Surveys and test excavations, 1999–2002
- Top of page
- Surveys and test excavations, 1999–2002
- Summary of finds
- The steamboat Heroine
In 1990 the Red River experienced one of its periodic floods. Among the affected areas was a stretch of river in Choctaw County, Oklahoma, which shifted about 0.4 km (¼ mile) to the north of its previous channel (Fig. 1). When the flood waters subsided the remains of a wooden-hulled vessel were discovered eroding from the northern bank of the river. In September 1999 a local resident brought the wreck's existence to the attention of the OHS. Archaeologists William Lees and John Davis, together with INA archaeologist J. Barto Arnold III, undertook a preliminary evaluation of the site. A 15.24 m, (50 ft) portion of the vessel's stern was exposed above the river-bottom sands, but the overall length of the hull could not be determined since the forward end was still buried. Machinery elements on the wreck, including the portside paddle wheel, the port and starboard main shafts, and a pair of centre-mounted flywheels, indicated that the vessel was a western river-type steamboat that likely dated to the late 1830s or early 1840s (Lees and Arnold, 2000). The site was designated ‘34Ch280’ in the Oklahoma archaeological inventory.
Figure 1. Map showing the location of the wreck of Heroine, near Fort Towson in south-eastern Oklahoma. (Drawing: Kevin Crisman, INA/TAMU)
Download figure to PowerPoint
Discussions between the OHS, the US Army Corps of Engineers (Tulsa District), and the Oklahoma Attorney General's Office in 2000 concluded that neither federal nor state governments had jurisdiction over the wreck. Negotiations were therefore conducted with Mr Ricky Martin, the owner of property containing the site. In 2001 Mr Martin generously donated the wreck to the OHS, enabling further archaeological research.
Panamerican Consultants, Inc. was contracted by the OHS in 2000 to carry out remote sensing and test trenching, but unusually high waters during the summer and fall permitted only a magnetometer survey. Undertaken in November 2000, this operation documented a number of anomalies downstream from the wreck, but there were no substantial hits to suggest the presence of boilers or the engine cylinder (Krivor, 2001).
Funding to conduct archaeological and historical research and artefact conservation was secured in early 2001 as part of a larger grant to the OHS to develop transportation exhibits for the new Oklahoma History Center in Oklahoma City. This grant was provided by the Oklahoma Department of Transportation through the transportation enhancements programme of the federal Transportation Efficiency Act for the 21st Century (TEA-21); it has funded both the excavation and the conservation of recovered materials. Artefacts from the wreck have been treated by the CMAC's Conservation Research Laboratory at TAMU.
Major fieldwork began in 2001 with a three-week OHS field school led by Sheli Smith and Annalies Corbin, with overall project direction by William Lees. The partially exposed structure in the stern underwent preliminary mapping, and test excavations in this area yielded several wooden barrels, one of which was recovered. The selected barrel contained a saponified mass of pork flesh and bones. After conservation the bones were analysed by Elizabeth Reitz and Gregory S. Lucas at the Zooarchaeology Laboratory, Georgia Museum of Natural History, University of Georgia (Lucas, 2005).
In 2002 Kevin Crisman joined the project as the principal investigator for TAMU and INA, and further surveys were completed. A 50 x 300 m (165 x 985 ft) swath of river surrounding the wreck was examined by sidescan sonar which clearly revealed exposed portions of the wreck, the sculpting of river-bottom sediments by the current passing around the hull, and confirmed the lack of exposed wreckage other than the stern (Fig. 2). The buried forward portion of the hull was subsequently delineated by a combination of probing and test pitting. The results indicated that most of the hull was preserved to the level of the sheer, that portions of the main deck were preserved at the stern, amidships, and at the bow, and that the overall length of the wreck was about 42.5 m (140 ft) (Crisman and Lees, 2003).
Figure 2. Sonar image of the wreck in the Red River in 2002, showing the exposed stern structure. (Sonar image by Brett Phaneuf, INA/TAMU)
Download figure to PowerPoint
Intensive excavation began in 2003 as a combined INA-TAMU-OHS project (additional technical assistance was provided in 2003 by the Lake Champlain Maritime Museum). Over the course of five weeks in the summer and two weeks in the fall the interior of the stern's starboard side was uncovered and two trenches were excavated across the more deeply buried port side (Fig. 3). A total of 23 starboard frames were documented, and nine frame sections were taken. The portside trenching allowed recording of two complete frame sections (Fig. 4), and exposed the remains of a non-watertight plank bulkhead that separated the stern compartment from the main hold. Excavators also discovered intact main-deck planking on the port side of the stern and a small companionway into the after compartment. Artefacts recovered from within the stern compartment and adjacent hold area included a barrel containing pork bones, a wooden soap-box, two hand-trucks, rope, one pair of wrought-iron barrel hooks, shoes, machinery parts, an iron stove grating, various tools, and many smaller finds. Elements from the steamer's steering system (the tiller, starboard wheel-rope block, and complete rudder), were recovered for conservation and study (Crisman, 2005).
