As is common in many other places, the Spanish Flu visited Newfoundland in several waves. Figure 2 shows the epidemic curve for influenza cases and for influenza plus pneumonia occurring between 1918 and 1920. This figure indicates a somewhat ambiguous series of waves. As recognized in other locations worldwide, there is clearly a small "first wave" during the summer of 1918. This is followed by a much larger "second wave" during the fall of 1918. There is a decline in deaths during December 1918, followed by a moderate increase in January before a gradual decline that lasted through June 1919.

It is possible that this pattern signifies the end of Wave II in December of 1918 followed by the arrival of a "third wave" in January that corresponds to Crosby’s Wave III (Crosby 1989); however, the wave Crosby observed in US data occurred in March and April, significantly later than the second Newfoundland Wave II peak. The onsets of both Wave I and Wave II on the island were significantly delayed relative to the United States and Western Europe (recall from our discussion above that in these regions Wave I spread in April and May 1918 and Wave II in September and October 1918), probably in large part because of the relatively remote location of Newfoundland. Thus, it is more likely that the entire period from September through June represents Wave II, and that this wave exhibited a bimodal and elongated pattern in Newfoundland. Specific reasons for the bimodal distribution of Newfoundland’s Wave II may become apparent when our present aggregate level analysis is supplemented by analyses at the sub-district and community levels.

Initial analyses of within-district epidemics indicate that in some districts the flu returned in April, May, and June 1919 after disappearing for a few months. This later outbreak is barely noticeable in Figure 2, however, and was much smaller than either Wave I or the 1920 outbreak; nonetheless, it may indicate some presence on the island of the "third wave" noted by Crosby (1989). Finally, the last (1920) wave apparent in Figure 2 clearly corresponds to an "echo" epidemic that also hit other parts of the world in 1920.

The predominance of Wave II is also reflected in our interview data. Although there were some areas where residents had extensive knowledge about Wave II (see below), we did not interview anyone who was aware that there had been more than one wave.

Figure 2 also clearly shows a strong relationship between influenza and pneumonia deaths. This problem was well known during the epidemic itself as indicated by the US Census Department’s annual report on mortality statistics for 1918:

it is not believed to be best to study separately influenza and the various forms of pneumonia, bronchitis, and the respiratory diseases, for doubtless many cases were returned as influenza when the deaths were caused by pneumonia, and vice versa. The best method therefore seems to be to study as one group deaths from influenza and pneumonia (all forms), disregarding deaths from the other respiratory diseases, which were comparatively few. (U.S. Bureau of the Census 1920, 27)

As is consistent with even modern CDC and WHO practices, because of the confounding of influenza and pneumonia, in most of the analyses presented below we combine deaths from these and related causes. It should be noted, however, that although this is standard practice, combining deaths recorded as either pneumonia or influenza can introduce other problems, because pneumonia is always a relatively common cause of death during the winter, and, in almost all cases, it is impossible to determine from written death records whether a person who died from pneumonia had also been infected with influenza or whether the death was due to pneumonia alone.

As in most regions of the world, Wave I in Newfoundland was fairly mild. As Figure 2 shows, it is difficult to determine exactly when the epidemic began since there are scattered flu and pneumonia deaths throughout the early part of 1918. The onset of Wave I was somewhat arbitrarily set at 1 May 1918, largely because there is a short period before this during which flu deaths were absent and they were dispersed in time throughout the first four months of 1918. The end of Wave I was set at 31 July 1918 because there is a break in recorded influenza deaths from that date until 10 September 1918. During Wave I, a total of 74 influenza deaths and 117 pneumonia deaths were recorded. Influenza deaths were recorded in all districts except Carbonear, Fogo, Harbour Grace, and Harbour Main. Pneumonia deaths were recorded in all districts, although Carbonear, Ferryland, Fogo, and Harbour Main experienced fewer than five deaths from influenza and pneumonia combined. The Trinity, Burin, and Bonavista districts recorded the highest number of deaths from influenza, but after standardization for differences in population size, Burin, Fortune Bay, and Trinity were highest. When pneumonia deaths were considered in addition to influenza deaths, St. John’s, Trinity, and Bonavista had the highest numbers of deaths, but Burgeo/La Poile, Fortune Bay, and Burin had the highest standardized rates of death (see Table 1).

