An Iron Ore Deposit in the Potsdam Sandstone of Eastern Ontario: The Hematite Deposit Between Delta and Furnace Falls

An early report that describes the rocks of Lanark County is the Report on the Geology of a Portion of Eastern Ontario by R.W. Ells that was published in 1904 by the Geological Survey of Canada. The report covers the northern three quarters of Lanark County and covers adjacent parts of Renfrew, Addington, Frontenac and Carleton counties. In his report Ells traces the Potsdam sandstones from Renfrew down to south of Perth. One part in the report that caught my eye included comments on iron ore deposits in the Potsdam sandstones. Ells commented:
"While the sandstones as a rule are grayish or sometimes yellowish; in places they are often coloured red through the presence of hematite. This sometimes assumes the form of ore deposits of this mineral, capable of being mined. These deposits are usually in the form of irregular pockets and vary in quality from the impregnated sandstones, in which the siliceous matter predominates to an iron ore of considerable purity."I found that intriguing because, while I am aware of numerous former iron ore mines in the Precambrian Shield of Eastern Ontario, I could not recall anyone mining iron ore in Potsdam sandstone. Unfortunately, the map that accompanied Ells’ report does not show the location of the hematite deposits.

I decided to do further research on the matter and started by trying to determine the source for the iron ore that had been smelted at Furnace Falls, Ontario. I chose this location because Furnace Falls, which is now known as Lyndhurst, is the location for Ontario’s first iron ore smelter and because directly to the north of Lyndhurst (and south of Delta, Ontario) are many dark red beds of Potsdam sandstone (and it is known that the red colouring of the sandstone is due to iron oxide). Lyndhurst is located about 25 kilometers south of Lanark County in the United Counties of Leeds and Grenville. I struck gold (or, more accurately, iron), as the hematite that was smelted at Furnace Falls was mined from Potsdam sandstone beds approximately five kilometers (three miles) north of Lyndhurst (and one kilometer south of Delta, Ontario).

An historical plaque on a rock cairn in Lyndhurst provides a brief history of the smelter.

The plaque states:
"Lansdowne Iron Works
Forges Lansdowne
While the existence of local ore was well known and various petitions had been made to erect a foundry, it was not until 1801 that Wallis Sunderlin, a Vermont founderer, established the first iron works in Upper Canada at Furnace Falls. The works, which included both a furnace for the production of cast iron and a forge for the manufacture of wrought iron, were operated with limited success by Sunderlin and his associates until destroyed by fire in 1811. Attempts in 1815-16 to re-establish the works to supply the Kingston dockyard were ended with the agreement to limit armaments on the Great Lakes."

A newsletter published in 2004 by the Leeds and Thousand Islands Historical Society provides additional information on the iron smelter: "By 1795, it was known that the big falls on the Gananoque river had all the resources needed to develop an iron smelter. Wallis Sunderlin, an ironmaster from Tinmouth, Vermont, received assent from the Executive Council in 1800. The Lansdowne Iron Works was completed in 1802 being the first in Upper Canada. Sunderlin was granted 1200 acres to supply fuel for the blast furnace. The Furnace produced domestic cast iron items, pig iron, and wrought iron. ... In 1811 the smelter and mill complex was destroyed by fire. Sunderlin died that year and his family moved back to the United States... With the outbreak of war in 1812 no effort was made to rebuild the furnace."

A number of publications of the Geological Survey of Canada and the Ontario Department of Mines confirm that the iron ore that was smelted at Furnace Falls came from hematite deposits in Potsdam sandstone mined from Lot 25 (and possibly Lot 24) in the tenth concession of Bastard Township (near the township line with Lansdowne township), between Delta and Furnace Falls.  Murray (1852) of the Geological Survey of Canada provides an early description of the ore deposit:

"The cliff of ferruginous sandstone, which occurs on the twenty-fifth lot of the tenth concession of Bastard, displaying a vertical height of about thirty feet, brown in the lower and deep red in the upper part, owes its color to the presence of peroxide of iron, which is mingled with the siliceous grains, apparently cementing them together, and sometimes being pulverent, staining the fingers with a red shining powder. In a three feet bed, which occurs within about three feet of the top, the oxide passes into the form of strongly coherent scaly red iron ore, in which thin seams and spangles of crystalline specular iron ore occur. The parts so marked run in layers in the bed, and alternate with layers of the sandstone of a yellow and less ferruginous character. The concentration of the ore is greatest towards the middle of the bed, where nodules and patches of pure red hematite, running with the stratification, occur at intervals of a few inches, the thickness they display not exceeding a couple of inches. About forty years since an attempt was made to mine the ore for the supply of a furnace erected at Furnace Falls, but the quantity in the locality worked was not sufficient to give a profitable result."

