Three Flathead House Locations

A New Perspective on the Legacy* of Saleesh House

*The Merriam-Webster Dictionary defines legacy as:
"something that happened in the past or that comes from someone in the past"

Too often an interest in a particular historical event is so intensely focused on a single event that the context within which it occurred is lost. When that happens history can become very misleading and sometimes even factually distorted.
In my opinion, such is the case with David Thompson and the legacy he left behind following the four years he spent in the vast, unexplored wilderness of the Pacific Northwest at the beginning of the 19th Century. In fact, the legacy he left was one of an extremely successful business that lasted for more than 60 years.
On April 30, 1803, U.S. representatives in Paris, working under the direction and authority of President Thomas Jefferson, agreed to pay $15 million for about 828,000 square miles of land—about $2.00 an acre—that stretched from the Mississippi River to the Rocky Mountains and from the Gulf of Mexico to Canada. Very little of the remote reaches of the lands involved had been explored.

Louisiana Purchase

Map courtesy of the History Channel

Jefferson's next step was to authorize the Lewis & Clark Corps of Discovery with a single goal in mind—to find an overland route to the Pacific Ocean. The expedition began their journey from a spot near St. Louis, MO, in May 1804, and spent the winter of 1804–1805 with the Mandans on the upper Missouri.
The Corps reached the shore of Pacific Ocean in November 1805, and built Fort Clatsop where they spent the winter of 1805–1806. On the return trip, Lewis and Clark split up in order to explore more territory and look for a more direct route home. To accomplish their mission the Corps traveled more than 8,000 miles by boat, on foot and on horseback.
The Corps of Discovery was primarily a military venture. Its mission was to find a direct route from the Mississippi River to the Pacific. Along the way they were to map the country and maintain journals recording their observations about people they encountered along the route as well as produce drawings and written descriptions of native flora and fauna. They performed with military precision and did an extraordinary job for the President.

Lewis and Clark crossing the Clark Fork River

Lewis & Clark crossing the Clark's Fork River near Missoula
on their return trip to St. Louis. Painting by E. S. Paxson

In the fall of 1809, three short years after the Corps of Discovery had passed through the region, North West Company (NWCo) trader-explorer-surveyor David Thompson reached the valley of the river named by Lewis and Clark—Clark's Fork of the Columbia. Thompson called it the Saleesh River.
Thompson was the first European to establish a permanent business in the region when he built two trading posts along the banks of the Saleesh. The first, Kullyspel House, was located near Hope, ID. The second, Saleesh House, was built near Thompson Falls, MT. Thompson remained in the region until the spring of 1812 when he returned to Montreal and retired from the fur trading business.
Kullyspel House was closed after only two seasons of operation. The trading enterprise started at Saleesh House remained a busy, economically successful business run by the NWCo until 1821. That was the year when the fur trade in the region took a major turn. The North West Company and the Hudson's Bay Company (HBC) merged their businesses or in their words, "amalgamated" their holdings. Their agreement provided that all trading posts would be owned and operated by the HBC despite the fact that most of the men working for the new partnership were Nor'westers—former employees of the NWCo.
Following the acquisition of all NWCo posts by the HBC, the name of Saleesh House was changed to Flathead House. Trade from the original post continued until about the mid 1820s. At that time, the original post, built by Thompson, was abandoned and a second "Flathead House" was built at a spot a short distance downriver from Weeksville, MT. The HBC continued trading from that location until a decision was made to move it yet again.
In 1846, construction of a third "Flathead House" was started on Post Creek between the towns of St. Ignatius and Ronan, MT. Angus McDonald, who had been in charge of second Flathead House during the 1839 outfit year was transferred from Fort Hall (ID) in 1847 to take charge of the third Flathead House. Shortly after he arrived he re-named it Fort Connah and remained in command of the post during outfit years 1848–1854 before being transferred to Fort Colvile (re-named Colville) and appointed Chief Trader for the entire HBC operation west of the Rocky Mountains and south of the 49th Parallel.
The HBC continued to operate at the Fort Connah location until the US Government ordered it permanently closed in 1871. It was the last HBC post to operate in what is now the United States.
A map in the National Atlas of Canada shows the locations of HBC fur trading posts for the period 1600–1870. The HBC acquired many of their posts from other companies. In such cases, the owner of the post before acquisition by the HBC and the total length of time it was in operation are provided. A portion of the Atlas' map covering posts located in the Pacific Northwest and southern British Columbia is shown here.