It is impossible to know exactly when and where Wave I entered the island. The low rates of mortality during this wave meant that it may have been causing illnesses in various parts of the island well before the first death was recorded. Further, the locations of the first deaths are widely distributed across the island. Even if only flu deaths are examined, the first seven deaths after 2 May 1918 occur in four different districts — Ferryland (5/3), Burin (5/13, 5/13, 5/20, 5/23), St. John’s (5/14), and Placentia/St. Mary’s (5/21). When pneumonia deaths are included, the epidemic is even more widely distributed; during the same period seventeen deaths from pneumonia occurred in nine districts (seven of which did not record influenza deaths) — St. George (5/4, 5/6, 5/21), St. John’s (5/9, 5/11, 5/14, 5/20), Trinity (5/13, 5/16, 5/17), Bay de Verde (5/18, 5/22), Harbour Grace (5/19), Burgeo/La Poile (5/21), Carbonear (5/21), Placentia/St. Mary’s (5/23), and Twillingate (5/23). In addition, most of the remaining districts recorded their first influenza and/or pneumonia deaths only a short time later. The primary exception to this is the GNP — it is clear from both influenza alone and the combined data, that Wave I did not reach this area (a sub-district of St. Barbe) until either June or July 1918, significantly later than in the rest of the island.

According to an article in the 29 May 1918 edition of the Evening Advocate,a daily newspaper published in St. John’s, in late May 1918 a substantial outbreak occurred on the western coast of the island with "a great many people ... suffering from this ailment and several deaths" (4). The article also reported that "Hundreds of people here in the city [St. John’s] are at present laid up with La Grippe and in one firm 16 out of 20 employees are indisposed." On 30 May 1918, the Western Star, a weekly paper published in Birchy Cove, near Corner Brook on the western side of the island, noted a large outbreak of influenza with much illness but little death that affected both the coastal steamers Kyle and Fogota and the towns of Channel and Port-aux-Basques.

These articles confirm that the epidemic, like all influenza outbreaks, was much larger than the death records would indicate. This is also evident from the hospital records in St. Anthony. A diagnosis of influenza or related illnesses (e.g., influenzal pneumonia) was given for fourteen patients and one other had a note to indicate that he had experienced the illness during the summer of 1918 (possibly in Labrador) before he entered the hospital for other reasons. Of these fifteen patients, only four died (although one of these four was not recorded in the provincial death records and two were listed as dying from other causes, specifically meningitis and pneumonia). These hospital records only represent those individuals who were ill enough to require hospitalization — the total number of individuals sickened during the outbreak was undoubtedly substantially higher. Nonetheless, it is clear from Figure 2 that the impact of Wave I was miniscule compared to that of Wave II, to which we now turn.

For the sake of simplicity, we will usually refer to the period from 10 September 1918 through 5 June 1919 as Wave II, although this period may have contained both a second and third wave in some districts. The date of 10 September 1918 was chosen for the start of Wave II because it marks the first influenza death following the end of Wave I (several pneumonia deaths were recorded in the interim between Wave I and Wave II, however). The end of Wave II was chosen to coincide with the last official government report of influenza cases and deaths during the epidemic.[1] A total of 901 influenza deaths and 278 pneumonia deaths were recorded in the provincial death records during this period, representing 80 percent of all influenza deaths recorded during the three-year period from 1918 to 1920 and 62 percent of influenza and pneumonia deaths combined. The drop in percentage of pneumonia deaths when pneumonia and influenza are combined may simply be the result of a constant rate of pneumonia deaths during the times when more flu deaths are occurring. It is possible, however, that this pattern was partially a result of how causes of death were assigned — at the height of the epidemic when health authorities all over the island were well aware that an epidemic of flu was occurring, the cause of death from any respiratory illness may have been more likely to be attributed to influenza.

Nearly every district on the island experienced a major outbreak during Wave II. Only two districts, Burgeo/La Poile and Ferryland, recorded influenza death rates under 10 per 10,000 population; only Burgeo/La Poile recorded a death rate under 20 per 10,000 population for combined deaths from influenza and pneumonia (Table 2). The highest Wave II death rates for both influenza alone (> 80 per 10,000 population) and for influenza and pneumonia combined (> 90 per 10,000 population) were recorded in the districts of Placentia/St. Mary’s, St. George, and Fortune Bay.

