1809 Trip to McGillivray's River

Kullyspel House to McGillivray's River

David Thompson's 1809 travels from October 11th to October 20th


This OBServations paper traces David Thompson's travel from the Big Bend in McGillivray's (Kootanai) River and back as he explored this country for the first time in 1809. The maps below cover the areas discussed in this paper.



In 1809, David Thompson made his way down McGillivray's River to Bonners Ferry, ID. He camped at the mouth of Deep Creek located a short distance west of the town at 48° 42' 22" N, 116° 23' 04" W. One of Thompson's maps shows that to be the location of a North West Company (NWCo) post. However, he never mentioned trade at that place in his journal. The only reference I have found is one referring to it as "McDonald's House" likely named for Finan McDonald.

Saleesh House
Making his way overland with horses loaded with trade goods and supplies, he arrived on Kullyspelum Lake. Today we recognize it as Pend Oreille Lake located near Sandpoint, ID. Horses and help for the trip had been proffered by friendly Indians camped at the Lake.

The post was located on a peninsula opposite Memaloose Island at 48° 22' 07" N, 116° 17' 30" W. On the 23rd, Thompson and his crew had completed work on their "Store House". With winter coming on they immediately went to work on comfortable living quarters for the men.

On September 11, 1809, Thompson and his men were, at last, ready to begin construction of Kullyspel House. Tools were "helved", that is, handles were cut and shaped to fit items such as axes, saws and chisels. The post would soon be open for business.

With post construction going well, on September 27th, Thompson set out on a exploration trip down the Saleesh River (today's Pend Oreille River) hoping he would find a quick canoe route to the Columbia River. Unfortunately, the lower reaches of the Pend Orielle, a relatively short distance from the Columbia, turned out to be impassible. October 6th, he arrived back at Kullyspel House. After a couple days of rest and satisfied that things were going well with the construction, he decided to set out up the river to the east.

On October 11, accompanied by two of his men and an Indian guide, he left Kullyspel House for what would end up being a circuitous loop back to McGillivray's River. He would be following a well used Indian road which he referred to as the Saleesh Road to the Buffalo. On the 14th, he and his small party rode out onto what is today called Thompson's Prairie located a few miles east of the town of Thompson Falls, MT.

When Thompson crossed the Rocky Mountains in 1809, he was well ahead of his clerk, James McMillan, who was in command of several canoes loaded with men, trade goods and supplies for the new posts. His intention was to intercept McMillan at the Big Bend in McGilivray's River and guide him to the Saleesh River post. The new overland route he sought would have precluded traveling all the way back to Kullyspel House by way of Bonners Ferry, Deep Creek and Pack River to Pend Oreille Lake then back up the Saleesh River to get back to his October 14th campsite.

Always on the lookout for promising locations for a trading post, he camped near the mouth of today's Thompson River. Based on his journal notes we know that his camp was most likely located on the west side of the river about a quarter mile southwest of the Thompson River Bridge at approximately 47° 34' 22" N, 115° 14' 32" W. Within a month, following his return from McGillivray's River with James McMillan, construction of Saleesh House would be underway very near the old camp location. He had already decided it was a good spot.

Using his considerable skills as a surveyor, Thompson took the time to stop almost every clear day to take a sunshot with his sextant. Doing so allowed him to determine his location expressed in terms of Latitude and Longitude. Using this data as well as time, mileage and compass bearings along the way, he could calculate the approximate relative direction each observation was from those already recorded. In my book, Sometimes Only Horses to Eat, you can follow many of his routes in some detail based on information from his journals.

Likely using his data from sunshot observations, he determined that the Big Bend was located almost due north of his camp on Thompson's Prairie. Not familiar with the system of Indian roads in the country, Thompson hired a Indian guide to lead him back to the Big Bend. He also hired the man's son as a hunter to provide food along the way. As part of the deal, the young man was provided with Thompson's personal blanket, rifle and accoutrements such as powder, ball and patching. This was one business agreement that did not work out well for Thompson. As it turned out, the guide did not know the way back to the big bend in McGillivray's River. Consequently, he got them lost and ended up leading Thompson miles off course to the southeast.

