Arrived at Assiniboine House, he found it in charge of John McDonnell, brother of the well-known Miles McDonnell, who, a few years later, became Lord Selkirk's first governor on Red River. Ensconcing himself in the comfortable quarters at Assiniboine House, Thompson wrote up in ink his journals, maps, astronomical observations, and sketches which he had taken in crayon, thus giving them more permanent form. He had now been in the employ of the North-West Company a full year, and in that time had been fully gratified by the work he had done and by the cordial reception given him in all the forts to which he had gone.
Assiniboine House, or, as he called it, Stone Indian House, was found to be a congenial spot. It was on the north side of the Assiniboine River, not far from where the Souris River empties its waters into the larger stream, though the site has been disputed.
One of the astronomer's clearly defined directions was to visit the Mandan villages on the Missouri River. He was now at the point when this could be accomplished, although the time chosen by him, just as winter was coming on, was most unsuitable. His journey reminds us of that made by Verendrye to the Mandans in 1738.
The journey was carefully prepared for. With the characteristic shrewdness of the North-West Company, it was so planned as to require little expenditure. Thompson was to be accompanied chiefly by free-traders, i.e. by men to whom certain quantities of goods would be advanced by the Company. By the profits of this trade expenses would be met. The guide and interpreter was René Jussaume (a man of very doubtful character), who had fallen into the ways of the Western Indians. He had lived for years among the Mandans, and spoke their language. Another free-trader, Hugh McCracken, an Irishman, also knew the Mandan country, while several French Canadians, with Brossman, the astronomer's servant man, made up the company. Each of the traders took a credit from Mr. McDonnell of from forty to fifty skins in goods. Ammunition, tobacco, and trinkets, to pay expenses, were provided, and Thompson was supplied with two horses, and his chief trader, Jussaume, with one. The men had their own dogs to the number of thirty, and these drew goods on small sleds. Crossing the Assiniboine, the party started south-westward, and continued their journey for thirty-three days, with the thermometer almost always below zero and reaching at times 36 deg. below. The journey was a most dangerous and trying one and covered 280 miles. Thompson found that some Hudson's Bay traders had already made flying visits to the Mandans. On his return, Thompson's itinerary was, from the Missouri till he reached the angle of the Souris River, seventy miles, where he found abundant wood and shelter, and then to the south end of Turtle Mountain, fourteen miles. Leaving Turtle Mountain, his next station was twenty-four miles distant at a point on the Souris where an outpost of Assiniboine House, known as Ash House, had been established. Another journey of forty-five miles brought the expedition back to the hospitable shelter of Mr. McDonnell at Stone Indian House. Thompson now calculated the position of this comfortable fort and found it to be 49 deg. 41´ (nearly) N. latitude and 101 deg. 1´ 4´´ (nearly) W. longitude.
The astronomer, after spending a few weeks in making up his notes and surveys, determined to go eastward and undertake the survey of the Red River. On February 26th, 1798, he started with three French Canadians and an Indian guide. Six dogs drew three sleds laden with baggage and provisions. The company soon reached the sand hills, then called the Manitou Hills, from some supposed supernatural agency in their neighbourhood. Sometimes on the ice, and at other times on the north shore of the Assiniboine to avoid the bends of the river, the party went, experiencing much difficulty from the depth of the snow. At length, after journeying ten days over the distance of 169 miles, the junction of the Assiniboine and Red River, at the point where now stands the city of Winnipeg, was reached. There was no trading post here at the time. It seems somewhat surprising that what became the chief trading centre of the company, Fort Garry, during the first half of this century should, up to the end of the former century, not have been taken possession of by any of the three competing fur companies.
Losing no time, Thompson began, on March 7th, the survey, and going southward over an unbroken trail, with the snow three feet deep, reached in seven days Pembina Post, then under the charge of a leading French trader of the company, named Charles Chaboillez. Wearied with a journey of some sixty-four miles, which had, from the bad road, taken seven days, Thompson enjoyed the kind shelter of Pembina House for six days. This house was near the forty-ninth parallel and was one of the especial points he had been appointed to determine. He found Pembina House to be in latitude 48 deg. 58´ 24´´ N., so that it was by a very short distance on the south side of the boundary line. Thompson marked the boundary, so that the trading post might be removed, when necessary, to the north side of the line. A few years later, the observation taken by Thompson was confirmed by Major Long on his expedition of 1823, but the final settlement of where the line falls was not made till the time of the boundary commission of 1872.
Pushing southward in March, the astronomer ascended Red River to the trading post known as Upper Red River, near where the town of Grand Forks, North Dakota, stands to-day. Here he found J. Baptiste Cadot, probably the son of the veteran master of Sault Ste. Marie, who so long clung to the flag of the Golden Lilies.
Thompson now determined to survey what had been an object of much interest, the lake which was the source of the great River Mississippi. To do this had been laid upon him in his instructions from the North-West Company. Making a détour from Grand Forks, in order to avoid the ice on the Red Lake River, he struck the upper waters of that river, and followed the banks until he reached Red Lake in what is now North-Eastern Minnesota. Leaving this lake, he made a portage of six miles to Turtle Lake, and four days later reached the point considered by him to be the source of the Mississippi. Turtle Lake, at the time of the treaty of 1783, was supposed to be further north than the north-west angle of the Lake of the Woods. This arose, Thompson tells us, from the voyageurs counting a pipe to a league, at the end of which time it was the fur-traders' custom to take a rest. Each pipe, that is, the length of time taken to smoke a pipe, however, was nearer two miles than three, so that the head waters of the Mississippi had been counted 128 miles further north than Thompson found them to be. It is to be noted, however, that the Astronomer Thompson was wrong in making Turtle Lake the source of the Mississippi. The accredited source of the Mississippi was discovered, as we shall afterwards see, in July, 1832, to be Lake Itasca, which lies about half a degree south-west of Turtle Lake.
Thompson next visited Red Cedar Lake, in the direction of Lake Superior. Here he found a North-West trading house, Upper Red Cedar House, under the command of a partner, John Sayer, whose half-blood son afterward figured in Red River history. He found that Sayer and his men passed the winter on wild rice and maple sugar as their only food.