Hudson's Bay Company Posts

Pay special attention to the solid red triangles associated with several of the posts. These represent the posts that were owned by the NWCo and that operated for at least 16 years—long enough to be amalgamated by the HBC in 1821. Note that the date of construction for Flathead House was 1809—the year Thompson built the post and named it "Saleesh House."
Kullyspel House, originally constructed in 1809 by Thompson, was closed in 1811 after only two year of operation. The empty red circle indicates that it operated for three years or less. Thompson closed the post prior to the amalgamation—it was never operated by the HBC.
For some reason the HBC's Fort Connah, located on Post Creek between the towns of St. Ignatius and Ronan, MT, did not appear on this map despite the fact it remained in operation until 1871.

Hudson's Bay Company Posts

This map insert shows the fur company that originally constructed and operated each of the posts as well as the number of years it remained in operation. A statement on the document reads:
"The date applied to a post on the map indicates the first year in which it is known to have been operated. Names and dates have been applied to all posts that were in operation for four years or more, and where space allows, to those that were in operation for less than four years."
The red triangle for Saleesh/Flathead House indicated the post remained in business for more than 15 years. Sometime during the mid 1820s the post burned. Whether by accident or on purpose we do not know. What we do know is that following the 1821 amalgamation the post's official name was changed to "Flathead House" by the HBC.
About the time the original Saleesh/Flathead House burned, we begin finding references to it in the journals of several HBC employees who were assigned to Flathead House (aka Flathead Post) during the mid 1820s. Based on our research, I believe the new location—the second Flathead House/Post—was built a short distance downriver from where today's Weeksville Creek empties into the Clark's Fork.
An important clue for the location of the second Flathead House was found in the journal of William Kittson who was traveling with Peter Skene Ogden on his Snake Country Expedition. In September 1825, while returning to Spokane House from the Snake River country, Kittson arrived at Flathead House. He recorded the following about arriving at the Horse Plains, where Plains, MT, is now located:
"Friday 29th Encamped this afternoon at the North end of the Horse Plain... Saturday 30th Continued my march...Passed and visited the Flat Head House found it complete. Had some trouble in getting over a rock [Bad Rock] below the house... Sunday 31st Continued the march...Encamped a the foot of Falls [Thompson Falls], 12 miles from Flat Head Post..."
Note that in this journal entry, Kittson used both the terms "house" and "post" when referencing Flathead House. For orientation purposes, following is a clip from Kittson's 1825 map. His original map shows North being to the left side of the sheet. Most maps are oriented with North towards the top of the sheet. For ease of comparison with other maps and images, North is toward the top of the following section of Kittson's map:
  • Circle #1 shows where Kittson drew the location of the post. Note that on his map, Kittson calls the post Fort Flat Head.
  • Circle #2 shows the Horse Plains/Dog Lake area. The trail leading to the SE is the Road to the Buffalo along Camas Creek.
  • Circle #3 is the confluence of the Flathead and Clark's Fork Rivers.
  • Circle #4 is Flathead Lake.

William Kittson's Map

William Kittson's 1825 map

Compare the Kittson map with the following clip from Thompson's 84° Map. Thompson's map shows an important river crossing that did not appear on his original map published in 1814. This crossing was obviously the place where fording the river was possible for at least a part of the year.
For orientation, note the location of Saleesh House (labeled "N.W.C. Obs") near upper left and "Horse Plains" near the center of the map. Between these two sites is a trail (dotted line) leading south and connecting to another river. The trail can be seen on the south side of the river in the white circle. It followed today's Swamp Creek to its headwaters. Swamp Creek empties into the Clark's Fork about a mile downstream from the mouth of Weeksville Creek.

David Thompson's Map

From Thompson's 84° Map published in 1826.

