By the spring of 1913, Silas Tanner had been drilling water wells in Iron County, Missouri, for eleven years, and he had never needed a permit to keep working after dark. That April, he was on the Rourke farm, a dry 160-acre spread two miles south of Ironton, where Otis Rourke had hired him to drill a well deep enough to survive a dry summer. The cable drill pounded through red clay, then gray dolomite, and at 71 feet it struck something that stopped the drill head completely. Tanner had grown up on that same cable drill, apprenticed under his father before taking his own contracts.
He knew rock, but after three days of trying, he still could not get through. At 80 feet, the drill head came up intact, scratched but not broken. The markings were straight, parallel lines evenly spaced, the kind of scarring a blade leaves on a flat surface, not the random wear of natural rock. When he lowered a second hole eleven feet to the east to check the depth, he hit the same surface at almost the same depth.
Rourke, who had farmed the land since 1894 and found nothing sharper than field stones in eighteen years of plowing, came to see for himself. He later told the county clerk that the sound of the drill head hitting that thing was not the sound of rock. It rang. Tanner tried three different drill heads over the next two days.
Each one came out scored with the same parallel pattern, each dulled more than a full week of normal drilling would have cost. A local blacksmith named Everett Hill, who sharpened Tanner’s drill heads that spring, told his grandson decades later that he had never seen tool steel come back worn smooth instead of chipped, like a plow blade wearing against packed clay, not a drill head against stone. Sheriff White Combs, 42, in his second term, had grown up in Pilot Knob, the son of a furnace worker who had labored in the Iron Mountain mines during the 1880s. He knew the mineral district better than most men who had never held a survey chain.
He drove to the Rourke farm with a deputy and spent about two hours at the wellhead before making his decision. Rourke later said Combs lowered himself into the well as far as the casing allowed, holding a lantern, staying long enough that the deputy called out twice to make sure he was all right. Combs ordered Tanner to seal the hole. He told Rourke the well should be moved 30 yards to the north.
According to the account Rourke gave a neighbor two decades later, Combs said the land under that corner of the property still fell within an unresolved old mining boundary dispute dating to the 1870s, and blasting inside a contested claim without clear title was an invitation to a lawsuit the county did not want to arbitrate. That explanation satisfied the county’s official procedures, but it did not satisfy Tanner. He later said he had drilled inside disputed mineral boundaries before, and no sheriff had ever come in person to stop a routine well. He sealed the opening as ordered, moved his equipment, and finished a second well 30 yards north, which hit ordinary dolomite at 44 feet and was unremarkable in every way.
The first hole was filled in with its own spoil within a week. Rourke never drilled another well near that corner, and when he sold the farm in 1931, he told the buyer to leave the northeast corner alone and not ask why. The Iron County records office kept a permit log during that period, a simple bound book where the sheriff’s office recorded every application, location, and outcome. The log still exists today in the county archives.
The entry for Tanner’s initial application, dated the second week of April 1913, is present and legible. The later entry recording Combs’s decision to refuse the permit and order the well sealed is missing. The page that should have contained it was removed rather than left blank, cut near the binding with a blade rather than torn out by hand. The log resumes with three later entries for an unrelated quarry permit from June.
No one in the records office has an explanation for the gap. The Missouri Bureau of Geology and Mines sent a field geologist to the area that same summer, a young man named Foster Latham who had spent two years mapping igneous outcrops in the St. Francois range. Latham’s field notebooks from 1913, preserved today in the state archives in Jefferson City, contain a brief entry dated late April noting a visit to the Rourke property southeast of Ironton regarding an abnormal blockage reported in a water well.
The entry has four lines. It records the depth, 79 and a half feet, and the bore diameter. It does not record what Latham saw when he looked into the hole, what he said to Rourke, or what recommendation he made to the bureau. The many pages following the same notebook entry were also removed, cut the same way as the Iron County log page, near the binding, leaving no indication of how many pages were lost.
Latham’s published bureau work over the next decade concerned manganese ore chemistry and never mentions the Rourke property, the well, or anything like it. He left the bureau in 1921 and spent the rest of his career teaching mineralogy at a small college in Kansas, a significant career step backward for a man on the verge of a regional supervisor post. No letter, resignation notice, or bureau memo explains the change. The Rourke well was not the only report of its kind from the St.