Figure 3. Diagram of the wreck, showing the progress of excavations between 2003 and 2008. (Drawing by Kevin Crisman, INA/TAMU)
Download figure to PowerPoint
In 2004 the INA-TAMU-OHS group undertook eight weeks of field work, five in the summer and three in the fall. The summer work was slowed by unusually high river levels that swept the wreck with a powerful current and caused overnight backfilling of each day's digging. Excavation of the port side of the stern and recording of five frame sections were nevertheless completed. Artefact finds included six sections of sheet-iron stove pipe, a tin basin, three pork barrels (one of which retained a hardened block of its original contents), a large wooden double block and associated iron hook, the port wheel-rope block, and smaller items. The October project focused on the documentation of the exposed side-wheel machinery, allowing TAMU research modeller Glenn Grieco to reconstruct the engine and side-wheel assembly and build a 1/10 scale model for display in the Oklahoma History Center (Fig. 5) (Grieco, 2005).
Figure 5. A scale reconstruction prepared by CMAC modeller Glenn Grieco of Heroine's central hull and propulsion machinery. The pair of large flywheels in the centre is typical of pre-1840s single-engine western steamboats. (Photo by Wayne Smith, TAMU)
Download figure to PowerPoint
In 2005, the INA-TAMU-OHS team returned to the wreck for a third season of intensive excavation and recording. That year's focus was on the approximately 29 m (95 ft) length of hull forward of the side-wheel machinery (the part of the wreck entirely covered by sand). In contrast to 2004, the 2005 season was characterized by dry summer conditions and uncommonly low river levels, enabling the excavators to accomplish much over the course of nine weeks in the summer and early fall. During four weeks in June the portside guard (the portion of the main deck that overhangs the side of the hull), port paddle wheel, and a 12 m (40 ft) length of the main deck and its engine-supporting cylinder timbers were exposed and documented. A 1.5 m (5 ft) wide trench was opened across the hull forward of the flywheels, and a complete frame section was recorded at this location.
Divers excavating amidships, beneath intact main-deck structure, encountered a cache of both crushed and intact pork and flour barrels; examples of both types were recovered for study. Other June 2005 midships finds included two cast-iron rim sections from the flywheels; large sheet-iron plates with riveted edges and numerous brick fragments, most or all of them presumed to be from the boiler firebox and casing; and a small sheet-metal box painted black and decorated with painted flowers (a style known as tole) (Brown and Crisman, 2005).
Five weeks of excavation on the wreck in early fall 2005 concentrated on the forward-most 7.6 m (25 ft) of the bow. This part of the severely hogged hull was buried by up to 4.6 m (15 ft) of sand, but exceptionally well preserved, with the stem, frames, main deck, and port guard in good condition; some wood surfaces even retained their original paint. Openings through the fore deck included a large cargo hatch and two small, side-by-side companion hatches (Fig. 6). The companions provided access to a bow compartment separated from the forward hold by a watertight plank bulkhead that extended to the underside of the main deck. Known as a snag chamber bulkhead or collision bulkhead, this feature was employed on western steamboats during the 1820s and 1830s to prevent the entire vessel from flooding if the bow was pierced by a submerged log or similar hazard (Hunter, 1969: 80; Kane, 2004: 54, 90). This is the only example of a western steamboat snag chamber bulkhead ever seen by archaeologists.
Figure 6. Plan view of the bow of Heroine as uncovered by excavators in the autumn of 2005. (Drawing: Kevin Crisman, INA/TAMU)
Download figure to PowerPoint
Eight bent or broken iron castings were neatly stacked beside the keelson in the bow compartment: two fire gratings that had sagged from over-heating, five heavy bars that may have supported firebox gratings, and half of a cracked-and-splinted cam frame (part of the reciprocating mechanism that opened and closed the cylinder valves). These materials were probably destined for recycling at a foundry; a contemporary account book shows that broken or damaged cast iron had a scrap value of 1¢ per pound (0.45 kg) (Jefferson Foundry Journal, 1837). Other bow-compartment finds included two iron hammers and a wooden mallet, three block sheaves, and an intact glass medicine bottle. Four frame sections were recorded in the bow, and offsets were taken to record the curvature of the apron and upper stem (Crisman, 2007).
The fourth and final season of full-scale excavation, in 2006, was divided into two campaigns: five weeks in May and June, and four weeks in September. The work was again aided by drought conditions that significantly lowered the river's depth and current. The May–June work involved uncovering the 7.6 m (25 ft) section of hull between the bow and the cylinder timbers (the boiler and firebox assembly was originally mounted here on the main deck). The wreck was severely damaged at this location, with the port side broken and collapsed outward, the starboard frames missing above the turn of the bilge, and main deck gone. Much of the damage to the port side, we discovered, was caused by the ‘snag’ that sank the steamboat, a 3.76 m (12 ft, 4 in) softwood log that penetrated the hull at the turn of the bilge, adjacent to the forward end of the cylinder timbers. Judging by the shattered state of the frames and planking around the snag, the damage to Heroine was profound, causing the steamer to fill and sink very quickly. The keel and keelson of the steamer were also broken directly beneath the forward end of the cylinder timbers, a break that marked the beginning of the severe hogging at the forward end of the hull. These features were recorded in detail, along with the hull's general assembly pattern and three frame sections.