From Murray’s description it would appear that there is a well defined stratigraphic control for the ore body and, accordingly, that the iron deposit is the same age as the host sandstone.

The deposit appears to have been fairly extensive. Murray mentions that the cliff in Lot 25 extends into Lot 24, that similar rocks that contain small seams, patches and streaks of specular iron can be found in Lot 23 of Concession X of Bastard Township, in the adjacent ninth lot of the twelfth concession of Lansdowne Township and in the 9^th Concession of Lansdowne Township.

A little over a century after the smelter at Furnace Falls was destroyed the hematite deposit was tested for ore. The Ontario Department of Mines, and J.F. Wright of the Geological Survey of Canada, provide details of the testing of the property. Wright (1921) comments:
"From October 1918 Drainey Brothers of Toronto worked these deposits, except in winter, until November 1919. In August 1919 the Consolidated Iron and Steel Corporation took over the property. Three small shafts and some prospect pits were sunk and four carloads of ore were shipped, which according to the smelter records, averaged 68 per cent iron. About one carload of ore was left on the dump." The Ontario Department of Mines’ Twenty-ninth Annual Report (1920) mentions that The Consolidated Iron and Steel Corporation, Limited "had an option on some 600 acres in Leeds County about half way between Delta and Lyndhurst Stations on the Brockville and Westport railway in concession X of the township of Bastard and the adjoining portions of the township of Lansdowne." However, nothing appears to have come of Consolidated Iron’s testing of the deposit in 1919 and the shipment of four railroad cars of hematite. Four years later when M. B. Baker (1923) reported on the geology of Leeds County for the Ontario Department of Mines he commented that "a number of pits were opened up, but shipping ore was not found in any appreciable tonnage."

Wright (1921) provided a map of the deposit (see below) and additional details on the deposit, which he placed in Lot 23, Concession X, Bastard Township.


Shaft No. 1 (16 feet). The ore occurred in irregular shaped pockets. The ore is either massive hematite mixed with specular hematite, siderite and calcite or sandstone impregnated with hematite.

Shaft No. 2 (20 feet). The ore occurs along a fracture zone along which the sandstone is impregnated with hematite, and as veins of hematite, from 1 to 2 inches wide.

Shaft No. 3 (15 feet) . The ore body consists of two irregularly shaped veins of massive hematite from 3 to 6 inches wide.

Wright proposed the following origin for the ore: "The iron minerals appear to be due to solutions circulating along joint planes or fracture zones in the sandstone. The most likely source for the ore minerals is the weathering of overlying sedimentary formations containing iron, probably in the form of carbonate. ... The iron was probably carried down as ferrous carbonate and after or during precipitation was oxidized to the ferric oxide hematite. The solutions filled all the open spaces, spread into the sandstone, and replaced the calcareous cement but only slightly the quartz."

I mentioned above that Murray (1852) described a well defined stratigraphic control for the ore body–an ore zone that is a three foot bed three feet from the top of a 30 foot cliff of sandstone– and stated that the hematite runs with the stratification, alternating with layers of the sandstone. Wright’s (1921) description of the deposit (namely that the ore occurs in veins, fracture zones and pockets), and his suggestion for its origin, are at odds with Murray’s (1852) description. It is hard to reconcile Murray’s stratabound description for the ore zone with Wright’s description. It possible that Murray was describing different aspects of the deposit than Wright . Murray’s description was for Lot 25 while Wright described a deposit in Lot 23. It is also possible that the good stratabound ore that was present when Murray looked at the deposit had been mined out by the time that Wright looked at the deposit.

Those interested in locating the original ore deposit should consult the Geological Survey of Canada’s Map 1182A, Geology Westport Ontario, which can be downloaded free of charge over the internet from the GSC. An extract from the map is provided below.  Hematite occurrences in sandstone in Lot 25, Concession IX of Bastard Township are shown on the extract by "x hem".

I drove down to Delta and briefly looked for the deposit, but as it was deer hunting season did not venture into the woods. I did find an outcrop of massive hematite and  rock impregnated with hematite southeast of Delta (and within a kilometer of the original deposit)  in an area that is mapped as Potsdam Group sandstone by both the Geological Survey of Canada and the Ontario Geological Survey. However, the outcrop is too badly altered to hematite to positively identify it as sandstone. A photograph of the outcrop and a photograph of three specimens of hematite from the outcrop are shown below.   The outcrop consists of both (1) massive hematite mixed with specular hematite and (2) host rock impregnated with hematite.

