Tuesday, March 15, 2011

The site of the Northwest company post on Sandy Lake

http://collections.mnhs.org/MNHistoryMagazine/articles/7/v07i04p311-325.pdf


The post of the Northwest Company on Sandy Lake in
Aitkin County was the first enduring establishment of its type
west of Fond du Lac on Lake Superior, and, from the date of
its erection in 1794 to the close of the period of British occupancy
of the region after the War of 1812, it was one of
the most important fur-trading stations in the Northwest.
The location on Sandy Lake was determined by its proximity
to the route between Lake Superior and the Mississippi by way
of the St. Louis and East Savanna rivers, the Savanna Portage,
the West Savanna and Prairie rivers, Sandy Lake, and
Sandy River. This route, which gave access tO' the great
regions lying south, west, and north, had probably been used
by the Indians for centuries before the advent of the white
man.^ It was by this route that Du Luth in the summer of 1679
"penetrated with his lively crew of voyageurs tO' the Sandy
Lake country, being probably the first white trader upon the
head-waters of the Mississippi," ^
Here in 1794, William Morrison, according to his brother
Allan, built " the original fort on Sandy Lake," Whether or
not the fort built by Morrison is the one which became the
Northwest Company post is unknown. He himself says that
he went into the country " in opposition to the old N. W. Co.,"
and that he "opposed all the N. W. posts until 1805."^ He
" found Bousquai at Sandy Lake." Charles Bousquet was
the trader of the Northwest Company at Sandy Lake between
1 Jacob V. Brower, " Prehistoric Man at the Headwaters of the Mississippi
River," in Minnesota HLHorical Collections, 8: 238.
2 Reuben G. Thwaites, " The Story of Chequamegon Bay," in Wisconsin
Historical Collections, 13: 407.
3 Grace L. Nute, ed., " The Diary of Martin McLeod," ante, 4: 384 n.;
Jacob V. Brower, Itasca State Park, 47 {Minnesota Historical Collections,
vol. 11).
311
312 IRVING H. HART DEC
1794 and 1797,* It is reasonably certain, however, that in the
year 1794 the Northwest Company erected a permanent establishment
here, which, probably from its form and strength,
came to be known as " the fort." °
Zebulon M. Pike gives the first detailed description of
the Northwest Company post on Sandy Lake. Having left the
main body of his expedition encamped upon the banks of the
Mississippi, Pike and one companion pushed on northeastward
to the station on Sandy Lake. They evidently missed
the old portage trail that led from the river to the lake, for
Pike writes, " we traversed about two leagues of a wilderness
. and at length struck the shore of Lake de Sable [Sandy
Lake], over a branch of which our course lay. The snow
having covered the trail made by the Frenchmen who had
passed before with the rackets [snowshoes], I was fearful of
losing ourselves on the lake. . . . Thinking that we could observe
the bank of the other shore, we kept a straight course,
some time after discovered lights, and on our arrival were
not a little surprised to find a large stockade. The gate being
open, we entered." °
Assuming that Pike had veered to the right of the portage
trail, he must have struck the lake shore on the south side of
Fisherman's Bay, which would form the " branch " mentioned
by him. What is now Brown's Bay was probably at that time
a small lake connected with the main lake only at high water.
