Four industrial archeological studies of varying scope have been made at the Whitney Armory site in order to learn about the manufacturing processes that were carried on at the site during its use as a small arms factory in the nineteenth century. These studies of the material remains at the site were intended to augment the very skimpy written evidence available concerning those manufacturing processes. A pilot investigation in the spring of 1972 led to major excavations in the summers of 1974 and 1975, which were followed by a more specialized study in autumn, 1979. These studies have greatly improved present understanding of the structures that were on the site in the past, and given insight into the application of alternative metal-shaping technologies at the Armory in the latter half of the nineteenth century.

The pilot study in 1972 was a joint project by the New Haven Colony Historical Society and the Yale University archeology program. Robert MacDonald, executive director of the Historical Society at that time, and Yale archeology professor Richard S. Ellis directed Yale students in measuring, drawing and photographing the above-ground features of the site, such as the coal-storage shed and the water turbine, and in digging seven test pits. They uncovered a corner of what proved to be Eli Whitney Jr.'s 1860 factory building, and found some files, butt plates, and breech blocks of late nineteenth century Whitney arms.

Encouraged by this experience, the Historical Society two years later co-sponsored another archeological study at the site, this time together with the Historic American Engineering Record, a unit of the National Park Service. It was a more ambitious project, carried out by a team of historians, architects, and archeologists.2 While the historians sifted through the Whitney papers at Yale University and the Water Company papers at the New Haven Colony Historical Society, the architects measured and made drawings of the extant structures that dated from the time of Eli Whitney, Sr. -the barn, boarding house, and coal-storage shed. The archeologists reopened and expanded the 1972 excavation of the 1860 building's foundation, and tested the area of the 1798 filing and machine shop and its millrace, but concentrated most of the summer's effort on excavating the forge shop, which had been built by Eli Whitney, Sr. and had stood for more than a century. Locating its foundations underground was relatively easy, since it appeared, in progressive stages of dilapidation, in twentieth century photographs of the site and had only a few years earlier been bulldozed level. Penetrating its twentieth-century concrete floor proved a harder task; eventually a back-hoe was called in to augment the archeologists' hand-wielded trowels, shovels, picks and sledge hammers. The high water table hindered excavation of the tailrace, which was flooded before the diggers could reach bottom.

But by summer's end almost half of the building and a major section of the tailrace had been excavated. Below the concrete appeared a traprock floor with several trap-rock forge platforms, arranged only somewhat differently from the plan for "a blacksmith shop" that is among the Whitney papers at Yale. The excavation also revealed what was not shown at all in the photographs nor in the sketches for the blacksmith shop: that the building had straddled its mill race, so the waterpower — harnessed first by wheel, later by turbine- had been centrally located within the building.3 The tailrace was wide enough to have accommodated delivery of goods, such as coal, by barge navigating the Mill River, a capability that is alluded to by Yale president Timothy Dwight in his description of the Armory during the time of the elder Whitney: "The river is navigable to this spot for scows, carrying from twenty to thirty tons."4

In the summer of 1975 the archeologists returned to the site, to excavate the millrace of the 1798 filing and machine shop, which had been located at one end of its building. The excavators found material evidence confirming the changes to the water-power system in the 18408 that had been described by William P. Blake.5 An iron penstock, which had been connected to a turbine after 1848, lay in the millrace; some distance beneath it was the wooden planking that Whitney, Jr. had used in 1842 when raising the wheel to combat the problem of backwater.

Below the planking was a thin layer of industrial trash including gun parts, on top of a thicker layer of domestic trash - pottery, bottle glass, buttons, clay smoking pipe fragments - seemingly dumped there as fill to raise the level of the floor for the new wooden lining of the millrace. Industrial and domestic trash was also found on top of the floorboards - including bayonets, gunbarrels, triggers, trigger guards, hammers, buttplates, files, brushes, machinery belts, shoes, glass, smoking pipe bowls and stems - embedded in lumps of asphalt. The gun parts were identified as from the era of Eli Whitney, Jr. Other industrial remains found at the mill race were grindstones of a size suitable for water-power. Additional test pits dug in the presumed area of the 1798 building failed to locate identifiable remains of it.6

What the studies conducted in 1974 and 1975 confirmed with respect to Eli Whitney, Sr. was his careful construction and attention to detail even in such utilitarian structures as the wooden barn and the stone coal storage shed and water-courses for the forge shop, which Benjamin Silliman had praised in 1832.7 Those who had been hoping that the studies would answer some grand question concerning Eli Whitney, Sr.'s machinery and manufacturing methods were disappointed, for the gun parts and tools that were retrieved from the excavations were not- with the exception of a dozen gunflints found in the forge shop tailrace - identifiably from an era earlier than Eli Whitney, Jr.'s proprietorship. The reason for this disappointment was clearer in retrospect than beforehand: the persistence of industrial activity at the site through-out the nineteenth and twentieth centuries right on up to the time of the studies meant that machinery had been repeatedly scrapped and replaced with newer models, old iron was melted down, successive buildings were put up and knocked down and rebuilt in the same location. The building sites were therefore filled with construction rubble, mostly of twentieth-century origin, rather than relics of the flintlock era. David Starbuck, the archeologist who directed the excavations in 1974 and 1975, concluded that this disappointment had been inevitable in studies that gave higher priority to exposing building foundations than to locating industrial trash dumps. He strongly advocated looking for the Armory's trash dumps instead of the Armory's buildings in any future archeological study.8

The advantage of a trash dump for archeological purposes is that by definition it is a concentration of discarded objects. On an industrial site, a dump can be expected to contain any worn-out tools or defective products that have escaped recycling as scrap metal. A dump generally has a shorter period of active use than does a building, and once covered over, is unlikely to be disturbed by intrusion of more recent objects. Dumps are however more difficult than building foundations to locate, for they are regrettably though understandably omitted from the maps, pictures, business records, and letters that make up the historic record of a site. Attempts in 1974 and 1975 to locate old trash dumps of the Whitney Armory were unsuccessful. But in 1979, such a dump was found, just outside the present-day boundaries of the site, and was partially excavated that autumn by a Yale University class in archeological field methods, under the direction of Prof. Harry V. Merrick.

The dump proved to consist almost entirely of thousands of fragments of foundry crucibles, a fact which focused attention on the foundry that Eli Whitney, Jr. was known to have built on the east bank of the river. That the dump did date from Eli Whitney, Jr.'s time was confirmed by the dates "1849" an^ "1850" incised on a few crucible bottoms, and by other roughly datable objects found in the dump: a clay smoking pipe bowl, a fragment of dinner plate with its maker's mark, and a seamless champagne bottle fragment. The main puzzle posed by the dump was its size: so many broken crucibles implied a large amount of molten metal, for which the total amount of brass known about in all the small arms made by Whitney, Jr. didn't seem to be a sufficient explanation.9 The literature about 19th-century small arms manufacturing technology could not answer this puzzle, for it says nothing about casting, but concentrates instead on techniques of metal-cutting and metal forging. Adding to the puzzle, pictures and an 1886 map of the site showed that in addition to its crucible furnaces, the foundry had a cupola furnace projecting from the building on the river side. But cupolas were for melting iron, not brass. The bowls of two iron-casting ladles suitable for use with this cupola had been excavated from the river bank. These clues suggested casting of iron, yet cast iron is normally brittle and totally unsuitable for use in small arms.

The various pieces of this puzzle fell together in a coherent picture when an unfinished and discarded Whitney pocket revolver frame found underground elsewhere at the site was subjected to metallurgical analysis and found to be of malleable cast iron. As discussed in "Arms Production at the Whitney Armory," malleable cast iron was a viable alternative technique for shaping revolver frames, which had hitherto been assumed were shaped by forging. Subsequent scrutiny of discarded iron gun parts excavated in 1975, and in other informal findings at the site, found that some — hammers, triggers, rifle receivers, trigger guards — were clearly of wrought iron shaped by forging, but that pistol frames and a butt plate were of malleable cast iron. Metallographs of samples from completed Whitney revolvers of two different models — the pocket revolver and the Navy revolver made during the same period, the late 1850*5 and early i86o's — demonstrated that malleable iron casting had not been just a temporary experiment at the Armory, but a regular production technique.10

Thus, archeologically-obtained material evidence has revealed what written evidence had left unknown concerning the use of alternative techniques in igth-century small arms manufacturing at a single factory. A plausible explanation for the surprisingly large crucible dump and for the other material findings is that Whitney, Jr. used crucibles not only for melting brass but also for melting iron, to make castings to be malleableized. This wore the crucibles out at a more rapid rate than brass melting does, because of the higher temperature needed for melting iron, and resulted in a larger crucible dump. At some point Whitney, Jr. added a cupola to his foundry, for melting iron in larger batches.

Was Whitney, Jr. the only small arms manufacturer of his day to have made use of malleable iron casting? So far, metallographic analysis of the frames of two non-Whitney revolvers in the Museum's collection, dating from the 18708 and '8os, has shown them to be of steel, which superseded iron for many purposes in the late nineteenth century. To continue research into this question, the Museum welcomes donations or loans of small arms, or parts thereof, that were made by Whitney, Jr.'s contemporaries in the period 1842-1888, so that metals analysis can reveal how they were shaped.

As for future industrial archeology at the Armory site, the ground has only been scratched. Properly executed archeology is a slow and painstaking exercise, but it can both pose and answer questions that the written record has ignored about what went on in factories during our American industrial revolution.