A distinct clustering of deaths was observed between the middle of October and the end of January in all districts except Burgeo/La Poile, Ferryland, and Harbour Main. Deaths occurring in Burgeo/La Poile and Ferryland were too limited in number to generate significant clusters; deaths occurring in Harbour Main were scattered throughout the extent of Wave II and did not produce well-defined clusters. Most of the other districts recorded distinct clusters in October and November, with some experiencing second, usually smaller, clusters in December and/or January. Port de Grave was unusual in that the October/November cluster of deaths did not occur there — the only observed cluster occurred in the second half of January. Almost all districts continued to record sporadic deaths from influenza and/or pneumonia between February and June.

Determining when and where Wave II began in Newfoundland is a complicated task. Because there was a break in recorded influenza deaths between 31 July and 10 September 1918, it is tempting to assume that the epidemic disappeared from the island at the end of Wave I and re-entered to begin Wave II. It is important to keep in mind, however, that this midsummer break occurs only in the influenza deaths — pneumonia deaths are recorded throughout the interim period (see Figure 2). Given the confounding of pneumonia and influenza noted by the US Bureau of the Census (1920, 27) during this epidemic, it is not certain that the Spanish Flu disappeared from Newfoundland at the end of Wave I.

Nonetheless, in Newfoundland as well as the rest of the world, Wave II resulted in a much higher rate of severe complications and mortality than Wave I, leading Crosby (1989) to hypothesize that a mutant form of the virus arose in the interim between Wave I and Wave II. Consequently, attempts have been made to trace the Wave II "introduction" of the epidemic.

Several possible clues to how the second wave may have entered the island are found in an article published in the Evening Advocate on 30 September 1918:

A steamer is anchored on quarantine with the yellow flag at her foremast. The censorship regulations [related to the war] prevent us from saying anything as to the port. Three of the crew were removed to hospital this a.m. and Dr. Campbell, the port physician, is enforcing the strictest regulations. In Boston and other American cities this deadly disease is now raging and many deaths have occurred. A schooner which arrived here some time ago also had several of the crew sick with what is believed to be the same sickness. (4)

Parsons (1992) states that "the next day, the Daily News [another local newspaper] reported that two cases from the schooner Ariceen of Twillingate were taken to hospital." The Evening Advocate of 1 October 1918 reports that the mate of the Ariceen "died early today, attacked on way from Twillingate to this port [St. John’s]" (4).

Our evidence suggests, however, that these cases were not likely to be the first introductions of Wave II to the island. Although nearly all influenza deaths recorded in St. John’s occurred in October or later, there was a flu death recorded there on 13 September and four pneumonia deaths were reported in the latter half of September, casting suspicion on the purported role of the boats and their sailors as presented in these newspaper accounts.

As is the case with Wave I, we will probably never be able to determine exactly where and when Wave II appeared on the island, especially since nonfatal cases may have been present for weeks before producing the first death. Furthermore, given the fact that most deaths from the flu were ultimately due to co-infection with pneumonia, it is also possible or even probable that at least some of the known pneumonia deaths recorded in August and early September were ultimately due to an initial influenza infection.

Official deaths recorded in the provincial death registers include a total of seven influenza deaths recorded during the month of September. Those deaths occurred in the districts of Bonavista, Burin, Fortune Bay, Harbour Grace, St. John’s, and Twillingate. All of these districts except Twillingate also recorded deaths from pneumonia in September and/or the last week of August. In addition, Bay de Verde, Placentia, and Trinity recorded pneumonia deaths during this time, although they did not report any influenza deaths. Given the uncertainties in diagnosis of flu as a cause of death, any of these nine districts could have been the location of the first Wave II case.

Where else might we be able to glean evidence related to the introduction of the second wave? The most likely mechanism by which diseases such as influenza entered Newfoundland during the early twentieth century was through interactions with sailing ships visiting the island. Consequently, the arrivals of international vessels listed in the Lloyd’s List of Overseas Shipping Intelligence have been examined to see if they can help in determining the island’s most likely influenza foci early in the second wave.