This is the trail I believe Thompson traveled on his exploratory trip to the vicinity of the present day towns of St. Regis and Superior, MT, in 1811. The confluence of the Clark's Fork and Flathead Rivers (lower right) is about a mile above the town of Paradise, MT.
The Clark's Fork is shown flowing to the north at the lower edge of the map. At the point where Thompson shows the Swamp Creek trail leading to the Clark's Fork (in the white circle), the river abruptly changes directions, first to the southeast, then again to the north just before it converges with the Flathead River flowing in from the east. The towns of St. Regis and Superior, MT, are located on the Clark's Fork along I-90 a few miles south from the bottom of this map.
As previously noted, I want to address the longstanding argument about whether or not there were three separate "Flathead House" locations. In our opinion, the answer is YES!
We believe the facts support that conclusion, and here is how we got there after locating a map that was what we considered to be the final piece of that puzzle. After 1821, three Flathead Posts belonged to the HBC. However, at any given time between the years 1821 and 1871, only one was actually in operation.
The first was the original post near Thompson Falls, MT, built by David Thompson in 1809 named Saleesh House. It was renamed Flathead House in 1821 by the HBC when it was merged with the NWCo. This post continued in operation until burning down in the mid 1820s.
As a result Flathead House was relocated about ten miles upriver so that beaver hunters coming from the east by way of the Horse Plains did not have to cross Bad Rock using the trail shown in this photo. There is no doubt it presented a challenging impediment to travelers, especially old people, children and horses. The toe of the nearly impassable, solid-rock, outcrop dropped straight off into the river until construction of the Northern Pacific Railroad was underway and the base was blasted away to allow the track to be laid on level ground. The tracks can be clearly seen in this photo.

Bad Rock

This is a photo of Bad Rock taken from the south side of the
river showing where the trail is located. Photo by Linda Haywood

Father Pierre-Jean De Smet, a Blackrobe, working among the Flatheads and other regional tribes, noted in his journal that on November 3, 1841, when traveling along the Saleesh Road to the Buffalo, his party had to cross Bad Rock.

Father De Smet

Courtesy of Gonzaga University

This is his description of that experience, one of the best available, copied directly from De Smet's Letters and Sketches, 1841–1842, published in Philadelphia in 1843:
"On the 3rd of November, after prayer and instructions to the savages, we continued our march. We were on the borders of the Clarke Forks, to which we were obliged to keep close during eight days, whilst we descended the country bordering the stream. The river is at this place of a greenish blue, very transparent, caused probably by the deposit of a great quantity of oxygen of iron. Our path during a great part of the day was on the declivity of a lofty, rocky mountain; we were here obliged to climb a steep rough pass from 400 to 600 feet high. I had before seen landscapes of awful grandeur, but this one certainly surpassed all others in horror. My courage failed at the first sight; it was impossible to remain on horseback, and on foot my weight of two hundred and eleven pounds, was no trifle. This therefore, was the expedient to which I resorted: My mule Lizette was sufficiently docile and kind to allow me to grasp her tail, to which I held on firmly: crying at one moment aloud, and at other times making use of the whip to excite her courage, until the good beast conducted me safely to the very top of the mountain. There I breathed freely for a while, and contemplated the magnificent prospect that presented itself to my sight.

"The windings of the river with the scenery on its banks were before me, on one side hung over our heads, rocks piled on rocks in the most precipitous manner, and on the other stood lofty peaks crowned with snow and pine trees: mountains of every shape and feature reared their towering forms before us. It really was a fine view and one which was well worth the effort we had made. On descending from this elevation I had to take new precautions. I preceded the mule, holding her by the bridle, while she moved cautiously down to the foot of the "Bad Rock," (as it is called by the savages,) as though she feared stumbling and rolling with her master into the river which flowed beneath us."
With that vivid picture in mind, imagine what it must have been like trying to move men, women and children of all ages as well as sometimes hundreds of horses and dogs laden with gear and baggage across such a place. This is how the actual tread of the ancient trail appears near the top today.

Bad Rock

Original tread crossing Bad Rock. Photo by Jim Manley

The next two photos are history buffs Ron Petrie and Tim Leventhal on the Bad Rock trail during a five-day primitive canoe camping trip on the Flathead and Clark's Fork Rivers in September 2018. They launched their canoes below the dam at Polson, MT, and finished their journey at the sandy canoe landing where Saleesh House was located.

Climbing Bad Rock Trail.

Ron Petrie on the Bad Rock trail. Below is Hwy 200, railroad
tracks and Clark's Fork River. Photo by Tim Leventhal

Climbing Bad Rock Trail.

