diurnal range for 1855 was expected to be less than the ranges of the surrounding years because it would be more conducive conditions for Aedes aegypti.
Exploratory data analysis indicated a long-term pattern in which diurnal range decreased over time (Figure 8). A scatter plot (Figure 9) suggested that there was a negative linear relationship between the two variables. The Pearson correlation coefficient is a measure of linear association between two numerical variables, such as the year and diurnal range values considered here. For the diurnal range data, Pearson's r = -0.622 (n = 22), indicating that there is indeed a negative relationship between year and diurnal range. The r-squared value of 0.387 indicates that change in year accounts for approximately 39% of the variation. The statistical significance of this correlation, i.e. the probability of obtaining these results if chance alone were operating, may also be computed.
Two-tailed Bivariate Correlation Test
Null hypothesis: There is no linear relationship between diurnal range and time.
Alternate hypothesis: There is a linear relationship between the two variables.
 A two-tailed correlation bivariate test indicated that the there is indeed a negative linear relationship between diurnal range and time (p < 0.001, Pearson's r = -0.622, n = 33). It is extremely unlikely that the null hypothesis of no linear relationship between diurnal range and time is true. This decrease in diurnal range over time is unusual for it should remain relatively constant and implies that nighttime temperatures are higher than normal.
Based on the results of the correlation, a t-test was run to see if the diurnal range of 1855 was significantly narrower than the surrounding years.
One-tailed, One Sample T-test
Null hypothesis: The mean diurnal range for the year 1855 is equal to the population mean.
Alternate hypothesis: The sample mean is not equal to the population mean.
A one-tailed, one sample t-test indicated that there is no statistically significant difference between the diurnal range of 1855 and that of the surrounding years (t = -0.371, df = 2, p = 0.374, n = 3). As one would expect in looking at the bar graph, the null hypothesis of no difference between the mean diurnal range for 1855 and the population mean is true. Therefore, the conclusion is that the diurnal range of 1855 was not significantly narrower than the surrounding years.
The next factor examined was monthly precipitation for 1854 to 1857 using both sources (Table 4). After performing some exploratory data analysis and removing outliers, 1854 was observed to have the least amount of precipitation (Figures 10 and 11). After an abrupt rise in rainfall during 1855, precipitation decreased during the following years.
A scatter plot indicated that there was no correlation between the amount of precipitation and time (r-squared = 0.018). The amount of precipitation for the entire year in the Crichton's Store data was examined by subgroups (1854-5, 1856-7) to increase the number of cases. There was a higher mean amount of precipitation in 1854-5 (Figure 12).
The EDA indicated that the data should be evaluated using a one-tailed independent samples t-test. A one-tailed independent samples t-test compares two sample means to the population. The statistical hypotheses are as follows:
One-tailed Independent Samples T-test
Null hypothesis: The mean precipitation for the two samples is equal. Alternate hypothesis: The sample means are different, and therefore from different populations.
 This test indicated there is a statistically significant difference between the mean amount of precipitation for 1854-5 and 1856-7 (t = 1.60, df = 43, p = 0.06, n = 24). It is fairly unlikely that the null hypothesis of no difference between the 1854-5 sub-group and the 1856-7 sub-group is true. Thus, the conclusion that can be drawn is that there was a greater amount of precipitation in 1854 and 1855 compared to the surrounding years.
For the analysis of the number of precipitation days (Table 5), the data was examined first in both sources and then for Crichton's Store alone (Figures 13 and 14).
After performing some exploratory data analysis and removing the outliers, both box plots indicated that the mean number of precipitation days was similar for most of the years. From the exploratory data analysis, it was observed that the hypothesis of significantly more precipitation days in 1855 was not true and it was unnecessary to produce any inferential statistics. Therefore, the conclusion is that the year 1855 did not have more precipitation days than the surrounding years.
For the analysis of frost days (see Table 6), Crichton's Store data was examined separately because Fort Monroe only provided data for summer months. Both the box plot in Figure 15 and the bar graph in Figure 16 indicated that 1855 had approximately
 the same number of frost days as the surrounding years except 1856, which had the lowest amount.
Seeing that my hypothesis of significantly less frost days in 1855 was not true, it was unnecessary to generate any inferential statistics. Therefore, the conclusion is that the year 1855 did not have a lower number of frost days than the surrounding years.