The outcrop is on both sides of County Road 42 a kilometer southeast of Delta, about 50 meters before the Hicock Road turn off to Lyndhurst. It is possibly altered Covey Hill sandstone and conglomerate (the lower part of the Potsdam Group). Twenty meters closer to Hicock Road , and at a higher elevation, is an outcrop of that is obviously sandstone (possibly Nepean Sandstone - the upper part of the Potsdam Group). The outcrops can be seen in Google Street View. I assumed there was an unconformity between the hematite bearing beds and the beds that are possibly Nepean sandstone, but didn't look for it.

Addendum (January 25, 2022): Ken Watson and Art Shaw have researched the sources for iron ore smelted at Furnace Falls. It appears the Murray (1852) was wrong and that I was wrong, as the deposit just south of Delta, and east of Lower Beverley Lake, was discovered after the smelter at Furnace Falls ceased operations.  This is set out in detail in my January 25, 2022 blog posting.  Worth reading are the recent articles by Art Shaw and Ken Watson that are posted on  the web site for the Old Stone Mill at Delta, Ontario, and Lieutenant  Baddeley's (1831) report on the Lower Beverley Lake Deposit found in ‘An Essay on the localities of Metallic Minerals in the Canadas, with some  notices of their Geological associations and situations etc.’ published in the  Transactions of the Literary & Historical Society of Quebec, volume 2, 332- 426  at pages  336-7, 347, 383- 386  https://archive.org/details/transactionsofli02lite/page/332/mode/2up

Christopher P. Brett
Perth, Ontario

[To Comment: on the line below, click on the word "comment" or "comments".]

From Trails, to Mud Cracks to Evidence of Microbial Mats: Different Theories For Curved Lines in the troughs of ripple marks in Sandstone

The above photographs shows curved lines in the troughs of ripple marks in sandstone. The ripple marks are evidence of wave action in an aquatic environment, but it is the curved lines that are the most important feature of each photo. These sinuous, circular, branching lines that are confined to the troughs between wave ripples are a special kind of microbial mat shrinkage structure. They are fairly common in the fossil record, in the past were given the unpronounceable names Manchuriophycus (for sinuous curved lines) and Rhysonetron (for corrugated circular and sinuous lines), and have now been reinterpreted as resulting from the shrinkage of a microbial mat.

The photographs show two specimens that were found about a five minute drive directly north of Perth in Drummond Township, Lanark County in an area that is mapped as March Formation, which is generally thought to be Early Ordovician in age. The March Formation (the Theresa Formation in New York State and in the Province of Quebec) is younger than the underlying Nepean Formation sandstones. The Nepean Formation (Keeseville Formation in New York State; Cairnside Formation in Quebec) sandstones form the upper part of the Potsdam Group in Eastern Ontario.

Others have reported microbial mat structures in the underlying Potsdam Group sandstones of the Ottawa Embayment. A specimen showing circular and sinuous cracks confined to troughs in wave ripples was reported in Potsdam Sandstone from Canada in a paper read before the Geological Society of London in 1890 by Sir J. William Dawson, one of Canada’s famous geologists. Below is a photograph of the drawing from his paper.

Dawson described the figure as follows:

"Fig. 14 shows a rippled surface in Potsdam Sandstone with marks of worms or molluscs, arranged in the hollows of ripples. The marks are simple trails, of that curious circular or chain-like form sometimes observed, and seem to have been made by animals creeping in the furrows between the ridges of the ripples in the ripple-marks."

[Dawson (1890), On Burrows and Tracks of Invertebrate Animals in Palaeozoic Rocks, and other Markings, Vol. 46, The Quarterly Journal of the Geological Society of London, 595 at 610-611.]




While almost everyone would now agree that Dawson was wrong, he made the above observation approximately 100 years before the feature was commonly accepted as resulting from the shrinkage of a microbial mat. Today I can easily identify the structure as a mat shrinkage structure because I spent part of this past winter reading a number of papers on microbial mat features preserved in sandstones and this is one of the most common textures pictured in the articles. I would not have made that identification without the benefit of those papers.

It is interesting to look at the literature to see how others interpreted these structures. Similar structures were found in the middle of the last century in sedimentary rocks of the Precambrian Shield and attracted varied interpretations. The following is a sampling of these interpretations.