David Thompson Saleesh House

The little party paused long enough for a sunshot near the old town of Weeksville located at 47° 31' 30" N, 114° 59' 39" W. Thompson knew that they should heading more to the north; the party was not headed in the right direction. Trusting his guide, Thompson followed. That night the men camped on the western edge of the "Horse Plains", near the town of Plains, MT. The next day the guide led them up the river to the southeast. Thompson was certain they were not headed in the right direction. What he didn't know at the time was that the trail they should have been on actually turned north at Weeksville!

Leading the party away from the Indian road, the guide headed out on a northeasterly bearing. In about two miles they encountered a "long Pond of Water". The general outline of the pond can be seen in this photo. Settlement, irrigation and streambed alteration have undoubtedly changed it's character but the swampy, mossy water is still home for a significant number of muskrats whose fur was valued during the fur trade period.

The guide led them to the southeast along the pond and after 45 minutes they ended back on the main Indian road! To say that Thompson was not happy would have been an understatement. He now realized that his guide was leading him farther and farther from the Big Bend and his supply canoes. But still he followed probably hoping they would soon encounter a main road to the north. He noted in his journal that:
"...set off N55E 2 M but went round a long Pond of Water & was obliged to return on our Road which our Guide did not remember by which we lost abt 3/4 of an Hour - havg found the Road we held on 'till 5:40 PM..."
From Horse Plains the Indian Road then swung to a northeasterly direction to a spot about a mile west of Dog (Rainbow) Lake where they camped for the night. The blue line is a best-guess of the likely route.

David Thompson Saleesh House
The red line indicates the bearing and distance, N55E 2 M, they followed until they encountered the long pond.

The following morning the party veered to the southeast following the road along Camas Creek onto today's Camas Prairie and continued along their southeasterly course. The blue line represents Thompson's route and lies within a narrow corridor where Linda and I think it was probably located. The traditional thought is that Thompson rode due east across Camas Prairie, turned north over the meadowed ridges to a campsite on Warm Springs Creek. The coordinates for that campsite is approximately 47° 25' 25" N, 117° 25' 25" W. Our research does not support that idea. Following is additional information that convinced us that we were finally on the right track and strongly supports the route shown above, or at least something very close to it. This area is located within the boundaries of the Flathead Reservation.

The Indian road likely followed Camas Creek out onto the Prairie then swung to the south to the Flathead River a short distance above the old town of Perma located at the intersection of Highways 200 and 382 located at 47° 22' 07" N, 114° 31' 35" W.

In the photo below, you will notice that Thompson's "likely route" terminates where it intersects a red line. The red line is based on a journal entry for October 16th regarding the general compass bearing they had been following.
"...went up & thro grassy Hills, then down them to a Rill of Water & stopped at 11:20 AM to refresh the Horses, havg large Plains & the rocky Mountains full in our View bearing N60E our Co has been I think N60E 2M - a bend of the River lies N60E from me the Mountains lie abt N40W & S40E forming a very wide & sublime View, they are loaded with Snow as in the depth of Winter - Obsd for Latde 66..35 (&) 1/2..."
David Thompson Saleesh House

The entry provides the bearing from what was probably his position on the ridgetop (located in the vicinity of 47° 26' 05' N, 114° 29' 35" W) to the unmistakable Big Bend on the Flathead River. He also noted the same straight-line bearing along their route prior to reaching the OBS location shown in the valley to the east. The location is identifiable since it is the place where the Little Bitterroot River veers to the east. The site is located on the Magera ranch (originally homesteaded in 1910 by the Melton family). A mile below the OBS location it empties into the Flathead River.