Based on the results from the exploratory data analysis and inferential statistics, it seems that the weather in 1855 was conducive to the mosquito, Aedes aegypti. The weather was warmer in the summer months of both 1854 and 1855 with a mean temperature near 79° F. There was also less of a cooling trend in the summer of 1855, with higher nighttime lows, because it had a fairly narrow diurnal range compared to previous years. Potentially, the higher temperatures increased the activity of mosquitoes. There was also a greater amount of precipitation, which increased the humidity in the area and the number of breeding places for A. aegypti. These weather conditions aided A. aegypti by increasing the frequency of its activities and providing more breeding places for the insect. Thus, it can be concluded that with the ideal weather conditions for mosquito activity and breeding, conditions were conducive to an increase in both the A. aegypti population and the cases of the yellow fever.
 Chapter Four: Relevance and Conclusions
This paper has examined the 1855 yellow fever epidemic of Norfolk and Portsmouth as a window into the past and the future through an anthropological analysis of the community response to the epidemic and the role of climate. In Chapter Two, two problems were introduced. Studying social responses to epidemics allows medical anthropologists to learn about a particular society's perspectives of disease and its relationship with the environment. Through the analysis of climate and aspects of the environmental context, the influence of weather conditions on the spread of infectious diseases was evaluated. In Chapter Three, the relevant evidence was presented. Four stages of social response were identified in the 1855 epidemic. The climate during the outbreak was compared to surrounding years using statistical analysis, which determined it was conducive the yellow fever vector, Aedes aegypti.
Finally, in this chapter the patterns in evidence are evaluated with respect to the study of disease and human behavior in an ecological setting. The first section compares the social stages of response in Chapter Three to ones defined by medical anthropologists. The second section evaluates the climate research in respect to current issues. The final section merges the complementary results and discusses the relevance of this topic in today's world and emphasizes the environmental and health risks created by culture.
 Social Response to Disease
Chapter Three presented a four-stage model of social response in the 1855 yellow fever epidemic, compared to other medical anthropological studies which have identified up to seven stages. Many of the stages that medical anthropologists identify are very specific, whereas the four stages in this study are defined through the evaluation of seven attributes that describe the spread of the disease, number of cases and deaths, social perspectives, and impacts on the community. For instance, Humphreys' (1992) seven stages can be grouped to form four, more broadly applicable stages. Rosenberg (1989) identifies four stages but not one describes the social response to the climax of the epidemic. Both Doka's (1997) and Smelser's (1963) stages are very general and some occur throughout the epidemic. For example, Doka's stages of transcendence and of art and literature do not necessarily occur at the end of an outbreak but may be present throughout it.
Identifying overly specific stages of social response to epidemics may be unrealistic. As medical anthropologists have learned, different cultures can respond uniquely to disease based on their perceptions of it, which may pose a challenge in universally categorizing all aspects of human behavior. Cultures have various expectations, understandings, and knowledge of disease that influence how people respond to epidemics. Science is another part of the culture that influences social response to disease. For instance, many contemporary societies believe that medical science can control disease and that Western society will not experience infectious epidemics (Doka 1997:52). People also rely on past personal experiences during firsthand encounters with disease. For these reasons, defining over-specific stages of  social response to disease may be misleading or overlap may occur. For example, the four stages identified in this study incorporate all of Humphreys' (1992) seven stages. These four stages are most pragmatic because they are generalized and widely applicable, defined by consistent attributes with varying degrees of expression at each stage. However, even within this case of the 1855 yellow fever epidemic, some overlap occurs between Stages Two and Three, with the impact the epidemic had on community services and affairs in Norfolk and Portsmouth. The four-stage model may vary from place to place and over time, given a variety of socio-cultural factors which include society's knowledge of the disease and faith in medical science. For instance, there will be a difference in degree and duration of each stage depending on the particular disease and culture. Further studies should test this model against other cases to refine its application.
This case study can benefit similar, future studies by allowing medical anthropologists to anticipate social stages of response to an epidemic. In Stage One of an epidemic, anthropologists may expect people to act irrationally as a result of the government's failure to recognize the epidemic and implement necessary preventative measures, which allows the disease to spread. Citizens in Stage Two become more panicked and attempt to escape the disease through flight. Gradually, there is a loss of community cohesiveness as people turn against one another through prejudice and blame, and by fleeing from the situation. Stage Three is marked by a severe impact on services caused by the drastic increase in cases and deaths from unstoppable spread of disease. People begin to take desperate measures and lose hope. The final stage involves complete social disruption and community attempts to regain control and recover from  the epidemic. At this stage, people realize the limits of medical science, instigating new research.