Tim Leventhal on grass covered trail across Bad Rock. Note the
absence of grass above and below the trail. Photo by Ron Petrie

Ron has surmised that the grass growing in the trail on top of the rocks continues to exist because of the tons of horse, mule and dog manure left behind by thousands of pack animals that undoubtedly traversed this treacherous trail for hundreds of years. I tend to agree since there would have been very few, if any, places to get off the trail from one end to the other.
This relatively short stretch of trail climbed some 300 feet up, across and down the other side of Bad Rock in a distance of no more than ¼ mile. It would likely have taken more than an hour to pass, especially when small children and elders were along.
The second Flathead Post was constructed east of Bad Rock on the north side of the Clark's Fork River near the mouth of Swamp Creek, which flows in from the south.
This 1889 GLO map shows where a remnant of the original Indian road still exists where it crossed a ridge near Flathead House to avoid a meander and slough of the river that existed prior to the construction of the railroad in 1882.

Indian Road

This is how a remnant of the original trail looks today.

Indian Road

Photo by Linda Haywood

We plotted the route using the GLO map shown above as the base then measured and plotted the distances from the nearest section or quarter-corners to where the road crossed external and internal subdivision lines. Once that was completed, we pulled up the area on Google Earth, overlaid it with the U.S. Geological Survey (USGS) quadrangle map covering the area, marked those spots and connected the dots. This is the result that should be very close. The red line is the location of the original road.

Indian Road

The east termination point, the white pin marked OBS, was where Thompson noted in his Journal on October 15, 1809, that he took an observation of the sun at noon to determine his latitude. He recorded it as 47° 32' 08". That parallel is actually about a half-mile north of the Indian road. In my opinion, his OBS would have been from a spot very near the trail, not a half mile away on a mountainside.
The mouth of Swamp Creek can be seen on the south side of the river at the lower left. My original conclusion about where the post stood is shown in the rectangle on the north side of the river. I now believe I was wrong. I wasn't far off initially but I want to keep you informed of my latest thoughts about it as we have dug more deeply into the subject and spent more time on the ground in the area.
When picking a site for a new trading post, it was primarily about three things: location, location and location. There had to be (1) good water, (2) pasture for horses, (3) a good camping area for hunters and their families to pitch tents nearby and (4) good trail access to and from the site to other major Indian roads.
The red line on the image above as well as on the following image is where Thompson's Saleesh Road to the Buffalo was located because of the original river channel. Construction of the railroad in 1882 (where the interpretive signs are now located) cut across the old channel, forever changing the hydrodynamics of the river. The spot where Thompson took his OBS was most likely where the white marker labeled OBS is placed. The yellow pin near the top of the image marks the actual parallel of latitude he recorded.

Thompson's Observation

The guide Thompson hired to get him back to the Big Bend, in time to meet his supply brigade coming down McGillivray's River, was less than honest. In fact, as it turned out, the guide was not at all familiar with the shortest route and got them lost.
This was Thompson's first trip on this trail and he noted several times in his Journal that they should be traveling north, not east. Finally, after reaching the Little Bitterroot River, he fired his "guide," then managed to find his own way back to the rendezvous spot. Despite the many miles of travel in the wrong direction, he missed his supply brigade by only a few hours, catching up with them a short distance above what is today known as Kootenai Falls, a few miles upriver from Troy, MT.
The following GLO map of the south side of the river shows two primitive trails leading to the location of the second Flathead Post. Keep in mind that it remained in operation for about 25 years. Also note the trail along Swamp Creek that appeared on Thompson's 84° Map shown previously.

Flathead House Trails

Note the two trails leading to the second Flathead House location, one from
the mountains to the south, the other from the southeast along Swamp Creek.