A map of Sandy Lake in i860 shows such a small lake lying
southeastward of another somewhat larger body of water,
which is probably Bass Lake.'' Pike relates that the " fort
* Jean Baptiste Perrault, " Narrative of the Travels and Adventures of
a Merchant Voyageur," in Michigan Pioneer and Historical CoUections,
Z7 •• S04, S70, 573. 574-
5 William W. Folwell, A History of Minnesota, 1:68 (St. Paul, 1921).
" Zebulon M, Pike, Expeditions to the Headwaters of the Mississippi
River, 1:138, 139, 281 (Coues edition. New York, 1895),
^ This is one of the numerous manuscript maps among the Alfred J,
Hill Papers, in the possession of the Minnesota Historical Society,
1926 THE SANDY LAKE POST 313
at Sandy Lake is situated on the S, side, near the W, end,"
He " marks the site on his map, and gives it as ij4 m, S. of
the discharge of the lake into the short thoroughfare by which
this reaches the Mississippi." ° George Henry Monk, Jr., writing
in 1807, also locates the fort on the south side of the lake.°
\ .'Fun CoMfAMT POST
.Prawn 6r
MAP OF SANDY LAKE, SHOWING THE SITES OF THE POSTS OF
THE NORTHWEST AND AMERICAN FUR COMPANIES
Coues's statement that " The N. W. Co. house where Pike
was entertained stood on the W. shore of Sandy 1., next to
the Mississippi " is inaccurate. The editor of the Pike journals
made a canoe voyage to the source of the Mississippi at
8 Pike, Expeditions, i: 138 n., 281.
8 See ante, 5 : 36.
314 IRVING H. HART DEC
the time that the first government dam was being constructed
on Sandy River, and might, had he cared to do so, have definitely
located the site of the Northwest Company post. He
seems, however, to have depended for the location of this post
solely upon Pike's original map, which is drawn to a very
small scale and which in the nature of the case could not show
the location with any degree of accuracy. Coues's own " Historico-
Geographical Chart of the Upper Mississippi River " is
drawn to a scale only a little larger than that of Pike, but is
more accurate with regard to this location than is his statement."
After the treaty of 1783, which ended the Revolutionary
War and which endeavored to fix the boundary between the
possessions of Great Britain and those of the United States
in this region, and even after the withdrawal of the British
garrison from Mackinac in 1796 as a result of Jay's treaty,
the Northwest Company continued to occupy and exploit the
upper Mississippi Valley. It was not until after the War of
1812 and the treaty of Ghent that the jurisdiction of the United
States over the Sandy Lake region was definitely established.
This treaty and the passage by Congress in 1816 of
an act restricting the Indian trade to American citizens
brought to an end the activities of the Northwest Company
here. The company sold all its posts and outfits south of the
Canadian boundary to John Jacob Astor, who had previously
organized the American Fur Company." In 1820, when the
Cass expedition passed through Sandy Lake on its way to the
headwaters of the Mississippi, the agents of the American
Fur Company were found established in the old fort, which
Schoolcraft describes in almost the exact words used by Pike.
Between 1820 and 1832, when Schoolcraft for the second
"> Coues, in Pike, Expeditions, i: 138 n., 283 n.; and maps accompanying
volume 3 of the same work.
11 Folwell, Minnesota, i: 132, 133. See also Morrison's statement in
Brower, Itasca State Park, 47.
1926 THE SANDY LAKE POST 315
time visited Sandy Lake, the American Fur Company's post
was moved to a point just north of the mouth of Sandy River,
where it flows into the Mississippi, later the location of the
Libby post office.^^
Edmund F. Ely, a Congregational missionary to the Indians,
who was stationed at Sandy Lake in 1833 and 1834, refers
several times in his journal to the old fort, at that time
occupied by a man named Abbott, who had there a fur-trading
station in competition with that of the American Fur Company
at the mouth of Sandy River. Ely speaks of coming
back from Abbott's post at " the old fort " to the station of the
American Fur Company across the ice of the lake, and says,
" a strong N. W. wind in my face rendered it [walking]
quite tedious." "
In the year 1833, William Johnston, a representative of a
rival of the American Fur Company, writes of a visit to Sandy
Lake:
We arrived at the trading house of one of our clerks; it is
pleasantly situated on a point of land extending some distance into
the Lake. And the woods having been cleared, when it was
occupied by the North West Company gives it the appearance of
a White settlement; And it commands a view of the Savan river,
and the one which empties into the Mississippi; which is necessary
for a trading post, in order to watch the movements of the
opposition and Indians.^*
Johnston's reference to the site of the post is somewhat obscure.