The vessels arriving in Newfoundland in the month between 23 August and 22 September include the following:

• August 23 — the Cascapedia arrived in St. John’s from Halifax

• August 24 — the Rijabergen arrived in St. John’s from England

• August 24 — the Faraday arrived in St. John’s from unknown location

• August 29 — the Bluenose arrived in St. John’s from Cadiz [Spain]

• August 30 — the Evelyn arrived in St. John’s from Cadiz

• August 31 — the Terra Nova arrived in St. John’s from Seville [Spain]

• August 31 — the Svalen arrived in St. John’s from Iceland

• August 31 — the Wellington arrived in St. John’s from Cadiz

• September 1 — the Mackay Bennett arrived in St. John’s from unknown location (probably local)

• September1—the Solburg arrived in St. John’s from Christiansund [Norway]

• September1—the Stella II arrived in St. John’s from Cadiz

• September2—the Freedom arrived in St. John’s from Cadiz

• September2—the Viola May arrived in St. Jacques from Oporto [Portugal]

• September2—the Gigantic arrived in Harbour Buffet from Gibraltar

• September2—the Elsie L. Corkum arrived in Burgeo from Oporto

• September 5 — the R. Fabricius arrived in St. John’s from the Faroe Islands [North Atlantic]

• September5—the Andres Roca arrived in St. John’s from Cadiz

• September4—the Jennie E. Ritcey arrived in St. Anthony from Gibraltar

• September 6 — the Robert J. Dale arrived in St. John’s from Pernambuco [Brazil]

• September 6 — the Carrie and Nellie arrived in Burin from unknown location

• September 11 — the Success arrived in Marystown from Oporto

• September 11 — the Nita M. Conrad arrived in Burin from Oporto

• September 13 — the Prospero arrived in St. John’s from Philadelphia

• September 13 — the Eskason arrived in St. John’s from Halifax

• September 16 — the Fritz arrived in St. John’s from Iceland

• September 16 — the Ludvig arrived in St. John’s from Iceland

• September 20 — the Hebe arrived in St. John’s from Iceland

• September 22 — the Cascapedia arrived in St. John’s from Halifax

• September 22 — the Sunset Glow arrived in St. John’s from Barbados

• September 22 — the Emily E. Selig arrived in Burin from Oporto

• September 22 — the Burleigh arrived in Burin from Oporto

As this list shows, most international ships entered Newfoundland at St. John’s, but there were a number of arrivals at other ports during early September. With one exception (the Jennie E. Ritcey that arrived in St. Anthony on the Northern Peninsula), ships arriving at ports other than St. John’s docked at towns on the southern coast — St. Jacques is in the Fortune Bay district, Harbour Buffet was at the northern end of Placentia Bay in the Placentia/St. Mary’s district, Burgeo is on the south central coast in the district of the same name, and Burin and Marystown are located on the eastern side of the Burin Peninsula in the Burin district (see Figure 1). This list suggests that if the epidemic did arrive by sea, it most likely entered the island at St. John’s, although other points of entry, especially the Burin Peninsula, are also possible.

What do we know about the occurrence of the epidemic in the ports from which these ships set sail? Unfortunately, examination of the epidemic experiences in these ports does not clarify the picture much. Only two ports can be removed from consideration — Iceland, which did not get the second wave until mid-October (Cliff et al. 1986) and Barbados, which remained unaffected by Wave II until late December (Killingray 1994). In addition, the timing of the second wave in the Faroe Islands and in Nova Scotia is not known. There is evidence, however, that Wave II reached Spain, Portugal, and possibly Brazil in late August (Pyle 1986; Patterson and Pyle 1991; Echeverria 2003), and during September it made its way to England, Norway, Philadelphia (Pyle 1986; Patterson and Pyle 1991), and Gibraltar (Sawchuk, pers. comm). Any one of the nineteen ships derived from these ports could have brought the second wave to the island during the critical time period.

International vessels were not the only vessels sailing to and from the island, however — other boats that do not appear on the Lloyd’s List might have brought the second wave to other ports, including those on the north and northeast coast of Newfoundland. These initial analyses make it clear that shipping lists, although informative, are not sufficient to pin down the epidemic’s source. Additional information to indicate whether ships’ crew members were experiencing flu-like illnesses may be available in ship’s logs and other marine archives.

