In 1700, the French fur-trader Pierre-Charles Le Sueur (1657-1704), while ascending the Mississippi River, described saltpeter caves along the west side of Lake Pepin, in Minnesota. This is the earliest record of cave saltpeter in America according to Shaw (1992, p. 52). Le Sueur’s Journal, under the dates September 10 to September 14, 1700, reports that:
In these regions, a league and a half to the northwest, there is a lake named “Pein” [Pepin] which is six leagues long and more than a league wide. It is bordered on the west by a chain of mountains; on the other side, to the east, there is a prairie, and to the northwest of the lake a second prairie two leagues long and wide. Nearby there is a chain of mountains which must be two hundred feet high and more than a half league in length. Many caves are found there in which bears hibernate in winter. Most of these caverns are more than forty feet deep and between three and four feet high. A few have very narrow entrances, and all of them contain saltpeter. It is dangerous to enter them in summer because they are filled with rattlesnakes, whose bite is very dangerous. M. Le Sueur saw some of these snakes that were six feet long, although usually they are only about four feet.
Le Sueur’s comments about the caves being inhabited by bears in winter and rattlesnakes in summer suggests that they were visited (by someone) throughout the year and presumably there would have been a reason for this.
Since the early seventeenth century there was an extended French presence in the Upper Mississippi Valley (UMV) that required gunpowder and the traditional assumption is that all of it was imported. But it’s also possible that French fur traders manufactured their own gunpowder, as they did in Missouri at a later date. Gunpowder is composed of 75 percent saltpeter (potassium nitrate) combined with sulfur (10%) and charcoal (15%). Charcoal is readily available in the wilderness, but where would the saltpeter and sulfur have come from?
Most American saltpeter caves are found in the southeastern United States, providing the vital ingredients for gunpowder during the War of 1812 and the Civil War. The exact location of Le Sueur’s saltpeter caves in Minnesota—far outside the classic saltpeter belt—had been a topic of speculation at least since the 1981 Saltpeter Symposium, summarized in a special issue of the NSS Bulletin (October 1981). But to this writer’s knowledge, no one had actively searched for these caves before the present endeavor.
Geological Detective Work
The location of Le Sueur’s saltpeter caves is not identified on historic maps. Franquelin’s 1697 map of the Upper Mississippi region, which had Le Sueur’s input, predates the cave visit, and although a five-part map of the Mississippi River was compiled by the famous French cartographer Guillaume Delisle from Le Sueur’s notes in 1702, and this map does label some mineral resources such as copper and lead mines, it does not depict the saltpeter caves (Fig. 1). Indeed, perhaps Le Sueur’s best-known exploit was the establishment of “Fort Green” at the site of a supposed copper mine on the Blue Earth River, in 1700, shortly after visiting the saltpeter caves.
At this point we fall back on the narrative. The “chain of mountains” mentioned in the quotation from Le Sueur above tallies reasonably well with former islands in the ancestral Mississippi River in Goodhue County, MN. This series of east-west ridges were once mesa-like islands in the river, huge slices of bluff cut off by stream erosion, composed of bedrock, and not to be confused with low, sandy islands such as nearby Prairie Island. Of the several possible Goodhue County “islands,” which are now part of the shore, the one that best matched Le Sueur’s narrative is the bluff towering over the 1856 logging ghost town of Sevastopol, east of the city of Red Wing, MN.
In 2004, a cave matching Le Sueur’s description, but somewhat shorter, was discovered by the author on private land along the Sevastopol bluffs (Fig. 2). While there is no way of knowing for certain whether this was one of Le Sueur’s caves, it became conceptually important by revealing that the caves he described match a particular type of bluff crevice found among the outcrops of Ordovician-age Oneota dolomite. They form where the bluff is slumping away from the main rock mass along a rock joint, leaving a long, narrow, bluff-parallel crevice. But tectonic caves 40 feet long, such as Le Sueur described, would represent lines of weakness in the bluffs. With this insight, it was surmised that the actual historic crevice(s) might no longer exist, having tumbled down the bluff at some point during the past 300 years, to be replaced by a subsequent generation of newly-widened rock joints as the Lake Pepin bluffs undergo erosion and slope retreat.
Chemical Fingerprints
When floor sediments or “petre dirt” from the Sevastopol crevice was tested it contained elevated levels of nitrate, probably in the form of calcium nitrate. This became the inspiration for a wider sampling program, involving caves along both sides of the Mississippi River as far south as the Shawnee Hills of Illinois, not merely those in the immediate area suggested by Le Sueur’s Journal. But the most heavily prospected area, because of the public accessibility of the rock outcrops and the abundance of voids, was Frontenac State Park in Minnesota.