There is only one " point of land extending some
distance into the lake " which commands a view of the river
emptying from the lake into the Mississippi, and that is
12 Henry R. Schoolcraft, Narrative Journal of Travels througli the
Northwestern Regions of the United States to the Sources of the Mississippi
River, 218 {Albany, 1821) ; Coues, in Pike, Expeditions, 1 : 138 n.
13 Ely Diaries, October 12, November 22, December 7, December 24,
1833. The originals of these diaries are in the collection of the St. Louis
County Historical Society at Duluth; copies are in the possession of the
Minnesota Historical Society.
^^ Michigan Pioneer and Historical Collections, 37:173.
3i6 IRVING H. HART DEC
Brown's Point. There is no point on the lake which commands
a view of both the Savanna River and the Sandy River.
Johnston's statement can be reconciled to local geography only
by omitting the " and " following " Savan river " and assuming
that by the " Savan " river he means Sandy river. Johnston
nowhere mentions the name of the clerk of his company on
Sandy Lake, but he was probably the Abbott mentioned by
Ely.
No significant references have been found to the location
of the old Northwest Company post between 1833 and 1894.
In the latter year Brower wrote, " On the south shore of Sandy
Lake are visible the old landmarks of the trading post and
station of a hundred years ago which Lieut. Z. M. Pike so
carefully described, . . . It is now an abandoned waste, soon
to be obliterated farther by the flood from the government
reservoir dam about to be completed." ^°
Within a comparatively few years following Abbott's occupation
of the old fort, the site seems to have been abandoned
entirely by the trading companies, and, according to
local tradition, the few available maps, and the more reliable
evidence of findings all along the north shofe of Brown's
Point, the old post and its surrounding clearing became again
the site of an Indian village.^° The Hill Papers contain a map
1^ Brower, in Minnesota Historical Collections, 8: 238. The first government
dam on Sandy Lake was in the process of construction at the time
that Brower wrote. As a matter of fact, neither this dam nor the second
dam, which is now standing, served to " obliterate " the site of the old
fort, as it lies entirely above the high water mark. At times of high
water in the Mississippi, Sandy River sometimes flowed back into the
lake, causing great floods which covered the country for many miles
around. Ordinarily, however, it would seem that the level of water in
the lake was considerably lower than at present. See Newton H. Winchell
and Warren Upham, The Geology of Minnesota, 54 (Geological and
Natural History Survey of Minnesota, Final Report, vol. i — Minneapolis,
1881).
1' Brown's Point had evidently been occupied by a Chippewa village at
a much earlier date. Warren's detailed account of the battle of Sandy
Lake between the Chippewa and the invading Sioux clearly indicates that
1926 THE SANDY LAKE POST 317
drawn in 1886 which shows the location of the Northwest
Company's post on Sandy Lake. It is based upon data furnished
by Ely from a map drawn for him on the ground by
the Sandy Lake Indians in i860. There are two obvious historical
errors on this map, one of which Hill has noted and
corrected. The station at the mouth of Sandy River is labeled
" Site of old N, W, Fur Co,'s post or Aitkin's trading
post in 1832," whereas there is no evidence that the Northwest
Company ever had a post on this spot, Aitken's post
was that of the American Fur Company, occupied after the
old fort on the south shore was abandoned by the company.
About halfway up the point north of Fisherman's Bay on the
same map is a spot marked, " Fort according to E, F, Ely
here (i860) but evidently wrong 1" Hill has corrected the error
in the location of the Northwest Company post by marking
a location on what is now called Brown's Point with the
notation, " Suppositious site of fort." This map also locates
an " Indian village in i860 " near the " suppositious site " of
the fort. From the evidence of this map it might be thought
that Ely was unaware that the site of Abbott's post was the
same as that of the original Northwest Company station of
1794, It is more probable, however, that his indication of the
location of the post was inaccurate or was misunderstood by
Hill. A small map drawn by Lieutenant James Allen, a member
of the Schoolcraft expedition of 1832, confirms the locations
of the two posts. Allen marks the post at the mouth of
Sandy River, " Trading H of A. F, Co." and that on Brown's
Point, " Old Trading House." "
this point was the site of that memorable encounter. See William W.