By examining the social response of the citizens of Norfolk and Portsmouth, it may be possible to take effective preventative measures to especially avoid Stages Three and Four, when there is a total breakdown of the community with loss of control of the epidemic. During these stages, fear shapes society's response to disease which affects medical treatment and policies. Only when this fear is no longer present can the disease be viewed and treated more rationally (Doka 1997:57). Learning about social response to disease enables medical anthropologists to interpret human behavior with respect to society's perceptions, faith in medical science, knowledge, and culture, and to explain why people behave as they do. With this evidence, they can effectively contribute to improving control programs and health care, with a uniquely anthropological perspective.
Climate's Role in Epidemics
Diaz and McCabe (1999) recognized a possible connection between the 1878 yellow fever epidemic in Memphis, Tennessee and El Niño/Southern Oscillation (ENSO) in their study. ENSO causes changes in atmospheric pressure and winds which alter the course of storms and their properties, mainly in the Pacific-North region, subtropical western Atlantic, and South America. During El Niño, rainfall increases along the equator in response to the warming of underlying sea surface temperatures. However, along the periphery of this wet zone, reduced rainfall occurs. The opposite effect is experienced during La Niña (Climate Diagnostics Center 2000). Diaz and McCabe (1999) believed that as a consequence of one of the strongest El Niño episodes,  exceptional climate anomalies occurred in 1877-8 which were partly responsible for the severity of the yellow fever outbreak in Memphis (Diaz and McCabe 1999:21). For example, they saw the irregularly high temperatures and amounts of precipitation as contributing to the increase in the mosquito population (Diaz and McCabe 1999:22). However, this research indicates that weather conditions of 1855 in Norfolk and Portsmouth, which was not an El Niño year, were as conducive to Aedes aegypti as those of 1878 in Memphis. Thus, this work demonstrates that mosquito-borne infectious disease epidemics can occur during non-El Niño years.
Diaz and McCabe (1999) analyzed the same climate factors for 1877-78 as were examined here for 1855—precipitation, temperature, and number of rain and frost days. However, they did not evaluate diurnal range, the difference between the minimum and maximum temperatures in a day, which is important because it can influence the activity level of Aedes aegypti. The findings in both studies reach similar conclusions. For example, in both cases it is shown that higher amounts of precipitation the year before yellow fever epidemics, along with consistently or unusually warm conditions throughout the spring and summer, are partially responsible for yellow fever outbreaks by promoting high mosquito populations.
While Diaz and McCabe's (1999) results are sound, ENSO years are not the only years when yellow fever can occur in severe outbreaks. This study shows that yellow fever may occur any time weather conditions are conducive to Aedes aegypti (i.e., warm temperatures, narrow diurnal range, greater amount of precipitation) once it is introduced. In effect, localized conditions after the introduction of the disease can lead to horrific conditions, too. However, weather is not the only factor involved in yellow fever  epidemics. The results of this study determined that the weather conditions were ideal in 1855, but some of the surrounding years were also warm, wet and humid. The time of year the infection arrives in the area is also important to its survival. The leading factor is whether or not there is an introduction of yellow fever. Further studies should examine these factors.
There are some limitations in this study of the yellow fever epidemic of 1855 in Norfolk and Portsmouth. Only two weather data sources, Fort Monroe and Crichton's Store, were used. The completeness of these sources was not always consistent, usually for temperature readings. For example, in the Crichton's Store record, periodically for 1854-1855, temperature was only recorded twice instead of three times daily while a few days were even blank. These records became more consistent over time. Second, with only two sources, there were too few cases for some descriptive and inferential statistics such as clustered box plots and ANOVA, so combining the data into subgroups, rather than examining each separately, became necessary. This may have biased some of the statistical results. In future studies, more sources should be consulted which include both daily data and frost dates.
This paper provides statistical evidence that warm-wet conditions increase opportunities for outbreaks of infectious disease. It also explains the stages of social response that may develop as a result of incidents of disease, especially ones which are not well understood like the yellow fever epidemic of 1855.
