Posted on

‘At the Junction of the Bad and Red Deer Rivers.’ Searching for Peter Fidler’s Long Lost Chesterfield House. Have We Finally Found It?

Wooden statue of trader, mapmaker, Peter Fidler, Elk Point, Alberta, Canada. Fidler served at the nearby Hudson’s Bay Company Buckingham House (c.1792-1800), located along the North Saskatchewan River, just southeast of Elk Point. https://farm4.staticflickr.com/3529/3967919062_060d0fee79_z.jpg

Note: This is a revised and condensed version of an article we recently published in the Saskatchewan Archaeological Newsletter Quarterly, May, 2021 edition, regarding our search for the the Chesterfield House fur trade sites in Spring, 2021. Readers are referred to this edition of the Quarterly for a more detailed version of our findings.

Time and the Unknown

Ah, the mystery of the unknown! It’s one of the things that first drew me to history and archaeology. The thrill of discovering new facts, objects or places, lost or abandoned centuries ago. It didn’t matter if they were only minor footnotes in the bigger picture of human history.

One of the most rewarding and challenging experiences in my career was searching for the many lost fur trade posts in western Canada. The remains of some lay hidden in front of our very noses. Others, so remote and covered by nature, it took considerable effort or sensitive equipment to eventually find them. Still others guard their hiding places well, and to this day, elude discovery.

The remains of the last Hudson’s Bay Company Fort Edmonton (c.1830-1915), located on the Alberta legislature grounds. A fort, whose location was known by only a few historians and archaeologists. In a survey, conducted while excavating this fort in the early 1990s, we discovered that over 60% of the public had no idea the original fort was located on the current Alberta legislature grounds in the heart of Edmonton, Alberta, Canada.

This is a story about a search for one of those fur trade post that has eluded us for many years – Chesterfield House. A search that began in the mid-1960s. But for me it began in c.2005 and continues to this day. Because no one has yet found Chesterfield House.

Searching for Canada’s Fur Trade Forts

In an earlier blog I talked about explorer and mapmaker David Thompson. One of the world’s most remarkable geographers and mapmakers. Thompson visited many western fur trade forts and wrote about them or mapped them. Often he left behind clues for us relocate them. Such as the NWC/HBC Fort Vermilion I (c1798-1830) site in northern Alberta. (https://canehdianstories.com/wp-admin/post.php?post=1894&action=edit)

In this post I focus on another lesser-known but equally competent trader, surveyor and mapmaker, Peter Fidler of the Hudson’s Bay Company (HBC). And in particular, his brief, and sometimes scary stay in southern Saskatchewan at the confluence of the South Saskatchewan and Red Deer Rivers where he would build his fort.

In the fall of 1800 Fidler built Chesterfield House for the HBC. Soon after the North West Company (NWC) built alongside the HBC post, followed by the XY Company. Many (including me) have searched for them but, to this day, they have never been found.

The confluence of the South Saskatchewan and Red Rivers today, near the Saskatchewan-Alberta border, Canada. Somewhere down there on the river flats are the remains of three fur trade forts, over two-hundred years old. Their whereabouts remains a mystery.

Peter Fidler

Born at Bolsover, Derbyshire, England, Peter Fidler (16 August 1769 – 17 December 1822) joined the HBC in 1788. He was trained in surveying and astronomy by Philip Turnor who also trained David Thompson. Fidler became the Company’s chief surveyor and map-maker, much like David Thompson for the NWC.

While acting as trader, explorer, and mapmaker, Fidler also observed and wrote about the Indigenous peoples of the region. He married a Cree woman and learned Native languages to carry out the trade. Occasionally he convinced his Native informants to draw maps of their territories for him. Today these are some of the few surviving Native maps of western Canada (see a former post on the Ki-oo-cus map of southern and central Alberta. (https://canehdianstories.com/wp-admin/post.php?post=266&action=edit). His journeys, largely undertaken in western Canada, covered an estimated 48,000 miles on horseback, foot, canoe and dog team.

Fidler’s superiors admired his toughness and fortitude. For example, while traveling and wintering with the Chipewyan in northern Alberta and the NWT, a near-starving Fidler mentioned what parts of a game animal they ate to stay alive: “We eat everything except the manure.”

Fidler had some incredible adventures as a trader and explorer for the HBC. A few could have ended his life. One of these adventures required constructing a fur trade post on the Western Canadian prairies. He built the fort with the intent of trading with Plains First Nations peoples. After only a few years, Fidler and the other Companies abandoned their forts, barely escaping with their lives.