Warren, " History of the Ojibways," in Minnesota Historical Collections,
S: 177, 225-234. At the present time, rice holes dot almost the entire
central and eastern portions of the point. Numerous graves may be
located by rectangular depressions on the ridges at the western end. Excavations
for cabin sites near-by have opened up many of these with the usual
revelation of Indian relics.
I'' Allen's map appears with his report in American State Papers: Military
Affairs, 5: 313.
3i8 IRVING H. HART DEC
In 1900 the point on which the Northwest Company post
was situated was purchased by Thomas Edward Brown, according
to his daughter, Mrs. Jessie Brown Cleaves. Plis log
DIAGRAM OF THE STOCKADE AND POST OF THE NORTHWEST
COMPANY ON SANDY LAKE
[Drawn from the description of Lieutenant Zebulon M. Pike, by
W. P. Ingersoll. The buildings inside of the stockade were probably
located somewhat closer to the walls than is indicated on this diagram.
The two cellars completely excavated were located underneath and
south of the residence of the superintendent. The fireplace may have
been on the east rather than on the south of this building. The remains
of stockade posts were found at the northeast corner and extending
westward and southward from this point. Remains of corner
sills at the northwest angle of this building were found together with
other remains of sills extending south and east. Excavation beneath
the storehouse in the southwest corner of the stockade area showed
the presence of a filled-in cellar. The remains of a stoned-up well
were found at the location indicated on the diagram.]
1926 THE SANDY LAKE POST 319
cabin stood there until four years ago. Before Brown's death
the north shore of the point was subdivided into lots, which
have since been sold for cabin sites. Today it has the appearance
of the usual tourist resort. Mr, WilHam P, Ingersoll,
at present a resident on the point, says that Mr, Norbert
Ohrenberg, who worked for Brown while he was clearing his
land for cultivation, stated that innumerable evidences of Indian
occupancy were turned up by the plow. Stone hammers,
tomahawks, and other indestructible articles were heaped up
and hauled away, while bones and all combustible materials
were piled up and burned. Stone walls or foundations were
also said by Mr. Ohrenberg to have been found, but the locations
of these have been forgotten.
The task, therefore, of determining with any degree of accuracy
the location of the old Northwest Company post was
one for which exact data were not available. The descriptions
of the fort by Pike and Schoolcraft were the most definite
and helpful. The Ely Diary and the Hill maps served to
confirm the Pike data. Fortunately one of the Hill maps
shows the location of the " suppositious site " of the fort with
reference to the government survey, marking it on the mainland
southwest of the northwest corner of section i, township
49 north, range 24 west, which description locates the fort on
Brown's Point beyond the possibility of question.
Pike says that the fort consisted of " a stockade 100 feet
square, with bastions [blockhouses] at the S. E. and N. W.
angles, pierced for small-arms." Pike's description continues
:
The pickets are squared on the outside, round within, about one
foot diameter, and 13 feet above ground. There are three gates:
the principal one fronts the lake on the N., and is 10 x 9 feet;
the one on the W. 6 x 4 feet; and the one on the E. 6 x 5 feet. As
you enter by the main gate you have on the left a building of one
story, 20 feet square, the residence of the superintendent. Opposite
this house on the left of the E, gate, is a house 25 x 15 feet,
the quarters of the men. On entering the W, gate you find the
320 IRVING H. HART DEC
storehouse on the right, 30 x 20 feet, and on your left a building
40 x 20 feet, which contains rooms for clerks, a workshop, and
provision store.