What can we say about the overall patterns of spread once Wave II took a firm hold on the island? In general, as is typical of most infectious disease epidemics, the second wave started relatively slowly, but towards the end of the second week of October, the epidemic exploded and it became clear that this was not an ordinary influenza outbreak. Judging from the date when well-defined clusters of deaths began to be recorded, the earliest district-level epidemics occurred in St. John’s, St. George, Placentia/St. Mary’s, Twillingate, and Harbour Grace, which all showed evidence of death clusters in early to mid-October 1918. Influenza deaths did not cluster significantly in Bay de Verde district until the third week of October, but there were several pneumonia deaths in the second week of October; so if pneumonia and influenza are both considered, this district would join the group with early clusters. Deaths in Bonavista did not begin to cluster significantly until the third week of October, but rose to a peak almost immediately. Outbreaks in Placentia/St. Mary’s, St. John’s, and Twillingate also began to peak around the beginning of the third week of October. During the week of 15-21 October, eighteen deaths from influenza occurred in Placentia/St. Mary’s and fourteen deaths occurred in St. John’s; five and seven additional deaths from pneumonia occurred in Placentia/St. Mary’s and St. John’s, respectively. The following week the influenza deaths rose to 33 in St. John’s, with fifteen additional deaths from pneumonia. Twillingate recorded nineteen deaths from influenza and three deaths from pneumonia over the two-week period, and Bonavista recorded seven influenza deaths from 15-28 October, as well as another seven on 29 October. The St. George death cluster began around the same time as in St. John’s, Placentia/St. Mary’s, and Twillingate, but it did not peak until early November. Harbour Grace experienced only a very small cluster at this time.

Observations of the geographic locations of these districts make it clear that identifying the early foci of the epidemic will not be possible at the district level — by the middle of October Wave II was strongly affecting all regions except for long stretches of the south coast and the GNP. Future research will include analysis of the temporal and spatial distribution of deaths within districts, and it is possible that this analysis will better enable us to determine the most likely early foci for the epidemic.

On the island as a whole, the epidemic reached its primary peak around the middle of November,[2] with nearly every district experiencing significant influenza and pneumonia death rates. It was also at this time that the median date of death was recorded for most districts (Table 2), although this corresponds to a well-defined epidemic peak within only a few districts (Fogo, Fortune Bay, St. Barbe, St. George, and Twillingate). As discussed above, there is a drop in incidence in late December, with a moderate rise again in January, but by the beginning of February, the incidence of deaths from influenza settled down to a fairly low level. Several districts contributed to the rise in January, but it was mostly fuelled by delayed epidemics in Harbour Grace and Port de Grave, and moderate increases in deaths recorded in St. John’s and Trinity. Influenza deaths continued to be reported from somewhere on the island sporadically and without breaks of any significant length through mid-summer of 1919, although as discussed above, because of the ending of official reports, deaths from 5 June through mid-summer were included within the interim between Wave II and the 1920 Echo period rather than as part of Wave II itself.

Interviews revealed that, as one would expect, the amount of knowledge about the flu was related to the severity with which it hit an area. In areas where no deaths were reported, such as along the Strait of Belle Isle, individuals often reported that the only thing they knew about the Spanish Flu was what they had read about in school or seen on television, and that they had never heard any stories about it being in their communities. Even in areas where a small number of Spanish Flu deaths had occurred it was common for many people to report that they had never heard about the Spanish Flu being in their area. One resident of St. Anthony was surprised to hear that there had been an outbreak in the Grenfell Orphanage because she worked as a tourist guide for the Grenfell Mission and had never heard that the Spanish Flu had been in the community.

This level of knowledge made the responses to questions about the Spanish Flu in areas that had major outbreaks all the more striking. In these communities we met many relatives of Spanish Flu victims, and these relatives provided many details about topics such as who had died, how the sick were cared for, and how people thought the disease had entered the community. For example, individuals in Lark Harbour (near the southern entrance to the Bay of Islands) were able to tell us about individuals who had been ill with the flu before the first resident had died, as well as detailed descriptions of the difficulty of keeping up with the burials of victims during the worst part of the epidemic. In most cases, this information would be in the form of stories passed down through the generations, but in a few cases residents were able to retell first-hand accounts from their childhood. One resident in Bonne Bay who had lived through the epidemic gave us details on how his father had travelled from house to house helping to feed and care for victims. He also reported that at the time people said the epidemic had been brought to the area by a sick sailor on a schooner from Boston.