 In today's world, the prospect of global warming is no longer speculative. The 1990s was the warmest decade on record (Shute 2001:44). The rise in temperature was most likely caused by the burning of fossil fuels that release carbon dioxide as well as other gases that create a "greenhouse effect," trapping heat in the atmosphere (Shute 2001:44). The United States is the single largest generator of greenhouse gases, contributing to one-quarter of the global total (Shute 2001:50). The UN-sponsored Intergovernmental Panel on Climate Change (IPCC) used data from satellites, weather balloons, ships at sea, weather stations, and computer models of the global climate system to predict that by 2100, temperatures will rise by 2.5° F to 10.4° F (more than 50% higher than predictions made half a decade ago) worldwide, which may cause unpredictable and violent weather as the warming accelerates the water cycle (Epstein 2000:50, Lemonick 2001:26, Shute 2001:47).
Most scientists forecast that mosquito-borne diseases, such as malaria and dengue fever, will become increasingly prevalent because their vectors are sensitive to meteorological conditions. Warmer temperatures increase the chance that the parasites will mature in time for the mosquitoes to transfer the infection. Increased climate variability is more important than rising temperatures in the spread of outbreaks of some vector-borne diseases because warm winters followed by hot, dry summers are favorable to the transmission of infectious disease (Epstein 2000:53). The developing world will be hit hardest and will be the least capable of coping with such effects of global warming owing to lack of resources for prevention and treatment (Balbus and Wilson 2000:37, Epstein 2000:50, Shute 2001:50). However, imported cases of infectious disease may be a significant threat to technologically advanced nations like the United States through  international commerce and travel as climate change increases incidence abroad (Balbus and Wilson 2000:2, Epstein 2000:50).
Global warming will allow mosquitoes to expand their range in both developing and developed countries (Balbus and Wilson 2000:20, Epstein 2000:52, Shute 2001:50). For example, since 1970, the elevation at which temperatures are always below freezing has ascended almost 500 feet in the tropics, and mosquitoes and mosquito-borne disease are following (Epstein 2000:53). Another example of regional spread of infectious disease is the West Nile virus outbreak that occurred in the Northeastern region of the United States. In the summer of 1999, New York faced an outbreak because drought conditions created stagnant water and ideal conditions for its vector, increasing transmission of the virus (Balbus and Wilson 2000:20). The alteration in insect migration patterns can serve as an indicator of climate change and disease. However, like any area of science, global warming and its effects have their uncertainties. There are also multiple determinants of vector-borne disease risk and transmission that make estimating future patterns difficult (Balbus and Wilson 2000:iv). For instance, some studies indicate that global warming probably may not cause a significant spread of malaria and other vector-borne diseases to currently unaffected parts of the world (Cuplee and Weiss 2000).
The possibility of rapidly changing climate is making countries consider how to become more practical and turn away from fossil fuels (Churchill 2000:56). With an increase in international economic and political links, countries share the burden of the effects of global warming (Balbus and Wilson 2000:2). In 1997, 170 countries, organizations, and institutions signed the Kyoto Protocol, an agreement to fight the battle against global warming. The Kyoto Protocol called for a worldwide reduction of  emissions of carbon-based gases. Recently, however, President George W. Bush abandoned the treaty, claiming that its mandatory pollution reduction would be too harmful to the American economy (Lemonick 2001:22, Yahoo News 3/29/2001). Unfortunately, many people, like Bush and other pro-business lobbyists, are still skeptical about global warming because they claim not to believe the scientific evidence and feel that the problem needs to be studied further (Lemonick 2001:27). People are also apathetic because they do not see global warming and its effects as a pressing issue.
However, most believe that there is a greater chance of global warming occurring and understand that preventative measures need to be taken now. One solution is to improve surveillance systems that would promptly recognize infectious disease or the vectors that carry them (Epstein 2000:56). The magnitude of the effects of epidemics depends on education, the ability to anticipate them, and social response that may decrease impact of disease (Balbus and Wilson 2000:2). A second one is to predict when climatological and other environmental conditions could become more conducive to outbreaks (Epstein 2000:57). Another suggestion is to address global warming itself by changing human behavior to conserve energy and use less fossil fuels (Shute 2001:50).
This study contributes to understanding how human behavior affects the environment and how climate may affect incidences of infectious disease. It also provides insight in understanding social response to epidemics in a four-stage model. In contemporary society, with the prospect of global warming, the incidence of infectious diseases like yellow fever and other mosquito-borne diseases are forecast to increase, especially threatening developing countries that do not always have the latest medical technology or receive assistance from other governments. However, global warming and  disease will also potentially affect even developed and technologically advanced nations, like the United States. This paper contributes to the awareness of the need for future preventive methods and education to prevent devastating effects on these peoples.