A map, by Peter Fidler, of the Upper Assiniboine and Swan Lake Regions. Fidler was a very accurate surveyor and cartographer. Not only did he map the lakes, rivers and important land features, he also accurately plotted the locations of the various fur trade post on those maps. This point becomes important later.
Like David Thompson, Peter Fidler used a line-track survey method when mapping the South Saskatchewan River. He would take a compass bearing and then estimate a distance to map that part of the river. This is part of Fidler’s survey of the South Saskatchewan River, up to the confluence of the Red Deer River, superimposed over today’s South Saskatchewan River route. Based on these and other evidence, we assume that Fidler was an accurate surveyor, especially calculating latitude.

Where did the Companies Build?

For many years I heard about the mysterious Chesterfield House(s) and attempts to find them. All searches ended in failure. But why? How could three forts of considerable size, just disappear, without a trace, in the valleys of the Saskatchewan and Red Deer Rivers? Or perhaps, as some researchers suggested, had those waters already swallowed them up leaving no trace behind?

In 2005, while visiting and hunting in the area, and intrigued with the lost Chesterfield House, I too joined the search.

The South Saskatchewan River Valley near Empress. So beautiful with its wide open prairie expanses and bright blue skies.

As with other similar searches, nothing is ever as simple as it first appears. This quest was no exception. It has taken me since 2005 to finally piece enough evidence together to make the modest claim that I might have a candidate where these fur trade forts were built. And I, like others before me, could be totally wrong.

Let’s start our search with Fidler’s Chesterfield House HBC journals (1800 – 1802). In them he gives only a few but very specific references to the fort’s location.

This photograph was taken from the east looking towards the forks of the South Saskatchewan and Red Deer Rivers. Fidler’s two references to the fort’s location are pretty specific. At first I thought he might have built on the island you see in this image which is in front (east) of the juncture of the two rivers. But then in the second quote he specifically says they built on the north side of the river(s). The ‘Bad’ River refers to the South Saskatchewan River. But Fidler has some reservations building on this spot: “Crossed the river to north side and looked out for a place to build at. The woods here are few and bad for building with.” (From Alice Johnson 1967:268. Saskatchewan Journals and Correspondence 1795 – 1802. The Hudson’s Bay Record Society, Volume XXVI) So, Fidler either moved to where there was more suitable wood to build with or he cut wood from elsewhere and hauled it to the junction of the two rivers.

Fidler gives the latitude of the south bank of the Red Deer River where he intends to build: 50o, 55’, 5” (50.9222o). Fidler’s latitude calculations were quite accurate. Longitude was not. But, if we take Fidler at his word, we really don’t need longitude because Fidler gives us a fairly precise east-west reference point where he built the fort – the confluence of the two rivers.

A satellite image of the confluence of the two rivers. The problem with river confluences, is that they can move. You can see the old Red Deer River channels in this image (shown in dark green). At one point in time it flowed into the South Saskatchewan River further north. Some researchers believed this was the original confluence in 1800 and looked for the forts in that area. But, Fidler’s 50o, 55’, 5” (50.9222o) is much closer to the present confluence than to the northern older one.
Peter Fidler’s longitude for the confluence of the rivers was out a considerable distance. Not unusual in those days when highly accurate time pieces were required to estimate how far west from Greenwich Mean time you were located. Fidler’s latitude however, was remarkably accurate, being approximately 15″ or +/- 450 metres out.

While rereading Fidler’s published journals (for the umpteenth time) this spring, I noticed at the end of the 1800-01 trading season a note by the editor: “[Meteorological and Astronomical Observations, made at Chesterfield House, covering 15 manuscript pages, not printed]” (From Alice Johnson 1967:268. Saskatchewan Journals and Correspondence 1795 – 1802. The Hudson’s Bay Record Society, Volume XXVI). I wondered if Fidler gave a more accurate reading of latitude and longitude for Chesterfield House in those unpublished notes. Fortunately I was able to get hold of a copy of his original journal, including the missing 15 pages.

Yes, indeed. Fidler gives two readings for latitude at Chesterfield House: 1) 50o55’12” (50.920o); and, 2) 50o55’21’’ (50.9225o). Both readings are relatively similar to his original north riverbank reading recorded at the confluence of the two rivers where initially he wanted to build. Had they been significantly different then it might suggest he built elsewhere (than right at the confluence).