On the W. and N. W. is a picketed inclosure of about four
acres, in which last year they raised 400 bushels of Irish potatoes,
cultivating no other vegetables. In this inclosure is a very ingeniously
constructed vault to contain potatoes, and which likewise
has secret apartments to conceal liquors, dry goods, etc.^*
To this description, Schoolcraft adds no significant details
other than that the pickets of the main stockade were of pitch
pine, and were " pinned together with stout plates of the same
wood." *=
The fact that several depressions, much larger than those
marking the old Indian rice holes and evidently showing the
locations of old cellars, were still visible on lots 16 and 17 of
the Sandy Lake Beach subdivision on Brown's Point, led the
investigators to begin excavations on this site under Mr. Ingersoll's
direction in August, 1926. The first spadeful of earth
turned up yielded an Indian medicine man's carved " swallowing-
bone," thus encouraging further search. Within a few
minutes unmistakable evidences of white man's work were
uncovered in the form of rotting cedar sills and the shredded
remnants of vertical posts, located at varying intervals and
covered by from fourteen to eighteen inches of sand, char,
and soil. Further excavations at different places near-by
showed without exception the presence of a well-defined stratum
of charred wood, as though a comparatively large conflagration
had at some time taken place. This layer of char
lies at a depth of from twelve to fifteen inches below the present
irregular surface.
The place at which the first excavations were made is about
thirty feet south of the beginning of the slope leading to the
edge of the lake. The finding of remains of vertical posts at
intervals on lines running west and south and the presence
18 Pike, Expeditions, i: 281.
1° Schoolcraft, Narrative Journal, ai8.
1926 THE SANDY LAKE POST 321
within the angle formed by these lines of a large depression
make it seem possible that this was the northeast corner of
the old stockade. In this case the depression may be what
is left of a cellar beneath the superintendent's residence. From
this northeast corner southward for some distance in extension
of the line on which remains of vertical posts were found
there is a ridge of earth one and one-half to two feet in
height, such as might have accumulated along the east line of
the stockade. Approximately one hundred feet to the west
a similar ridge of earth runs southward parallel tO' the first,
and this may be the line of the west side of the stockade. Some
thirty feet to the east of this west ridge, underneath a cabin
belonging to Mr. Enoch Johnson of Palisade, were found,
during the process of the erection of the cabin by Mr. Ingersoll,
the remains of what had evidently been a stoned-up well.
The location of this well within the putative outlines of the
stockade would place it east of the workshop and provision
store mentioned by Pike. Immediately back of the Johnson
cabin is another relatively large depression, possibly the site
of a cellar underneath the storehouse. Excavation here to a
depth of three or four feet showed a sandy loam which had
plainly been deposited at a comparatively late date and the
omnipresent layer of char. Nothing else was found that was
significant.
The site of the larger depression in the northeast corner
was excavated with exceeding care and for its entire extent
with very satisfactory results. The approximate dimensions
of the depression were fifteen by thirty feet, the longer dimension
extending north and south. A low ridge of earth
was to be seen running directly across the depression about
twelve feet from the northern end. The assumption that this
ridge marked a cross wall in the cellar or a partition between
two cellars was confirmed by the excavations, as beneath it
were found in their original position horizontal cedar poles
from four to six inches in diameter laid up to form a wall.
322 IRVING H. HART DEC
The poles were, however, so rotten that in spite of the greatest
care they crumbled away or caved in after a few minutes'
exposure to the air. At several places along the side walls of
the cellars, fragmentary remains of similar poles were found,
but nowhere else in such quantity or in such regular position.
Several of the pieces of timber were partially charred by fire.
IDEAL SKETCH, OF THE NORTHWEST COMPANY POST ON SANDY
LAKE, LOOKING SOUTHEAST
[Drawn from the description of Lieutenant Pike, by W. P. Ingersoll.]
A vertical cross section of the excavation in the smaller
cellar showed a stratum of sandy loam at the surface underlaid
successively by layers of char, ash, and sand, intermingled
with small boulders and bits of baked clay, some of the latter
showing still the impression of the stones which they had
joined. Underneath these was a stratum of fish scale from
two to four inches in thickness covering the entire area of the
compartment; and at the bottom, resting upon the undisturbed
1926 THE SANDY LAKE POST 323
native blue clay, was a layer of miscellaneous debris of neither
common composition nor consistency. Excavation to the
outer walls of undisturbed sand and clay showed that the
original dimensions of the two compartments had been twelve
by ten feet and twelve by twenty-four feet, respectively, with
a common depth beneath the present surface of approximately
four feet.