Of the nearly one hundred caves sampled in the UMV, about two-thirds had elevated levels of nitrate. High nitrate concentrations, up to 35,000 parts per million (ppm)—equivalent to 3.5 weight-percent—were widespread among the Lake Pepin caves and farther south. By contrast, none of the surrounding surface soils showed more than 10 ppm nitrate, because nitrate is highly soluble and thus quickly leached from soil by rainwater, or absorbed by plants as a nutrient. For comparison, the nitrate content of cave sediments at Mammoth and Dixon caves, classic saltpeter mining localities, “range between 0.01% and 4% [by weight]” (Hill, 1981). Many of the Minnesota cave sediments are thus sufficiently enriched in nitrate to be worth extraction.
Although it’s theoretically possible to manufacture saltpeter from the UMV cave sediments based on the high nitrate concentration, a serious problem is the small amount of sediment available, especially considering that Kentucky cave sediments historically yielded only several pounds of saltpeter per bushel of petre dirt. Moreover, the Lake Pepin crevices are small, narrow, widely separated, and often difficult to get to, being located at the top of steep talus at a considerable elevation above the lake, so that even if a sufficient number of them were available, the amount of human labor involved would be enormous.
Any consideration of wilderness gunpowder manufacture should also address the possible sources of sulfur used in the gunpowder formula. In Europe, sulfur was often obtained from the volcanic deposits of southern Italy. In America, French miners in Missouri in the 1720s had access to saltpeter from caves, made charcoal from willow trees, and probably used sulfur that was generated as a by-product of smelting galena (lead sulfide), according to Breckenridge (1925). However, no local source of sulfur was identified for potential use in the Lake Pepin area. Sulfur plays a complex role in the gunpowder formulation but Rae and Whitehead (2006) concluded that “the sulphur and carbon could be replaced by other readily oxidized materials such as wood flour.”
Although no mining tools or indications of mining were observed in the UMV caves, nearby archeological sites have the potential to contain mining artifacts. In 1727, Fort Beauharnois was established, traditionally assumed to have been located at Sand Point near the town of Old Frontenac, but during a systematic archeological excavation by the Minnesota Historical Society in 1976, no French cultural remains were found. This unfortunately precluded the possible identification of potential gunpowder-related implements in artifact inventories.
Conclusion
The formation of high-nitrate Le Sueur-type caves can be summarized as follows: the long, narrow crevices form by the widening of bluff-parallel rock joints, where the bluff is falling away from the main rock mass. Over time, the crevices fill with sediment derived from surface soil descending through rock joints from above. Animal traffic and plant debris add nitrogenous organic matter that undergoes bacterial breakdown, forming nitrate, which then accumulates because the cave roof protects the sediment from leaching by rainwater and by shutting out sunlight, permits this plant nutrient to remain in the sediment.
The occurrence of high nitrate concentrations in the sediments of UMV caves revealed by this study is enough to show that Le Sueur’s claim of finding saltpeter (more likely, a saltpeter precursor such as calcium nitrate) in caves along the shores of Lake Pepin in 1700 for making gunpowder in the wilderness is credible, but practical considerations raise serious doubts. Nonetheless it’s the earliest record of cave saltpeter in America and pre-dates Carver’s Cave (1766) as the earliest Minnesota cave reference.
As late as World War I, American prospectors in search of nitrates to make explosives, sought out natural deposits in a tradition stretching all the way back to Le Sueur. The prospector’s days were numbered, however, owing to the development of atmospheric nitrogen fixation, which involved converting the abundant nitrogen of the atmosphere into ammonia, which could be converted to nitrate. In 1913 the Haber-Bosch process became the first really successful nitrogen fixation method, being put into commercial operation in Germany and then around the world. Modern agriculture, with its dependence on nitrogen fertilizers, and the world’s present large population, could not exist without this industrial process.
For the author’s full report with complete references see Brick (2012).
References
Breckenridge, W.C. (1925). Early gunpowder making in Missouri. Missouri Historical Review 20: 85-95.
Brick, G. (2012). Le Sueur’s saltpeter caves at Lake Pepin, Minnesota, and wilderness gunpowder manufacture. Minnesota Archaeologist 71: 7-20.
Hill, C.A. (1981). Origin of cave saltpeter. NSS Bulletin 43(4): 110-126.
Rae, I.D. and J.H. Whitehead (2006). Rackarock: On the Path from Black Powder to ANFO. IN B.J. Buchanan (ed.), Gunpowder, Explosives and the State, pp. 367-384. Aldershot, UK: Ashgate Publishing Ltd.
Shaw, T.R. (1992). History of Cave Science: The Exploration and Study of Limestone Caves, to 1900, Second Edition. New South Wales: Sydney Speleological Society.