No early GLOs, or other maps I have seen cover this "Mountain Trail" farther than about five miles to the south. Consequently I decided to start working along another early, well-recognized and described, Indian road through the region later called, and better known, as the old Mullan Road. The route generally follows Interstate 90 along the St. Regis River to the Montana–Idaho border.
Again, using the earliest GLOs available, I began searching for segments of trails leading to the north that were generally on northeasterly bearings leading in the direction of my last available, mapped point on the trail south of Flathead House.
That attempt resulted in the discovery of other maps showing additional old trails that I believe tied into the network culminating at the Flathead House location.
We strongly believe that Flathead House once stood on the ground identified above. I also believe it is highly likely that at least some related evidence about the location still exists. However, archeological work is absolutely necessary if any remnant of the post is ever going to be found. I am an amateur historian and do not have the expertise in that field to even attempt to pinpoint the spot, nor have I been able to find anyone willing to put together such a project. Perhaps sometime in the future, before it is too late, someone else will pick up on the project and make an attempt.
Two roadside interpretive signs, one about Flathead House, the other about the "Saleesh Road to the Buffalo"—Kootenai Trail—can be viewed at the pullout on Highway 200 just west of Weeksville. The location of the pullout is shown on the Google image above.

Road to the Buffalo interpretive sign

Flathead House and Road to the Buffalo information
signs near Weeksville. Photo by Carl Haywood

For an even better perspective of the Flathead House location, this Google image provides terrain relief information at Weeksville.

Road to the Buffalo route

Angus McDonald knew exactly where the second and third Flathead Posts had been located. He had been in charge of each during his years with the HBC. Shortly after being placed in charge of the third post in 1847, he re-named it Fort Connah. He was eventually promoted to Chief Trader at Fort Colvile (later spelled Colville) in charge of HBC operations in the northwest.
Angus' son, Duncan, was placed in charge of the post after his father had been re-assigned and remained in that position until the U.S. Government ordered it permanently closed in 1871.
This is the HBC record showing who was in charge of Saleesh House/Flathead House for the years 1809–1871. Note that David Thompson's name is listed for the outfit year of 1809—the year he built the original post. The names of Finan McDonald and James McMillan also appear in the early years. Both had served as Clerks under Thompson. McDonald remained in charge of the post for the first three years after it was acquired by the HBC. Note that Angus McDonald and Finan McDonald were not directly related.

HBC Post Managers

Fort Connah

According to Angus' daughter, Christina, her mother, Catherine Baptiste, whose father was an Iroquois Frenchman who worked for the HBC, her mother was Nez Perce, a cousin of Nez Perce Chief "Eagle-of-the-Light."
She was born at Fort Hall (today's Fort Hall, ID) on September 20, 1847 "...on the Big Camas Prairie near what is now Boise, Idaho. When I was ten days old I was taken to the old Hudson's Bay Post at Post Creek, in the Flathead country. Shortly after, Father was transferred in charge of the Hudson's Bay Company post, Fort Colville."
During the time Angus was in charge of the third Flathead Post, he and his family claimed ownership of lands around the post and remained living in the area following its closure.
During the period 1860–1871, Angus made several trips between Forts Colvile and Connah on business and to visit family. The notes he jotted down along the trail during those trips provide great bits of information about where the HBC posts were located.
His comment about the location of second Flathead Post places its location almost exactly where we have been discussing near Weeksville. In a letter dated September 22, 1856, Angus noted that:
"...the old fort [Flathead Post] which is no more was abandoned in consequence of the building of the new fort in 1847."
The "new fort" McDonald mentions is best known as Fort Connah located on Post Creek just east of Highway 93 about six miles north of the town of St. Ignatius, MT. A Montana Department of Transportation information sign is posted at a pullout on the highway. There is no doubt in my mind that it was originally called Flathead Post. The information that removed any lingering doubts about the connection is presented near the end of this discussion.


Few people know that the lands included in what would become the Oregon Territory in 1846 were the only lands in what is now the United States to be claimed by proclamation. We know from David Thompson's Journal that he did just that by leaving a note in charcoal on a board at Saleesh House and planting a pole with a note attached at the confluence of the Snake and Columbia Rivers.
As a matter of interest, note the red circle on the map is known as "The Angle" and was surveyed by none other than David Thompson while employed by the International Boundary Commission. But that's another story.

The Angle

"The Angle" surveyed by David Thompson is located in the red circle.

Following the War of 1812, Great Britain and the United States agreed that the 49th parallel would be the official boundary between the US and Canada. However, they also agreed that joint occupation by the United States and Great Britain would continue within the boundaries of the future Oregon Territory.
As time passed, tolerance by the US Government for HBC operations and the rights of non-US citizens became more restricted. Articles in a related treaty signed in 1846 clearly defined a few important limits and restrictions that would be imposed if the HBC wanted to continue it operations on US soil. The treaty also dealt with legal rights of HBC employees and other non-citizens associated with the HBC.