Saleesh House

The reference to the Big Bend of the Flathead River and the OBS locations are consistent with Thompson's journal notes and with his 84° map of 1814. A portion of the map covering the area is shown here (NOTE: the OBS location near the center-bottom of the map was taken after Saleesh House had been established and is of little consequence to the 1809 trip.) Several parts of this map are clearly inaccurate.

Keep in mind that Thompson had never been in this country. His Indian guide was lost so there was an urgency to keep moving if they were to intercept McMillan and his canoes at the Big Bend on McGillivray's River. Other than the Little Bitterroot River (he referred to it as the "Root Rivulet") which he followed north (upstream) from OBS 1. The only other part of the region he actually viewed lying to the north and east is the south end of Saleesh Lake on March 1st, 1812. This map appears to be an attempt to combine his recorded information as well as detail from memory from the two trips. The Flathead River can clearly be traced from where it leaves the lake to the vicinity of the OBS 1 location.

Thompson's map shows the Indian road cutting across to the Flathead River a few miles southwest of the outlet. In fact, there was a road that ran along the river to the outlet of the lake but it was not located properly on his map. His route is shown correctly.

It followed the Little Bitterroot River from the OBS 1 location to the point where Sullivan Creek empties into the Little Bitterroot. That location was noted in the journal. From there he followed Sullivan Creek to a point a little south of the settlement of Niarada. Wild Horse Island, on the west side of the lake at the town of Elmo, can be clearly identified in both the Google photo and on Thompson's map. The green line on the photo represents the approximate location of the Indian road to the lake.

McGillivray's River route

The valley where the road was located is open and flat. Highway 28 closely follows the same route as the Indian road. In my opinion, there is little doubt the map is not entirely correct along the route.

From Niarada the road swung to the east and basically followed the same route as Highway 28 to near Elmo where it intersects Highway 93. It then swung to a northerly direction eventually falling into the Flathead Valley in the near of the city of Kalispell. Note the obvious jog in the road shown on Thompson's map above a little south of OBS 2. This map shows what is likely that switchback in the existing county road a short distance northwest of Niarada.

McGillivray's River route
Our conclusions about the most likely location of Thompson's route were based on a combination of using journal notes as well as plotting where early roads appeared on old maps. We then interpolated the most likely travel route between those points based on landmarks such as ridgelines, streams and other prominent and/or easily identified features.

It is important to remember that simply stringing his bearings and distances together, each connecting to the next, is not what Thompson did. The course was, rather, a series of individual observations based on a landmark that could be easily seen and identified from several miles away. At the Highway 2 crossing, Thompson's route as noted in his journal was about a quarter mile east from where Linda and I believe the road crossed.

McGillivray's River route

Thompson continued his northwesterly course from the OBS 2 location and continued recording bearings and distances. His route is shown in BLUE.
The accompanying photos show the route where Linda and I think the trail was likely located. We have not found any remaining, identifiable traces of the actual road during our field trips in the area. However, General Land Office (GLO) maps from more than 100 years ago show segments of several early roads and trails in the area. Plotting his route using bearings and estimated distances as recorded in his journal shows it to be about one quarter mile east of where we think the road was most likely located.

The green line on the map below shows the road Thompson and his party should have taken from Weeksville that led up Buffalo Bill Creek, down the Little Thompson to Thompson River. From there the trail led north along the east bank to the present location of the Bend Ranger Station located a short distance south of the Thompson Lakes. Detailed information about the route can be found in my book, Sometimes Only Horses to Eat.

The road Thompson had been following was leading him in precisely the right direction to reach the Big Bend. However, had he turned west along the upper reaches of Fisher River (In his journal he referred to it as "Rock River") he would have fallen onto the road he should have been following. Instead he continued on the bearing he knew would get him to his destination on McGillvray's River. That decision resulted in having to cross the mountains dividing the Fisher River and Wolf Creek drainages.
"Octr 19...At 7:50 AM set off & went N25W 3 M passed a Lake & I think N25W 1M passed another Lake at end of which we got lost & after looking for the Road we found it...we then held on about N50W or N60W say 6 or 7M as the ground was hilly & much fallen Wood to 1 PM when we crossed a Brook of abt 5 yds [across] with abt 2 ft of water & stopped to refresh our Horses &c the Brook we have crossed comes from N40E by the Compass...we put up always keeping close along the Brook our Co has been S72W full 8M..."
In the Google Earth image below, note that Wolf Creek does, indeed, flow from the northeast to near the location where the Indian road crossed Wolf Creek. Kavalla Creek empties into Wolf Creek near the crossing.

