Early in 2001, FOSA member Walt Landgraf was instrumental in starting a research project at the site of a forge in Killingworth associated with a colonial steelworks and famous metallurgical experiment. Working with Yale University historical metallurgist Prof. Robert Gordon, FOSA President Mike Raber, and Frances Kemmish, Walt helped arrange for cooperation and assistance from the Connecticut Water Company, which owns the site. So far, the project has involved background research, detailed surface survey of part of the site, and a short report, all completed between Spring and Fall 2001.

Killingworth minister and entrepreneur Jared Eliot (1685-1763) was an early investor in Connecticut's Salisbury high-grade iron-ore industry beginning in 1734. Pig and cast iron from this region eventually replaced much of the smelted bog-iron ore used at bloomery forges in the state's lowlands. By 1744, Jared's son Aaron (1718-1785) became involved in the earliest Connecticut attempts to convert bar iron into the steel needed for tool edges, working at a furnace in Simsbury perhaps begun as early as 1728. Aaron Eliot made steel by the cementation process. He bought bars of iron and packed them along with charcoal powder into sandstone chests placed over a firebox in a dome-shaped stone furnace. His artisans heated the chests to a bright red heat for a week or more with a wood fire. Carbon from the charcoal packed in the chests diffused into the iron, converting it to steel. Some of this carbon reacted with slag inclusions that were always present in bar iron to release carbon dioxide gas. Because the gas formed blisters on the surfaces of the bars, the product was known as blister steel.

Aaron Eliot built a similar furnace for converting iron to steel sometime before 1755 on Jared's Killingworth property on Pipe Stave Hill. High phospohorous content in locally-available bog ore made steel of low quality, and in 1761 Aaron added a finery forge to convert pig iron, which he would have to buy from distant sources, to bar iron. The Eliots, and other colonial ironmakers, already knew that black sand left behind by running water in brooks and streams, and on beaches, had enough iron content to have some potential as ore. In need of a ready source of pig iron, Jared and Aaron collected enough local black beach sand (possibly at Hammonasset) to fill their saddle bags in January of 1762, and their finer - stimulated by a bottle of rum to overcome his qualms about the endeavor - made about 52 pounds of iron using a bloomery process. Subsequent conversion of part of this iron to steel produced a satisfactory product, from which they made a knife blade which Jared sent to the Society for the Encouragement of Arts, Manufactures, and Commerce in London. The society promptly voted Jared its medal, which arrived in 1763 shortly after Jared's death.

Jared and Aaron had difficulty finding enough iron sand for further trials. They mixed what they did find with bog ore, and made useful, but not particularly good metal. News of the sand iron experiment spread rapidly, and reports of other locations where iron sand could be collected arrived in Killingworth. Jared sent several of his sons to investigate these, and betook himself to New Haven to make experiments with the use of permanent magnets to separate the richest grains from the ore. The Eliots accumulated enough iron sand to make a number of additional sand-iron blooms, which they used for further demonstrations of the suitability of this iron for steelmaking, but there is no documentary evidence of how long Aaron may have continued to use sand ore. He was able to continue making steel through the American Revolution, but probably had to purchase pig iron given the inadequate supply of the sand ore. After his death in 1785, the steelworks ceased operations but the forge was used by others at least briefly, perhaps as an ordinary bloomery smelting locally-available bog ore. The site may have been converted to gristmill use in the early 19h century.

The basic equipment needed at an 18th century finery or bloomery forge included a hearth where burning charcoal fuel provided the heat needed for fining pig iron or smelting ore, a power-driver pair of bellows to supply an air blast to the hearth, and a power-driven helve hammer. The helve was a massive, slow-acting hammer used to expel slag and consolidate the sponge iron produced in the hearth. Typically, separate waterwheels provided power for the bellows and the hammer, requiring a waterpower privilege including a dam, races or flumes of running water to the wheels, and tailraces discharging water to the stream or river below the forge. The Eliot iron works site is located on the Menunketesuck River off Ironworks Road in Killingworth. The forge was on the east side of the river, marked today by two large piles of slag. There is no road access on the east bank, and the forge must have been reached by a bridge from the end of the approach road on the west side. The earth and stone dam abutments survive, with large pieces of slag which appear to have been thrown on the upstream face and crest of the dam from time to time to protect it against wind-driven waves in the forge pond or erosion by flood-stage water. The western end of the dam is now missing. It was probably constructed of wood planks placed over a foundation of stones, and served as a spillway. The bridge to the forge probably crossed this section of the dam.

Abundant black earth with charcoal fragments on the west side of the river, on a section of level ground supported by a massive stone retaining wall, marks the site of the charcoal storage area to which fuel was delivered by wagons traversing a branch of the road that descends Pipe Stave Hill from the west. Remains of the steel furnace have not been found yet. While a steel cementation furnace contained a massive amount of stonework, it might have attracted later people looking for stone to re-use, including perhaps those who built some early 20th century cottages erected on the west side of the river. Since the steel furnace did not need mechanical power from a waterwheel, it might have been placed elsewhere on the property, in a location convenient for the delivery of wood, which was required in large quantities.

Slag is the principal artifact remaining at the Eliot site today. No metals, hammer parts, or furnace remains are present above ground. This absence of artifacts other than slag is typical of other forge sites in Connecticut. Salvagers or, later, scavengers, usually removed metal and machinery for scrap once a forge stopped operating, and forge hearth structures do not survive well in locations susceptible to flood damage. Archaeologists elsewhere often find the anvil base, a massive wooden block, and traces of hearth foundations when they excavate forge sites. No bloomery or finery forge site in Connecticut has as yet been excavated.

The slag at a forge site can yield evidence about the metallurgical processes used and the quantity of iron made. At the Eliot site only surface finds have been examined so far, and these are probably representative of the later period of forge operations. Excavation of the slag piles might yield information on evolution of the metallurgical techniques used during the lifetime of the forge. Prof. Gordon's work included looking at the size, shape, and surface texture of slag samples to estimate the associated metalmaking processes, along with mineralogical and chemical analyses of interior slag structures from thin sections cut with a diamond saw. Intact slag samples were interpreted as having solidified in the bloomery or finery hearth of the forge, which appears to have been about 12x14 inches in plan. Most of the remainaing slag consists of broken plate-like pieces that can be identified from their surface markings as slag tapped from the hearth during smelting or fining. Others are small pieces apparently removed from the hearth during or after the completion of a fining or smelting run. The specimens examined in the laboratory include examples of all these types of slag. Of the five samples on which mineralogical analysis was done, four were found to have resulted from bloom smelting of bog ore, and one from fining of pig iron. This distribution may represent the later stages of operation of the forge since all samples are surface finds. The forge may have been run as an ordinary bloomery smelting bog ore in its later years, perhaps after Aaron Eliot's steelmaking venture was abandoned.

Because of the setting of the forge on a side of the river lacking direct road access, it is unlikely that any slag was hauled away from the site after its abandonment for use as road metal or fill. The total quantity of metal made can be estimated from the amount of slag remaining on the site. Based on transit survey, the contours of the two slag piles were used to compute pile volume totaling at least 4330 cubic feet. In bloom smelting the volume of slag made is approximately equal to the volume of metal made. Based on an estimated 50% porosity within the piles, the slag volume correspond to production of 530 tons of iron over the lifetime of the forge assuming only bloomery smelting was practiced. From the typical 80-pound bloom suggested by the deduced size of the hearth, the number of blooms made would then have been 13,300. If production continued from 1761 to 1785 (the year of Aaron;s death), these blooms could have been made with the forge operating half fo each year at the rate of three blooms per day - which matches the usual experience with bloom smelting. The amount of slag made in fining is somewhat smaller than in bloom smelting. If future sampling and analysis shows that a significant part of the slag pile is fining slag, an upward revision of the production estimate will be needed.

In future research, samples collected from top to bottom of the slag piles might reveal the time sequence of changes in technique used at the forge. Among the samples examined so far, none can be identified as resulting from smelting sand ore. This material may be at the bottom of the piles. If the sand ore never materialized as an adequate source of iron for steelmaking, the quantity of such slag would be small. Using mechanical equipment, an archaeological excavation at the site might also locate surviving traces of the forge equipment. Search of the property for any remains of the steel frunace would also be worthwhile. No remains of a steel cementation furnace have been found in the United States, even though we know that at least two or three others were operated in Connecticut, and many more in Pennsylvania. The Eliot site remains significant today because of tis association with a leading colonial family, and as a place that holds material evidence of the colonial understanding and practice of metallurgical technology.