Wheeler and Quinlan (1951) reported on sinuous traces lying in the troughs on the bedding surfaces of ripple marked quartzites in Precambrian rocks in Idaho and Montana, and identified the sinuous traces as mud cracks. They rejected an earlier identification of similar structures in Huronian quartzites from Montana as "trails" or "burrows".

A slab with vermiform markings was found near Sault St. Marie in Precambrian (Upper Huronian) arkosic sandstone. These were interpreted by Frarey, Ginsburg and McLaren (1963) as bodies analogous to the tubes of modern annelids (i.e., worms). They considered and ruled out an origin by desiccation crack filling.

In 1965 similar corrugated specimens were found in Precambrian (Upper Huronian) arkosic rocks east of Flack Lake near Elliot Lake, Ontario. Hofmann (1967) described them as questionably organic, but did not rule out an inorganic origin (the possibility that they were mudcrack fillings, injection or crystal growth was considered). He assigned them the names Rhysonetron lahtii and Rhysonetron byei.

Grant M. Young in articles published in 1967 and 1969 in the Canadian Journal of Earth Sciences reported on similar structures in Huronian Rocks near Elliot Lake, Ontario. He described the beds as cherty quartzites and described the structures as crescent ridges and sinuous ridges in ripple troughs. In his 1967 article he found the origin difficult to explain in terms of inorganic processes, and favoured an organic origin, with some of the structures being probably the casts of vermiform organisms. In his 1969 article he reported that new evidence indicated that the corrugated vermiform structures were formed by the infilling of shrinkage cracks in fine grained sediments, that the cracks were not dessication cracks caused by subaerial exposure, and that the structures were probably formed in sediments containing water.

J. Allan Donaldson (1967) observed modern structures along the margins of ponds. He found "tunnel-like ridges in relatively flat algal mats... [where] the ‘tunnels’ appear as distinct linear, curved, and sinuous ridges that commonly branch and typically taper and disappear over short distances..." He proposed that "at least some of the Huronian vermiform structures may be related to algae rather than to metazoans" and noted that "a muddy environment is not essential for algal growth" and that "algal mats may completely decay subsequent to burial, leaving only the structures they served to create as a record of their former presence."

Hofmann (1971) in his publication Precambrian Fossils, Pseudofossils and Problematica in Canada, Geological Survey of Canada Bulletin 189, reported on a new specimen from the Precambrian rocks near Flack Lake which showed "that the corrugated spindles (Rhysonetron lahtii ) are arranged in a distinct shrinkage crack pattern, and that a biogenic origin can no longer be considered." However, he rejected a mud crack hypothesis (in part because of the paucity of mud), concluding that the specimens "make it evident that the rhysonetron stucture is a Manchuriophycus-type pattern that has undergone unusual diagenetic modification... that involve[d] the reduction, if not elimination, of the pelitic layer, possibly by solution under considerable pressure." He concluded that "Rhysonetron is a sedimentary-diagenetic structure, resulting from shrinkage crack filling, modified by compaction and injection processes, and ... accompanied by almost total removal of the pelitic layer."

Fast forward to the present. Bosch and Eriksson (2008) report on vermiform structures in ripple troughs in 2.1 billion year old sandstones near Pretoria, South Africa. They describe them as "connected or disconnected, curved spindles or rods along the troughs of the ripple marks, resembling worm burrows. These casts form very shallow moulds, up to 2mm deep, on the bedding surfaces and protrude up to 3mm above the bedding surfaces. In plan they curve, branch, taper and may also be longitudinal, and sometimes they overlap." In their analysis of these vermiform curved markings they follow Donaldson (1967), concluding that the sinuous forms found in the South African specimens equate to "a special form of microbial shrinkage crack, normally developed within the thicker mats that occur within the troughs between ripples...".

It is interesting that there have been many theories for these sinusoidal and curved lines in sandstone. As they resemble burrows, it is not surprising that they have been mistaken for burrows. As they resemble mud cracks, it is not surprising that they have been mistaken for mud cracks. However, the curved lines in sandstone resemble curved and sinuous ridges formed on modern microbial mats, and are best explained as a microbial mat shrinkage feature. The solution to their origin is an example of a basic concept of geology that the natural laws and process that operate now have operated in the past. The present is the key to the past.

Added November 12^th:   Below I’ve provided two photographs of specimen that shows a similar texture on a bed surface that does not show ripple marks. Here the spindles or rods curve and taper, and sometimes overlap.

Chris Brett
Perth

[To comment: on the line below click on the word "comment" or "comments".]