Thompson was most probably on the Kavalla Creek trail that ends at the Wolf Creek crossing. His journal entries with regards to his route down Wolf Creek that same day appear to indicate that to be the case. According to his journal, he and his men arrived at Wolf Creek at 1PM and remained there until 2:30PM. The party then turned to a southwesterly bearing along the creek until until 5:15PM when they made camp near 48° 14' 40" N, 115° 13' 44" W, about two miles above the confluence of Wolf Creek and Fisher River.

David Thompson

On the morning of October 20th the party traveled another two miles on southwesterly bearings. At that point, about a mile east of the Wolf Creek-Fisher Creek confluence, the Indian road Thompson was on made a swing to the north. Evidently Thompson could not see Fisher River from the road as he turned north. In his journal entry for the day I think it is clear he thought he was still along Wolf Creek:
"...always along the Brook which has more than doubled its Water havg no less than 6 or 7 Rills from the eastd Hills..."
Years after the actual trip, during the 1840's, Thompson wrote and published a narrative account of his travels west of the Rockies. The strength of the narrative is that it was based on notes from his daily journal but for some of the information he clearly relied heavily on his memory. Trying to remember details about exactly what transpired on a particular day 30 years ago is, at best, questionable.
"...At 1 1/2PM set off & went on the same Course about 1M when thank God we came to the River - across the river we saw two Canoes & a Tent of Indians, upon calling they came & crossed all our Things &c - here we found also Forcier & Roberge who informed us Mr McMillan & Canoes were 3 Points below..."
The camp was probably on the west side of the river where the Army Corps of Engineer's campground is located. Thompson knew both men. They had been employed by the NWCo and had worked with Thompson west of the Rocky Mountains. According to Catherine White, Roberge had wintered with him at Kootenae House in 1808-1809 and Forcier was one of the men that accompanied Thompson across the Rocky Mountains via Athabaska Pass to Boat Encampment on the Columbia River during the winter of 1810-1811.

The information provided by Forcier and Roberge that McMillan was "...3 points below...", indicates he was probably in the vicinity of Libby, MT, about ten miles as the crow flies below the Big Bend.

Although Thompson was finally at the river, the surprises kept coming. Not only had his guide gotten them lost, costing them valuable time, but Thompson had also hired the guide's son as a hunter to provide meat for the party. That turned out to be a bad decision. Not only was the young man a poor hunter, he was also a thief. Deserting the party, he took with him some valuable items belonging to Thompson. The day following his arrival at the Big Bend, Thompson noted in his journal:
"Octr 21st Friday...Equipped the Kootenai Lad with my Gun & a Blanket with Amm for hunting with which he has I think deserted."
Thompson's trip from Kullyspel House to the Big Bend of McGillivray's (Kootenai) River to intercept his clerk and the canoe brigade had been completed. Now he was back in familiar country. From the Bend, Thompson rode downstream where he overtook and joined James McMillan and his canoes.



Continuing down the river the brigade landed at the mouth of Deep Creek located at 48° 42' 22" N, 116° 23' 03" W, where the trailhead for what Thompson called "the Lake Indian Road" was located. Thompson's map shows a NWCo post at that place. The spot is located about three miles west of Bonners Ferry, ID. There the men set up camp, sorted goods and rested while waiting for horses and men to help pack their trade goods and supplies for transport over the final leg of their journey back to Kullyspel House.Thompson arrived there on October 29th; 16 days after setting out on his first trip up the Saleesh River. It would not be his last.




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