Further research can contribute to the improvement and expansion of this work. Some areas of expansion include using more primary sources to document social behavior and climate of the nineteenth century during the yellow fever epidemic of 1855. Further studies should also evaluate other infectious disease epidemics to test the applicability of the four-stage model in this study.
Armstrong, G. D.
1856 History of the Ravages of the Yellow Fever in Norfolk. J. B. Lippincott and Co., Philadelphia.
Balbus, J. M. and M. L. Wilson
2000 Human Health & Global Climate Change: a Review of Potential Impacts in the United States. December.
Burrows, J. L.
1855 Death Arbitrary and Funeral Sermon. Manuscript on file, Special Collections, College of William and Mary, Williamsburg, Virginia.
Burton, H. W.
1877 History of Norfolk, Virginia. Norfolk Virginian Job Print, Norfolk.
Causes and Prevention of Epidemics
1857 Harper's New Monthly Magazine. 15: 194-203.
2000 The Big Meltdown. Time. 4 September: 52-56.
Christophers, S. R.
1960 Aedes Aegypti (L.), Yellow Fever Mosquito: Its Life History, Bionomics, and Structure. Cambridge University Press, London.
Climate Diagnostics Center
Cooper, D. B. and K. Kiple.
1993 Yellow Fever. The Cambridge World History of Human Disease. Cambridge University Press, Cambridge.
Cummins, G. D.
1855 The Pestilence—God's Messenger and Teacher. G. S. Gideon, Printer, Washington.
De Salle, R. (editor)
1999 Epidemic! World of Infectious Disease. The New Press, New York.
Diaz, H. F. and G. McCabe.
1999 A Possible Connection between the 1878 Yellow Fever Epidemic in the Southern United States and the 1877-78 El Nino Episode. Bulletin of the American Meteorological Society 80(1):21-27.
 Doka, K. J.
1997 AIDS, Fear, and Society: Challenging the Dreaded Disease. Taylor and Francis, Washington, D.C.
Drennan, R. D.
1996 Statistics for Archaeologists: A Commonsense Approach. Plenum Press, New York.
Epstein, P. R.
2000 Is Global Warming Harmful to Health? Scientific American. August: 50-57.
Forrest, W. S.
1856 The Great Pestilence in Virginia: Being an Historical Account of the Origin, General Character, and Ravages of the Yellow Fever in Norfolk and Portsmouth in 1855. Derby and Jackson, New York.
Goldfield, D. R.
1973 Disease and Urban Image: Yellow Fever in Norfolk 1855. Virginia Calvalcade. Autumn: 34-41.
Handy, I. K.
1855 The Terrible Doings of God. Sermon. Special Collections, College of William and Mary, Williamsburg, Virginia.
Henry, E. F., J. Chudoba, and H. C. Porter.
1959 Soil Survey of Norfolk County, Virginia. 1953(5): 1-2.
Hinton, S. W.
1855-1862 Diary. Special Collections, College of William and Mary, Williamsburg, Virginia.
Howard Association of Norfolk, Virginia.
1857 Report of the Howard Association of Norfolk, Va., to All Contributors who Gave Their Valuable Aid in Behalf of the Sufferers of Epidemic Yellow Fever During the Summer of 1855. Inquirer Printing Office, Philadelphia.
1992 Yellow Fever and the South. Rutgers University Press, New Brunswick.
Inhorn, M. and P. Brown (editors)
1997 The Anthropology of Infectious Disease: International Health Perspectives. Gordon and Breach Publishers.
Johnson, T. M. and C. F. Sargent (editors)
1990 Medical Anthropology: Contemporary Theory and Method. Praeger, New York.
 Knutson, A. L.
1965 The Individual, Society, and Health Behavior. Russell Sage Foundation, New York.
Lemonick, M. D.
2001 Life in the Greenhouse. Time. 9 April: 24-29.
Mascie-Taylor, C. N. (editor)
1993 The Anthropology of Disease. Oxford University Press, Oxford.
Mattingly, P. F.
1969 Biology of Mosquito-Borne Disease. Willmer Brothers Limited, Great Britain.
McDaid, J. D. (editor)
n.d. "Sadness to our Circle": Grace Whittle's Account of the 1855 Norfolk Yellow Fever Epidemic. Library of Virginia, Richmond, Virginia.