One possibility for the low impact of Wave II on the GNP and Burgeo/La Poile regions is that these regions experienced widespread outbreaks of flu during Wave I, which led to the development of immunity to influenza that protected residents from the more severe Wave II epidemic. As discussed above, the hypothesis that immunity resulting from exposure to Wave I might help explain Wave II patterns was first proposed in the immediate aftermath of the epidemic itself (Frost 1919; Vaughan 1919; Jordan 1927; Burnet and Clark 1942; Shope 1944, 1958). Our analyses so far, however, indicate that first wave immunity is unlikely to account for the observed variations in death rates at the district level, and is also less likely to be a major explanation for the small outbreaks in the GNP and in the Burgeo/La Poile district than other potential influences.

The immunity hypothesis was stimulated by our analysis of in-patient hospital records from the Charles S. Curtis Memorial Hospital in St. Anthony. Unfortunately, we have not been able to uncover comparable records from the Burgeo/La Poile district or elsewhere on the island, so our discussion of this hypothesis will consider only the situation in the GNP. Analysis of the St. Anthony hospital records confirmed the conclusion from study of the provincial death records that the outbreak of Spanish Flu in the GNP was very small. Only 26 individuals were treated in the hospital for influenza during 1918 and 1919, eight of whom died from influenza or complications of influenza such as meningitis or pneumonia (in the latter cases, influenza was mentioned in the diagnosis). One significant feature of these influenza cases is that 2/3 of the 26 recorded cases and four of the eight deaths occurred in the summer of 1918, not in the fall of 1918 when the rest of the island was hardest hit. One of the remaining deaths resulted from complications of influenza contracted in May 1919, leaving only three deaths during the fall epidemic. Because the summer cases resided in nearly a dozen different GNP communities, these results suggest that St. Anthony and other communities on the GNP may have experienced a sizable flu outbreak during Wave I, especially since hospital cases and deaths almost certainly represented only a small proportion of existing cases. We have, however, no reported cases of either wave of flu in some areas of the GNP.

It is surprising that the proportion of hospital cases occurring during Wave I was so much higher than that occurring in Wave II, especially since in most places, both in Newfoundland and worldwide, Wave II was much more severe in its impact. This lower Wave II proportion is also surprising because we know that this wave did reach the region. Not only do we have the Wave II cases and deaths indicated in the hospital records and provincial death records, but we also know that there was a very localized severe outbreak of flu at the St. Anthony Orphanage in November 1918. This outbreak was noted in the official orphanage report for 1918 and indicated that 34 of the 37 children at the Orphanage came down with the flu at that time, although only one of those children died (Spalding 1920). That death is recorded in the hospital records, but very few of the cases in children (at most one or two) appear in those records. Additional orphanage records indicate that there were no cases of flu in these children during the summer of 1918 when there were other cases recorded in the hospital records from St. Anthony.

These data merely indicate the presence of the epidemic in the region during Wave II; they do not inform us about the extent of the Wave I outbreak. It is important to note, however, that both the absence of flu in the Orphanage during Wave I and the relatively minor presence of flu in the region during Wave II suggest only limited interactions between the orphanage residents and the general public. Nonetheless, even if such interactions were limited, there were individuals working at the Orphanage who lived in the community and provided a potential conduit for flu to spread from the Orphanage to the town. That this did not happen to any great degree is one of the confusing aspects of the Spanish Flu experience in St. Anthony.

It is unfortunate that the existing data are not sufficient to determine whether the summer outbreak was of sufficient scope to generate the level of immunity needed to protect the region from the later wave. Data on overall case numbers, including individuals who did not seek medical care, are not available for any community on the GNP, and comparable hospital data are not available for other parts of the island, so we have no knowledge of potential case rates for Wave I in any other part of the island. Further, when death data are aggregated at the district level, in many areas presence of the first wave was followed by major outbreaks of the second wave. It is possible that stronger evidence for the role of immunity will be uncovered with a finer level of geographic analysis (i.e., by comparing locations of Wave I and Wave II deaths at the community level), but it is also possible that this evidence will further weaken the immunity hypothesis. A definitive answer to this question, if it can be found, must wait for future research.

Interviews throughout the island revealed that many people saw boat travel as a major means by which contagious diseases had been spread in the past. For example, in Bay Roberts several people told us of a story, whose details were vague, about three sailors who arrived at the port dying of disease and were buried somewhere in the vicinity in an unmarked grave. The association between boats and the spreading of disease is also clearly reflected in the many newspaper articles about boats being held in port and/or being placed in quarantine during the time of the Spanish Flu. Such concern is justified because the movement of ships internationally and along the coast of the island of Newfoundland almost certainly did play an important role in determining patterns of spread of the Spanish Flu epidemic.