Oregon Treaty of 1846

Treaty of 1846

"From the point on the forty-ninth parallel of north latitude, where the boundary laid down in existing treaties and conventions between the United States and Great Britain terminates, the line of boundary between the territories of the United States and those of her Britannic Majesty shall be continued westward along the said forty-ninth parallel of north latitude to the middle of the channel which separates the continent from Vancouver's Island, and thence southerly through the middle of the said channel, and of Fuca's straits to the Pacific Ocean: Provided, however, That the navigation of the whole of the said channel and straits south of the forty-ninth parallel of north latitude remain free and open to both parties."
"From the point at which the forty-ninth parallel of north latitude shall be found to intersect the great northern branch of the Columbia river, the navigation of the said branch shall be free and open to the Hudson's Bay Company, and to all British subjects trading with the same, to the point where the said branch meets the main stream of the Columbia, and thence down the said main stream to the ocean, with free access into and through the said river or rivers, it being understood that all the usual portages along the line thus described shall in like manner be free and open. In navigating the said river or rivers, British subjects, with their goods and produce, shall be treated on the same footing as citizens of the United States; it being, however, always understood that nothing in this article shall be construed as preventing, or intended to prevent the Government of the United States from making any regulations respecting the navigation of the said river or rivers not inconsistent with the present treaty."
"In the future appropriation of the territory south of the forty-ninth parallel of north latitude, as provided in the first article of this treaty, the possessory rights of the Hudson's Bay Company, and of all British subjects who may be already in the occupation of land or other property lawfully acquired within the said territory, shall be respected."
"The farms, lands, and other property of every description belonging to the Puget's Sound Agricultural Company on the north side of the Columbia River, shall be confirmed to the said company. In case, however, the situation of those farms and lands should be considered by the United States to be of public and political importance, and the United States' Government should signify a desire to obtain possession of the whole, or of any part thereof, the property so required shall be transferred to the said Government, at a proper valuation, to be agreed upon between the parties."
In 1855, the Hell Gate Treaty between the US Government and the Flatheads was ratified. It officially established the Flathead Reservation but did not retroactively replace or alter the terms of the Treaty of 1846.
The following map, published in 1901, still shows the "Flathead Post" name attached to what was undoubtedly Fort Connah. In my opinion, this should settle the argument about whether there was not one, not two, but actually three "Flathead Post" operations. All three belonged to, and were operated by, the HBC during the period 1821–1871.

Flathead Post

There is a longstanding question about whether the renamed post was Flathead House or Flathead Post. In my opinion, the answer seems to be fairly simple. I believe the primary, underlying problem is mostly related to semantics. During the early years of the westward expansion of the fur trade industry across the Rocky Mountains, the terms, "House", "Post", "Store" and "Fort" were often used interchangeably to describe a permanent trading facility. This could explain the various names.
That fact has apparently been overlooked, ignored or misinterpreted by a lot of historians for years. I think it is rather simple. People who visited the post prior to 1821 knew it as Saleesh House. They would likely have been inclined, out of habit, to continue calling it Saleesh House, the NWCo name. On the other hand, newcomers to the region after 1821 probably called it either Flathead Post or Flathead House—the HBC name. That's simply human nature.
In 1871, the US finally ordered the HBC to cease operations on American soil. Fort Connah was the very last HBC operation in the United States to close its doors. The bottom line to this discussion of "Flathead House" issues and history is that the fur trading business David Thompson established in 1809 continued to be profitable long after his departure—lasting for 62 years.

Fort Connah

Fort Connah — Third Flathead Post.
Painting by Peter Peterson Tofft c. 1860s

The Fort Connah Restoration Society has been involved in restoring the old post for a long time. The Society opens it to the public during and annual fur trade rendezvous re-enactment. To find out more, visit their website at:

Copyright @2018, Rockman's Trading Post, Inc. All Rights Reserved

Last updated: 9/24/2020

"Ordinary history conveys information; extraordinary history conveys understanding!"
Stewart Edward White-1931