Next I looked at a few historic maps of the area. One map shows the location of Chesterfield House, marked by a dot, on the north side of the South Saskatchewan River, some distance downriver from the forks. But Fidler stated, “…opposite the mouth of the Red Deers River where we are to build…”

This particular map of the western prairies, shows the confluence of the South Saskatchewan and Red Deer Rivers, with a dot depicting the location of Chesterfield House. That dot is east of the forks on the north bank of the South Saskatchewan River. It likely refers to the later post built by the HBC downriver from the forks in 1821 (which also has never been found). (https://earlycanadianhistory.ca/2018/06/18/what-peter-fidler-didnt-report/)

Then I found another map drawn by both Fidler and his Blackfoot informant, Ak ko Wee ak in 1802. Does the straight line across the Red Deer River, with Chesterfield House’ written on it, indicate where the fort was built? If so, it was built upriver from the confluence of the two rivers.

A map drawn by Fidler’s Blackfoot informant Ak ko Wee ak in 1802. Fidler wrote the names of places on the map. On that map there is a straight vertical line across the South Saskatchewan and Red Deer Rivers. If the line represents where the fort was built, then Chesterfield House was located some distance (distance unknown, because there is no scale) up the Red Deer River. (From HBCA PAM: E.3/2 fos. 103d)
On this map, drawn by Fidler’s informant, Ak ko mok ki, in 1802, there is a little drawing of Chesterfield House located on the north side of the river(s). But its location is very general. The fort could be anywhere within miles of the forks. (From HBCA B.39/a/2 fo.93)

Unfortunately no one, while searching for the forts, has taken the Native maps or Fidler’s latitude reading of the forts location very seriously. Keep in mind, Fidler was a very accurate surveyor for his day, especially when it came to calculating latitude. His readings were out by about 15 seconds of latitude, or +/- 450 metres.

If we ran his latitude for Chesterfield House as a straight straight line across a current map, assuming about 15 seconds (~450m) of error (shown by orange dashed lines on either side of the black line), where might the fort(s) be located?

Peter Fidler’s latitude projected onto a current map of the confluence of the South Saskatchewan and Red Deer Rivers. The orange dashed lines represent the margins of error around his calculations. The orange elliptical shapes are where other archaeologists have searched for the lost forts. The blue elliptical shapes are where no one has searched but are well within Peter Fidler’s margin of error for latitude.
Peter Fidler’s latitude for Chesterfield House, and margins of error, superimposed on a current satellite image of the area. The small yellow rectangle represents the area we are interested in and will discuss shortly. The other dashed line further north is another area archaeologists searched for the forts. The latitude of this area is well beyond Fidler’s margins of error.

Another little hint, where the Chesterfield forts might have been built, was a comment in Peter Fidler’s journals. “Dug up the small bateau that was laid up in the spring: the heavy rise of water in the summer had buried it four feet deep in sand.” (From Alice Johnson 1967:268. Saskatchewan Journals and Correspondence 1795 – 1802. The Hudson’s Bay Record Society, Volume XXVI). Presumably the boat was near the fort. If so, it suggests the fort was located on an inside meander of the river, where flood sediments are deposited. Instead of the outside meander where high water cuts away the bank.

We haven’t looked for these forts at all the possible places that are within Fidler’s range of error for latitude. And there are reasons for it. This is large area filled with dense wolf willow scrub and wild rose bushes that is not too pleasant to walk through, or find things. And Fidler’s reference to the forts being built at the forks of the rivers has perhaps been taken too literally. Would the London Committee reading his journals really care if he built a mile or two either way of the forks?

Searching for rock historic building chimney piles in the dense bush on the lower terraces of the Red Deer River. In some places visibility is poor and walking is tough. Currently there is no LIDAR (laser imaging, detection, and ranging coverage) for this area, which would help immensely to expose detailed surface ground contouring and possible evidence of the sites.

To add yet another obstacle to our search, not everything historical in this area is related to the early 19th century fur trade. This area was occupied and traveled over for thousands of years by First Nations Peoples. It became an important Metis settlement, Riviere La Biche, in the 1870s and 1880s, which would have left physical remains similar to those present at earlier fur trade forts.

A late 19th century chimney and fireplace, built by trappers, the Adsett brothers, still stands in one of the local farmyards in the area. The former Metis settlement of Riviere La Biche, was located around the confluence of the South Saskatchewan and Red Deer Rivers. People built chimneys and fireplaces, similar to these, during the early fur trade. These and other building remains, such as cellars, would be indistinguishable from one another without detailed archaeological exploration. However, unlike the Metis cabins which are scattered over a large area, the Chesterfield House forts, and their buildings, were built in a smaller area, surrounded by stockades.