As indicated in the list of international arrivals given above, ships coming to the island from international ports only rarely docked in the southern and western ports, and coastal boat traffic was heaviest in those areas hardest hit by the epidemic, including the northeast coast from St. John’s to the Baie Verte peninsula and the coastal regions of the Placentia/St. Mary’s and Fortune Bay districts. Both daily newspaper reports on coastal boat movements and interview data collected during the summer 2006 field season indicated that in a typical week during the early part of the twentieth century approximately half a dozen coastal boats would move between communities along the northeast coast, and about half that number travelled along the southern coast from Placentia through Fortune Bay. Rates of travel were lower elsewhere — typically only two boats per week travelled along the rest of the south coast, and only one boat regularly travelled along the west coast of the GNP. The inconsistency between the low coastal boat traffic along the three west coast bays (St. George’s Bay, Bay of Islands, and Bonne Bay) and the high number of flu deaths recorded from those areas is likely due to nearby railway lines (see below).

Because of the large role coastal boats likely played in the spread of the second wave, we collected data on all of their movements during that time period. Daily reports of coastal boat movements and other local news related to the flu were published in the Evening Advocate, and these reports have made it possible to identify several specific ways that the flu may have moved, and may have been prevented from moving.

Coastal boat travel between the island of Newfoundland and communities in Northern Labrador has been definitively shown by Budgell (1996) to have carried the Spanish Flu to Labrador. The vessel most strongly implicated in this was the Harmony, which left St. John’s on 20 October on its way to the Northern Labrador coast. Two other vessels, the Seal and the Sagona, also made similar trips and may have been responsible for the epidemic making it to other Labrador communities. The important thing about these ships for our purpose is that they made stops in Newfoundland ports while on their travels between Newfoundland and Labrador. For example, on 28 October 1918, the Evening Advocate carried this report from the Seal, which was attempting to return from Labrador to St. John’s:

October 28 [the Seal] detained on way south from Labrador at Catalina, Port de Grave, Hr. Grace, Bay Roberts, and other points landing crews brought from various ports on Labrador. Crew not allowed in any of ports. Heavy snows bring winter to Labrador. (6)

The Evening Advocate published a regular column called "Reid’s Ships and Trains" that appeared in nearly every edition. According to these reports, the Sagona made two trips back and forth to Labrador during Wave II and so could have carried the flu from St. John’s to the northeast coast of Newfoundland and to Labrador . The Sagona left St. John’s on 18 September on her way to Trinity, left Trinity on 19 September going north, left Wesleyville on 20 September, and was reported to be north of Twillingate on 23 and 24 September, before she sailed north of St. Anthony to Labrador on 25 September. On 7 October, the paper reported that the ship was on her return from Labrador to St. John’s. On 15 October the Sagona once again left St. John’s and sailed for Labrador. Ports of call along the way included Trinity, Catalina, and St. Anthony. On 30 October it was reported that the ship would be returning to St. John’s the evening of 31 October after going as far north as Hopedale, Labrador, and that she had been in St. Anthony again on 29 October on her way back to St. John’s. On 16 November the Sagona set sail again, this time for Burin. She returned to St. John’s on 19 November carrying reports of eight flu deaths in Harbour Breton and Sagona Island. There is no clear evidence that the ship carried the flu on its travels from either Labrador or Harbour Breton/Sagona Island, but it is possible.

In addition to the Harmony, Seal, and Sagona, other coastal boats that might have spread the flu along the northeast coast include the Dundee, Earl of Devon, Clyde, Susu, Petrel, and Home. Reports in the Evening Advocate during the autumn of 1918 describe how two of these six boats, the Petrel and the Home, were detained along the way because they were carrying people sick with the flu. Besides the Sagona, boats that travelled regularly along the south coast included the Fogota, Argyle, and Portia. Both the Fogota and the Argyle experienced disruptions in their service because of the flu. Illness was not reported on the Fogota, but this vessel was prevented from making its normal stop at Placentia on 25 October because of the flu, and the Argyle was held at Placentia for nearly two weeks in November due to flu in that community.

While the rest of the island was serviced by multiple vessels, the long west coast of the GNP was only regularly serviced by one coastal boat, the Ethie. Reports about the Ethie in the Evening Advocate during this period highlight the potential impact this relative lack of coastal boat traffic had on epidemic spread. Although such newspaper reports were commonplace at the time, hindsight imbues these particular reports with life or death drama. On 23 October the newspaper reported that the Ethie was about to leave Humbermouth in the influenza-ravaged Bay of Islands (St. George district) for the Strait of Belle Isle in the northern half of the St. Barbe district, and the start of this trip was confirmed the following day. However, the 25 October edition reports, "The SS Ethie has been hung up at Curling [in the Bay of Islands] the past three or four days owing to the difficulty in getting a crew because of the ‘flu’ epidemic" (8). On 29 October, the newspaper reported that the Ethie was back at Humbermouth because the influenza was so common in that area that they were unable to replace the sick crew. On 1 November the Ethie travelled from Humbermouth to Curling where it was stopped again because most of the crew were ill. Eventually the coastal boat, the Diana, was sent to replace the Ethie and on 1 November she was reported ready to travel northward to the Strait of Belle Isle on 2 November. The Ethie continued to be detained at Humbermouth until all of its crew recovered. It finally headed northward to the Strait of Belle Isle on 15 November.

If the Ethie had not been delayed in the Bay of Islands due to the sick crew, the pattern of the Spanish Flu on the island of Newfoundland might have been very different. A hint of the potential impact of Ethie’s postponed sailing is eerily noted in the Evening Advocate on 29 November. Upon the ship’s return back to Bay of Islands after finally making the trip northward, the crew is reported commenting that they feared the dire consequences of a Spanish Influenza outbreak in the Strait of Belle Isle because "In the Straits there is no medical service, and we would draw the attention of the authorities to this condition and trust an effort will be made to relieve the situation at once" (8). That such a feared epidemic never developed in the Straits may be due to the postponement of the Ethie’s voyage to the Strait of Belle Isle at a crucial time.

When asked about the possibility of the spread of the Spanish Flu by railroad travel, residents acknowledged this possibility. In contrast to boat travel, however, they only rarely brought up the idea that the train would have transported the Spanish Flu. This relative unconcern about railroad traffic in spreading the Spanish Flu is also seen in newspaper articles from the time period. While regularly presenting stories about ships being held up due to sick crew and/or being put under quarantine, the only attention given to trains was the reporting of their daily travel back and forth across the island from St. John’s to Port-aux-Basques, and along the spurs to Carbonear (west of St. John’s) and Placentia (the southern shore). This lack of concern about trains spreading the disease was even apparent when train travel was mentioned in a story about a recent outbreak of influenza. For example, in an article entitled "That Sickness at Port au Basques" the reporter states that "We also interviewed parties who came in from there by last evening’s express train, and were told that what is supposed to be a very severe type of influenza is prevalent at both Channel and Port au Basques, and many people are down with it." The article gives no indication of concern that these passengers may have just transported the disease, however (Western Star, 30 May 1918).

The apparent lack of concern is interesting because railroad travel was clearly of importance in the spread of the second wave through the island. The major rail line linked St. John’s in the northeast corner of the island to Deer Lake, Corner Brook, and Port-aux-Basques in the west. Branch lines led from the main line to most of the Avalon Peninsula, the Bonavista Peninsula, and Fortune Bay (see Figure 1). Interview data and newspaper reports both indicate that there was daily railway traffic throughout the northeast area of the island and from Placentia to St. John’s. There was also daily travel in both directions from Port-aux-Basques to St. John’s, and information collected in interviews during the summer of 2006 suggests that this travel probably directly contributed to the large epidemic that hit the St. George’s Bay area during Wave II. Interview data also indicate that train travel may have contributed indirectly to the epidemic at Bonne Bay in the southern St. Barbe district, because people would often travel by train to the Bay of Islands in the St. George district and then by boat to Bonne Bay.

The role of train travel in spreading the flu across the island, and to communities that might have otherwise escaped Wave II, is also supported by interview data and by the occurrence of deaths in a string of inland towns along its route (e.g., South Branch, Spruce Brook). For example, the Western Star reported that "Up to Saturday night there had been fourteen deaths from the influenza epidemic between Deer Lake and Spruce Brook, not including those at Lark Hr., and since several more deaths have been reported" (6 November 1918).