1998 Webster's Revised Unabridged Dictionary. MICRA, Inc., Plainfield, New Jersey.
Norusis, M. J.
1999 SPSS 9.0 Guide to Data Analysis. Prentice Hall, New Jersey.
Oldstone, M. A.
1998 Viruses, Plagues, and History. Oxford University Press, Oxford.
Parramore, T. C, P. C. Stewart, and T. L. Bogger
1994 Norfolk: The First Four Centuries. University Press of Virginia, Charlottesville.
1886 Sketch Book of Portsmouth. Library of Congress, Washington, D.C.
Rogers, D. J. and S. Randolph.
2000 The Global Spread of Malaria in a Future, Warmer World. Science. 289: 1763-1766.
2001 The Weather Turns Wild: Global Warming Could Cause Droughts, Disease, and Political Upheaval. U.S. News and World Report. 5 February: 44-52.
1853-1861 Crichton's Store Meteorological Record. Unpublished Meteorological Registers of the Smithsonian Institution, Washington, D. C.
 Storrs, R. S.
1855 Terrors of the Pestilence. Sermon. Virginia Historical Society, Richmond, Virginia.
Suplee, C. and R. Weiss.
2000 Global Warming and Malaria. Washington Post. 11 September.
Trask, B. H.
1996 Yellow Fever and Its Effects on Southern American Ports, 1859 to 1905. Mariners' Museum. 2: 44-57.
Tucker, G. H.
1972 Norfolk Highlights, 1554-1881. The Norfolk Historical Society, Norfolk.
United States Army
1824-1892 Fort Monroe Meteorological Record. Unpublished Record of the United States Army Surgeon-General. National Archives, Washington, D. C.
Wertenbaker, T. J.
1931 Norfolk: Historic Southern Port. Duke University Press, Durham.
1996 Yellow Fever, Black Goddess. Addison-Wesley Publishing Co., Inc., New York.
http://dailynews.yahoo.com/h/ap/20010329/pl/bush environment 1 -html 3/29/01
June 6 Steamer Benjamin Franklin arrived in distress from St. Thomas and quarantined June 7 Dr. Gordon, health officer of the port, inspected the Benjamin Franklin June 18 Dr. G visited vessel again and the Board of Health gave the ship permission to go to Page and Allen's Yard at Gosport June 19 Benjamin Franklin released from quarantine June 21 Dr. Upshur treats Palmer, crewman from Benjamin Franklin, for yellow fever June 22 Palmer dies June 24 First local case of yellow fever—Mrs. Fox at Scott's Creek near Portsmouth (Parramore et al. 1994) June 30 First case of yellow fever? (Armstrong) July 3 Carter from Richmond taken sick with yellow fever July 8 Carter dies, causing panic in Portsmouth; Benjamin Franklin sent back to quarantine July 16 Dr. Upshur treated first case of yellow fever in Norfolk but unwilling to admit it July 23 Seemed like epidemic would be confined to Gosport and Irish Row communities in Portsmouth July 25 Clerk in Gosport got yellow fever; cases appeared in Barry's Row July 26 Norfolk Board of Health quarantined vessels July 29 Norfolk Board of Health conceded existence of yellow fever epidemic after Dr. Upshur's report and vacated Barry's Row; 24 foot high board wall erected July 30 Fever in Barry's Row and spread rapidly; New York placed strict quarantine regulations August 1 Portsmouth in panic  August 3 2000-3000 abandoned Portsmouth August 7 Portsmouth deserted—3/4 population fled August 9 Barry's Row burned down; several cases on Main Street outside Row August 10 Several cases on Queen Street August 13 Norfolk deserted—1/2 population gone; first volunteers from abroad arrive August 14 Day set aside for humiliation and prayer August 31 Armstrong first noticed the "plague-fly" September 1 Climax—chill northeasterly storm September 3/4/5 Armstrong noticed greatest number of the "plague-fly" September 6 Martial law declared private vehicles might be seized for use by doctors and undertakers September 13 Armstrong noticed the "plague-fly" has almost disappeared; city government all but closed October 15 Refugees advised not to return until the first frost October 26 First frost—yellow fever epidemic ended October 29 Volunteers returned home and businesses reopened November 10 Last yellow fever death reported
(Armstrong 1856, Howard Association 1857, Parramore et al. 1994, Tucker 1972)