The Search Continues, Spring 2021

This spring (2021), when preparing this blog, I wanted a good satellite image of the forks area where Chesterfield House might be located. While doing so I noticed a long rectangle-shaped, light-colored outline on the satellite image.

This is the image of the Red Deer River flats on one of the meanders that I first looked at. At this height do you see what I see? Or, do I just have an supercharged imagination?
Here is a closer view of the satellite image. Do you see the long rectangle, lightly highlighted, oriented in a northeast-southwest direction in this image? Quite often simple aerial photography and other types of imagery can pick up features from the air, not seen on the ground by the naked eye.

This is the feature I see when looking the the satellite image. The long rectangle is relatively well-pronounced. The other lines to the west are not as definite. The NWC and HBC were built together, enclosed in a common stockade, which is explains the long, rectangular outline. Fidler states the XY Company built just west of his fort.

After reviewing the historic documents, satellite images, and constructing arguments that this might be the lost Chesterfield House forts, the next step was to re-revisit the site and look for physical clues on the ground.

So, my wife, Gabriella Prager, also an archaeologist, and I drove to the Empress area in April to see what we could see. When I visited this same location in 2005, I saw some rock scatters and slight depressions. It was time to reevaluate what those features might be, relative to this new-found evidence.

Once there, we looked for depressions, pits, rocks or mounds or any other evidence that could indicate a human occupation. The surface of this area is quite undulating and uneven from repeated flooding and scouring over the years. Just how much sediment covers the original 1800 ground surface is uncertain without excavating. However, based on other floodplains of this vintage (e.g., the NWC/HBC Fort Vermilion I site, northern Alberta), there could be as much as one-half metre or more sediments covering the original land surface and the remains of anything built on that surface. Fidler’s description of the bateau buried in over four feet of river sediments is most telling in this regard. And that was just one of many flooding events since then.

The area in question, where the long rectangular outline in the satellite photograph appears. Slightly elevated, the area contains little shrubbery, as was also the case in 2005.

When walking the area we noticed the ground was slightly elevated on the east and south sides. These elevated areas were likely responsible for the light-colored lines we saw on the satellite image. Normally, old stockade lines are slightly depressed, even after flooding. We did however also notice a few rock scatters and slight depressions with the rectangular outline.

Walking along the elevated ridge on the east side of the rectangle. This could be an old river terrace edge. The south edge is also elevated. However, there are no visible surface signs of anything where the west and north lines occur on the satellite image.
A small scatter of rocks. Possibly the remnants of a fireplace. But from what time period? Remember, this is a floodplain and rocks don’t float. So, it’s not a natural event. This definitely is evidence of human activities.
Gabriella Prager taking notes and GPS coordinates of a small depression on the site.

What We Concluded

It would be folly to state, without first excavating and testing this area, that we have discovered the Chesterfield House sites. We first need to find certain kinds of other archaeological evidence to suggest that these features, and that intriguing rectangular satellite image, are related to the early 19th century fur trade, and not some later period Metis household: 1) footer trenches representing palisades; 2) early 19th century artifacts representing the time period in question; 3) more building remains confined to the rectangle; and, 4) considerable amounts of animal bone from both human consumption of wild game and making meat provisions for the trip downriver.

There are things about this site that are troubling and do not fit what I expect to see on the surface of the ground; if this were a historic fur trade fort. First is the lack of more obvious visible surface features such as chimney piles and cellar depressions. Second, is the lack of visible faunal debris, or any artifacts. Given the amount of meat consumed, animal bone remains are typically considerable at forts such as this.

To some degree, this lack of evidence might be explained by the amount of flooding that has occurred in the area. If substantial, it may have covered any historic remains with considerable sediments and infilling most depressions that would be cellars, privies, and refuse pits. However, at other fur trade sites abandoned for over 200 years and constantly flooded, we have observed more pronounced surface features than we see here. However, currently we know little about flooding episodes and depositional rate of sediments of the Red Deer River, which could be quite different from our northern rivers.

To be clear, without further investigations, what we (and others) have found is definite proof of a human occupation of some sort at this spot. Based on the historic evidence regarding Chesterfield House, this location is a suitable candidate for these early NWC, HBC and XY Company forts. But, that’s as far as we can go presently. The area warrants further archaeological investigations to either refute or verify our claim.

EndNote

For those of who you who are aspiring students of history or archaeology, there’s a simple lesson here. Combining the evidence from two disciplines (history and archaeology) usually results in a more complete understating of human history. Not always, but better two independent lines of evidence to examine a problem of history, than only one. And perhaps, with the new remote sensing imagery, more than only two disciplines is necessary to eventually find these rather elusive historic forts.