Every shovelful of earth was examined as it was removed,
with the result of the discovery of gunflints, the trigger guard
and a part of the breechblock of an old flintlock gun, a light ax
of the type locally known as the " Hudson Bay" ax, a handmade
two-edged knife or dagger, a handmade metal lock, an
old metal door latch, many pieces of heavy old glass bottles
and demijohns, pieces of figured porcelain dishes, bits of mortar
and melted glass, and many scattered remnants of charred
and rotted timbers, all plainly evidences of the white man's
occupancy of the site many years ago. Few relics of Indian
origin were found during this excavation, although at other
places within the limits of the supposed area of the stockade
were discovered a carved " swallowing-bone," a hollow bone
ornament, a copper bracelet, hand-carved beads, and many
pieces of the blue and green figured porcelain of the trade
patterns common in this region. The Indian relics were generally
found just beneath the sod, while the evidences of white
man's occupancy were found buried often two feet or more
beneath the surface. No pottery remains whatsoever were
found, although fragments of potsherds are tOi be found in
quantities elsewhere on the islands of Sandy Lake and the
mainland. The failure to find any pottery here would seem
to indicate that the Indian occupancy of which these relics
are evidence occurred at a time subsequent to that of the white
traders.
About midway of the line which marks the beginning of
the slope from the north side of the site to the present shore
line of the lake, there are traces of what may well be the re324
IRVING H. HART DEC
mains of a gradual approach to the main gate of the stockade.
The site lies about a mile and a quarter south and a little east
of the entrance of the lake into Sandy River. To the west
lies a comparatively level tract of land upon which by actual
measurement a square four-acre garden might be laid out.
SANDY LAKE
SITE OF THE NORTHWEST COMPANY POST AND GARDENS ON
SANDY LAKE
Beyond this level tract rises a low ridge composed of boulders
and coarse gravel, succeeded in turn by a little valley and a
higher and heavily wooded ridge, an extension of those which
border the western shore of the lake.
1926 THE SANDY LAKE POST 325
The conclusion that these findings serve to determine the
site of the old fort has been reached only after the most careful
consideration of all the available data and of various conflicting
local traditions with regard to the location. No' other
place on Brown's Point satisfies all conditions set up by the
descriptions of Pike, Schoolcraft, and Johnston. This site
does.
It would seem probable that, after the final abandonment of
the fort, it was burned, the ashes and charred remains of the
conflagration forming the stratum of ash and char underlying
the whole area. The cellars were gradually filled with
soil washings and debris, and soon became mere depressions.
Timbers and lumber that escaped destruction and remained
exposed were doubtless used for firewood. The Indians reoccupied
the point, erected their tepees, dug their rice holes,
buried their dead, and used the clearing for their gardens,
thus giving to this area the name " Indian Gardens " by which
it was known to the early settlers. Second-growth timber
sprang up, and the site became, with the gradual decay and
removal of the Indian population, the " abandoned waste " of
which Brower wrote in 1894.
A temporary marker will be placed upon the site, and if,
after further study and investigation, the conclusions of the
present investigators are confirmed by adequate authority, a
permanent monument should be erected to mark one of the
most interesting and significant spots in all the great Northwest.
IRVING HARLOW HART
IOWA STATE TEACHERS COLLEGE
CEDAR FAIXS, IOWA
Copyright of Minnesota History is the property of the Minnesota
Historical Society and its content may not be copied or emailed to
multiple sites or posted to a listserv without the copyright holder’s
express written permission. Users may print, download, or email
articles, however, for individual use.
To request permission for educational or commercial use, contact us.
www.mnhs.org/mnhistory

No comments: