Midnight Ride, Industrial Dawn

by Robert Martello

  In June 1776, then-Colonel Henry Knox proclaimed, “The business of casting Cannon and making fire arms is of infinite importance to this Continent and cannot be too much encourag’d.” Although colonial American gunsmiths developed impressive skills in the art of making and repairing muskets and rifles, cannon casting remained a virtually unknown technology until the Revolutionary War. By the start of the war, Americans lacked significant numbers of cannon as well as the means or knowledge to make them. British troops carefully guarded cannon stockpiles, and merchants owned the only other cannon for use on their ships. Different colonies, beginning with Massachusetts in 1774, sought to purchase cannon for defense against British attacks. Early purchases and captured stockpiles succeeded in arming some colonial forts and militias with a confusing array of different-sized cannon, but the Patriots soon realized they had to learn to make cannon and ammunition if they wanted any hope of mounting an effective defense. Numerous ironworks had the theoretical capacity to produce cannon, provided they could learn how.33

  During the Revolution, various founders took on this challenge. Beginning with the Hope furnace in Rhode Island, the Salisbury furnace in Connecticut, and the Reading and Warwick furnaces in Pennsylvania, American foundry-men turned out increasing numbers of cannon. They had to produce iron cannon at first because of copper shortages and their unfamiliarity with copper casting. They soon learned the many disadvantages of iron: because iron is less elastic than bronze or brass, metalworkers had to make iron cannon larger to prevent them from bursting but this made them too heavy to move, more suitable for static defenses than battlefield action. Early American cannon had an unfortunate tendency to burst during testing or use, often with fatal results. In the sober words of George Washington, “It is a melancholy Consideration that in these cases we suffer more from our own Artillery than the Enemy.” Still, by the end of the war many hundreds of American-made cannon had been used or placed in fortifications.34 American Revolution–era ironworkers borrowed from existing casting technology, publications such as European “shot tables,” European foundrymen such as Louis de Maresquelle, and trial-and-error experimentation. Revere adopted a similar educational strategy in 1794, though by this time researching new metallurgical fields had become quite familiar to him.

  The cannon-casting process resembled bell or iron casting in many ways, but a cannon undergoes significantly different stresses over its operating life than just about any other object. Cannon casters over the centuries optimized their products to maximize strength and durability and minimize the overall size and weight of the cannon. In addition, casters standardized the production process as much as possible to reduce casting errors and guarantee that ammunition, itself a highly standardized product, securely fit the barrel without jamming. Critical steps in the production process involved selecting the proper alloy, understanding the casting process, and learning to properly finish and test the final product.

  Most early cannon casters preferred to work with “gunmetal,” an alloy of bronze consisting of approximately 90 percent copper and 10 percent tin. American founders often substituted brass for bronze until a special form of gunmetal consisting of 85 percent copper, 5 percent tin, 5 percent zinc, and 5 percent lead eventually supplanted other alloys. Bronze remained the preferred material for cannon casting because its low melting point enabled castings to be more reliably uniform, it resisted rust, and it could be worked relatively easily after being cast into final form. The ease of casting and greater elasticity of bronze also made bronze cannon less likely to burst than iron cannon, and when they did burst the gun tubes usually split without hurling deadly metal shards into the gun crew.35

  The cannon-casting process changed little between the Middle Ages and post-Revolutionary America, and even though Revere never described his own procedures in writing we can confidently assume that he followed the prevailing practices. First, the founder created a mold from baked clay. This mold contained three parts: the cannon barrel’s exterior, which usually included decorations such as coats of arms; the breech, or closed back portion of the barrel; and the core, a plug in the shape of the interior of the barrel that allowed the caster to produce a cannon with a hollow middle. Some founders reinforced their molds with iron bars to better withstand the stresses of casting. Second, the workmen secured the mold in a pit and filled it with molten metal. After the metal cooled, the workmen removed it by breaking the mold. Finally, the caster finished the cannon by carefully drilling or boring out additional metal to enlarge the hollow center, ensuring its proper size and uniformity.

  The traditional cannon-casting and finishing process left room for improvement. The core portion of the mold rarely remained in the appropriate location in the exact center of the barrel while molten metal poured in all around it. Casting difficulties led to many non-uniform cannon with irregular interior dimensions. The French first developed the “solid bore” casting method to correct these problems in the 1740s and 1750s. Cannon casters could cast the barrel of their cannon in a single heavy piece without a hollow core, and then drill out the bore with a powerful boring machine that usually made use of water or steam power and hardened iron drill bits. As a result of the accurate placement and uniformity of the center shaft, barrels more reliably fit the diameter of the cannonballs, reducing the risk of jamming and the loss of explosive force due to gases escaping around the edges of the ball. This new degree of safety allowed casters to reduce the barrel thickness and overall weight of the cannon. Improved cannon-casting technology spread to other countries, reaching England in the 1770s. Revere adopted this process, and subcontracted the drilling portion of his work at least in 1795 and 1796.36

  The high failure rate of colonial and Revolutionary cannon, both during the proving process and under battle conditions, illustrates the incomplete American mastery of casting technology. Many colonial founders initially had a lot of difficulty making sound castings that could withstand the stress of use, or the even worse stress of “proving,” the process of testing cannon for flaws. Revere described this process to a friend in 1795: “In proving my ten I lost one. Lt. Major Lillie was imployed by the Secretary to prove them he filled the Chamber full of powder, the shell full of sand and then Clayed the space between the shaft & Cillander, one of them recoiled 39 feet & all the rest above 30.”37 Proving typically involved discharging a cannon under unusually rigorous conditions, for example, by using triple the normal quantity of gunpowder, discharging the cannonball into the ground, or somehow plugging the barrel. Presumably, if the cannon survived this test it would endure ordinary operating conditions. Many cannon did not pass this test, leading Revere and other founders to complain (to no avail) that it represented unrealistic and pointless expectations.

  As late as December 1795, Secretary of War Timothy Pickering complained, “The casting of cannon has not been attended hitherto with the expected success.” He also stated that the government recently hired a “French gentleman” to overhaul the casting and boring process, and mentioned ongoing efforts to hire a “complete cannon founder” from Europe.38 This letter may refer to the same French gentleman whom Stephen Rochefontaine directed to Revere in 1795, as quoted in the previous chapter. Revere’s startup challenges echo the novelty and risk surrounding American cannon casting in general: even the so-called experts had a hard time turning out consistently excellent weapons.

  Perhaps the major difference between bell and cannon casting involved operational management. Bell casting took place on a much smaller scale, with a maximum of twelve sales in a given year, and therefore Revere could afford to wait for a contract before searching for metal. Cannon casting had more in common with Revere’s ironwork because it took place on a large scale and required more raw materials and a higher emphasis upon steady throughput. Fortunately, his skilled laborers stepped up to this new challenge thanks to the training they had received when working on iron casting. Likewise, Revere had personally acquired helpful managerial experience developing reli
able supply sources and tracking his income and expenditures. All systems were go.

  Revere initiated his cannon-casting career by contacting his old friend Henry Knox, now conveniently positioned as America’s first secretary of war, and offering to produce artillery for the government. He must have seen cannon casting as an opportunity to combine some of the best aspects of his earlier products: cannon equaled or exceeded the size of bells and therefore commanded high prices, and founders usually contracted for large sales involving many cannon at the same time. They embodied great craftsmanship as did his silverwork and bell casting, but they also served a utilitarian goal, all while helping his country cement its hard-won independence. Unfortunately, the full picture of cannon casting included negative aspects as well. Cannon casting involved frustration and setbacks due to the complex process and unreliability of most of Revere’s clients, particularly the federal government. But in spite of these impediments, Revere’s cannon-founding career certainly offered more pros than cons as it provided him with high-profile work that greatly expanded his workshop’s reputation, income, and metallurgical competence.39

  As with the bell-making process, cannon casting relied heavily upon skilled labor and personal judgment. Even experienced cannon founders had trouble getting the metal to solidify in a uniform manner, and a small misalignment during the boring process produced an off-center shaft more likely to burst or misfire. Revere and his workers had to climb a steep learning curve, and his first contract proved the most troublesome of his four client relationships between 1794 and 1800. The federal government hired him in July 1794 to cast ten howitzers, large army field pieces that fired six-pound cannonballs with an eight-inch caliber. Both the War and Treasury departments corresponded with Revere, occasionally in a repetitive or confusing manner, reflecting the unclear division of authority and responsibility in the early executive branch of the government. Over the course of this contract Revere and the government representatives all made mistakes that delayed completion of the cannon and the ensuing payment.40

  Government officials took an interesting initial step when they divided the howitzer contract between Revere and James Byers, a supervisor at the federal armory in Springfield, Massachusetts. The government wanted twenty howitzers, and asked each man to produce half of them for identical contract terms. In addition to spreading the risk between two facilities and producing the cannon twice as fast, this enabled a comparison between the two foundries to help determine the competence and efficiency of each. Indeed, this caused Revere some problems later on when his wasted metal stood in contrast to Byers’s better results. Although under other circumstances Revere and Byers might have seen each other as rivals, they quickly learned that they had similar personalities and technical interests and faced a common set of manufacturing challenges and frustrating bureaucratic hurdles. At this early point in the cannon-founding field, having a local colleague familiar with your trade outweighed the potential loss of business from competition. Revere and Byers chose to cooperate from the outset by exchanging molds, patterns, and advice, and they eventually became close friends. Letters between Revere and Byers display an increasingly friendly tone and the inclusion of personal inquiries in addition to business matters. For example, Revere begins his letters with “Friend Byers” and makes comments such as “My best regards to your good lady.”41

  The vast majority of problems with Revere’s first federal cannon contract centered on raw material procurement, again, an unappreciated harbinger of what Revere and his son faced repeatedly in the years to come. Copper and tin were scarce in early America and the quality of metal was often suspect. Revere initially bought raw materials with a specific product in mind, as he had during his early bell castings. In other words, if he knew he needed ten tons of copper for a set of cannon he would buy precisely ten tons. As his shop grew he gained confidence in his ability to procure contracts that would use up any metal he could lay his hands on, and consequently his raw material purchases became more regular, subject to merchant prices and availability. His shop maintained a stock of metal that he could apply to any project, and he confidently used his own funds to expand this stock whenever he could—a striking step toward large-scale industrial practice. His ledgers contain one furnace stock inventory performed by two of his sons in 1799, presented in Table 5.2.42

  By 1799, Revere had more or less eliminated his iron casting, as shown by the absence of iron in Table 5.3, to emphasize bronze and copper products. Gunmetal became his highest priority at this time, and the separate listing for gunmetal either represents an alloy that he had already created, or a stock of “old” gunmetal from a cannon or other source ready for recycling. Revere’s 410 pounds of bell metal likely refer to a broken bell awaiting recasting. The large weights of copper and tin that he kept on hand enabled him to create more bronze or produce pure copper goods, and the brass may have referred to either brass or bronze since practitioners often used these terms, as well as the alloys they described, interchangeably. The fact that Revere dealt with thousands of pounds of tin and did not have any zinc on hand implies that what he called brass was really bronze.

  Raw material stockpiling represented a logical response to the larger quantities of metal he used in his ever-increasing contracts. In 1795 Revere tallied all his metal expenditures for the federal howitzer contract at the government’s request. He bought 38,032 pounds of copper and 4,712 pounds of tin for 10 howitzers, of which he sent 15,473 pounds of copper and approximately 1,800 pounds of tin to Byers in Springfield at the government’s request. Revere’s final howitzers weighed 16,917 pounds, and although he had some leftover copper and tin at the completion of the contract, he still wasted as much as 8,200 pounds of metal, an enormous amount.43 Metal shortages had a dire impact on Revere’s operations. Both Revere and the government initially estimated the weight of the finished guns to be 1,500 pounds each, when in reality they weighed between 1,633 and 1,750 pounds each. Even though Revere’s cost estimate factored in 16 to 20 percent extra metal, “which is very great,” to account for errors and waste, he still fell short of the mark, and as a result he procured less metal than he needed to carry out his work.44 This miscalculation, among other issues, led to copper and tin shortages that slowed the rate of production while he scrambled to purchase additional materials. Revere also lost one howitzer during the proving process, forcing him to recast it and use even more metal. Since the government agreed to reimburse him for all metal purchases, this wasted metal did not come out of his own funds. But it did not benefit him either: Revere’s salary depended on the weight of the final product, a fee of 17 cents per pound of finished cannon. This fee had to cover his labor, fuel, and other incidental charges, which increased each time he began a new casting or labored to correct an error.

  Table 5.2. May 1799 Furnace Inventory

  In addition to slowing his operations, Revere’s wasted metal greatly concerned the government. Although the War Department’s first contract with Revere agreed to provide him with all needed metal, government officials quickly asked Revere to take on this task. In this case Revere’s purchasing network proved itself more effective than the government’s. He ended up writing many letters to various merchants, cobbling together enough small purchases of copper, tin, and “old” metals to supply both himself and Byers in Springfield. The large percentage of waste reflected the quality of the metal as well as his inexperience. The government limited him to a price of 1 shilling, 2 pence per pound of copper and 1 shilling, 8 pence per pound of tin, which forced him to buy inferior material.45 Added restrictions on his allowable shipping costs forced him to send the highest-quality copper to Springfield because its density made it the cheapest to transport. In contrast, Revere had to work with the bulkier recycled copper and brass implements that had to be melted and recast. In addition, reused copper contained small pieces of iron that either accumulated in the furnace or entered the final product as imperfections. In justifying his larger waste percentages in
comparison with Springfield, he felt the need to defend his honor and judgment: “I was as carefull and as prudent as if the metal belonged to my self & what is deficient was wasted.” He finished his howitzers by January 1795, but the government questioned his excess metal usage and delayed paying his contract until May 1796, more than a year later, when the different departments convinced themselves of his honesty. He grew more efficient as he mastered the casting process and wasted far less in future contracts.46

  Revere’s first federal contract reveals the inexperience present on both sides. Revere took on too many responsibilities and accepted constraints that excessively limited his operations. He could not procure high-quality metal at the price the government set, and wasted time and money working with inferior “old” copper. He also, understandably, had not mastered the casting process as quickly as he expected. His burst cannon and high amount of wasted metal triggered a government review process that delayed his reimbursement more than a year. The government’s division of authority between the War and Treasury departments also complicated all transactions, as Revere often received two sets of identical questions from Secretary of the Treasury Knox and Commissioner of Revenue Tench Coxe, while his own concerns often went unanswered. By the time he finally received his payment, the extra labor for the burst cannon and the lost use of his money during the long delay reduced his profit.