Francois region in that era. Two similar cases survive in the records, sharing enough with the Rourke well to suggest a pattern rather than coincidence. In 1896, a Reynolds County landowner named Amos Dell, drilling for water on a hill above the Black River about 20 miles southwest of the Rourke farm, found what his own account described as a floor of hard gray stone, smooth as the surface of a pond, which stopped his auger at 63 feet. Dell’s well was on land adjoining a manganese claim, and the claim owner paid to have the well filled and covered within a month rather than allow further drilling.
No geological report on Dell’s blockage was ever found. The second case comes from outside Missouri entirely, from the Baraboo Range in south-central Wisconsin, a formation similar to the St. Francois Mountains in one specific respect, being an ancient Precambrian island of hard purple quartzite about one and a half billion years old rising unexpectedly through much younger sedimentary layers. In 1889, well records from Sauk County, Wisconsin, documented a farmer near North Freedom finding, in the surveyor’s own words, a rock layer more regular than nature could make, at a little over 70 feet in ground where the surrounding geology would have predicted at least twice that depth to reach hard bedrock.
The Wisconsin file was never physically altered, unlike the Missouri record. It was simply never followed up. A single line in the county surveyor’s annual report notes the anomaly and recommends further study, which never happened, according to any surviving records. Each case shares a second feature with the Rourke well.
In all three, the obstacle was reached at roughly 40 to 60 percent of the depth local geology should have required to reach hard Precambrian bedrock. Modern drilling records held by the Missouri Department of Natural Resources confirm that the top of the Precambrian surface across the St. Francois region is genuinely irregular, rising and falling over a range of more than 400 feet within a few miles in some places. Unusually shallow elevations are not exceptional in themselves, but three unusually shallow elevations, each flat enough to be taken as a floor, each followed within weeks by an order to stop drilling, are hard to classify as separate geological accidents.
Two candidate hypotheses exist in geological literature to explain a surface like the one Tanner found. The first is natural. Rhyolite in the St. Francois range sometimes forms sills, sheet-like masses wedged between older rock layers, and an exposed sill might present a perfectly flat upper face after a billion years of erosion removed the softer surrounding rock.
The second is artificial and far more recent. The Pilot Knob Iron Company operated smelting furnaces and ore-processing floors four miles from the site for over 50 years. A buried furnace hearth or a slag floor, poured and leveled by iron workers in the nineteenth century, would explain both the flatness and the metallic ring under Tanner’s bit without requiring any explanation older than living memory. No sample from the Rourke well remains to settle the question either way.
The filled-in well was never reopened, and the only physical trace of what Tanner found was a fist-sized piece of gray stone said to have been kept on Rourke’s mantelpiece for the rest of his life, too small and too worn by the tool that struck it to have been examined for anything. Rourke’s granddaughter, in a 1981 interview with a regional folklore researcher, said the family still kept it in a cigar box, and that none of the three generations had ever taken it to a laboratory. That ambiguity itself is the most honest thing about the Rourke case. The record does not end with a revelation.
It ends with a gap in the exact place an answer should be. Sheriff Combs died in 1926, and no interview, letter, or memoir about the incident ever surfaced. Foster Latham died in Kansas in 1954, after 33 years of teaching mineralogy to undergraduates, never returning even once in any published or archived text to the four lines he wrote about the rock well as a young field geologist. Whatever Combs and Latham concluded as they stood at the wellhead in April 1913, they kept it to themselves.
The Rourke farm changed hands twice more after 1931, and according to county tax records, the current owner has held the property since 1997. No well has been drilled in the northeast corner since Tanner’s crew filled it in, and no survey team from the Missouri Department of Natural Resources has ever recorded a request to examine that land, not in the formal geological mapping contracts that followed, not during the 1960s campaign to reassess the region’s iron and manganese reserves, not since. The corner sits fenced for cattle, indistinguishable on the surface from the rest of the pasture, grass grown over a hole that for three generations of owners was only a rumor, with instructions simply to leave it alone.
Whether it covers a layer of ancient rhyolite exposed by billions of years of erosion, or a furnace floor built and abandoned within living memory, it has not been measured, sampled, or photographed since a field geologist from the Missouri Bureau of Geology and Mines climbed down a wooden ladder in April 1913 and wrote four lines that said nothing.