The excavation turned up four cast-iron elements from the boiler's mounting framework and the front face of the firebox. Also found were numerous bricks used to line the boiler firebox and large and small pieces of thin sheet-iron that were riveted together to form the boiler casing. Related finds in 2006 included another cast-iron fire grating, a section of cast-iron pipe with a flange at one end and a 45° bend at the other, and a T-shaped assembly of cast-iron pipes with a check valve that supplied water to the boilers. Cargo-related remains consisted of a broken but unused grindstone, and the oaken staves and wooden hoops from two provision barrels. The excavations also yielded miscellaneous finds such as a cast-metal US Army uniform button, a narrow-bladed caulking iron, a cold chisel, and a small collection of plain and transfer-decorated whiteware ceramic fragments.
The objective for the September 2006 project was to recover the steamboat's paddle-wheel components; these included the port and starboard main shafts and their attached flywheels, the port paddle wheel's shaft, flanges, arms and buckets, and five pillow blocks that supported the shafts. Conservation and analysis of these unique examples of early steam technology had long been contemplated by the project partners, since examples of early steamboat machinery are extremely rare (Krueger, 2012: 3, 194–242). The machinery data collected by Grieco in 2004 proved useful to the disassembly and lifting operations. Pneumatic and hydraulic tools were prepared for the work, but crew members found that even after 168 years in the river the majority of the bolts and nuts securing the machinery to the support timbers could simply be unscrewed with large adjustable and socket wrenches. Bottle jacks were then used to lift shafts and pillow blocks off the timbers. The three shafts with their flanges, each weighing between 1179 and 1542 kg (2600 and 3400 lbs), were blocked in place to await transport off the wreck-site by helicopter. The intact lower halves of the two flywheels were taken apart by separating the oaken arms from the flywheel flanges and unscrewing bolts that held the cast-iron rim sections together. The resulting 159 kg (350 lbs) rim-and-arm sections were then lifted out of the hull and transported to the conservation laboratory.
The starboard guard was missing from the wreck, and we believed it likely that the starboard paddle-wheel assembly fell alongside the hull when the guard collapsed. A test trench dug outboard of the main shaft in June 2006 proved that this was the case, for it revealed one half of the inboard paddle flange with six wooden paddle arms (Fig. 7). Further digging in September uncovered the starboard paddle shaft with the complete outboard flange still attached.
Figure 7. Part of the starboard paddle wheel recovered in 2006. Note the reinforcing struts, called blocking, that were toenailed between the wooden paddle arms. (Photo: Rebecca Sager, INA/TAMU)
Download figure to PowerPoint
The three heaviest pieces of the side-wheel machinery (the port and starboard main shafts and their flanges), as well as the port paddle shaft and its two flanges, were fitted with slings and lifted off the wreck by the St Louis Helicopter Company in November 2006. Placed on the north (Oklahoma) side of the river, they were subsequently transported to the conservation laboratory at Texas A&M.
The starboard paddle shaft and attached outboard paddle flange, as well as two pillow blocks remained on the wreck after 2006. A two-week project to recover these elements and to conduct a further search for boiler elements off the wreck's port side was therefore undertaken in September 2008. During the first week the machinery pieces were excavated and lifted using a pair of rafts and a chain hoist. The test excavations and metal-detector survey off the port side during the second week yielded no additional boiler parts, but did reveal more of the port guard assembly and previously unknown details of the main deck's construction. The completion of this work marked the end of the fieldwork portion of the steamboat study.
Summary of finds
- Top of page
- Surveys and test excavations, 1999–2002
- Summary of finds
- The steamboat Heroine
Contemporary newspapers and correspondence tell us that Heroine was partially salvaged at the time of its sinking and was periodically exposed for about five years afterwards; not surprisingly, the wreck was neither as intact nor as heavily laden with goods as the spectacularly complete lower hulls and cargoes found on the later western steamboats Arabia (1856) and Bertrand (1865) (Petsch, 1974; Hawley, 2005). Nevertheless, substantial portions of Heroine have survived along with associated artefacts. For the purposes of research and analysis the finds have been sorted into four broad functional categories: 1) the hull, detached timbers and fasteners, as well as the equipment used for the boat's daily operations and maintenance; 2) the propulsion system, including machinery and pieces of the boiler assembly, and the tools used to maintain or repair the engine; 3) cargo and cargo-handling equipment; and 4) the miscellaneous possessions, tools, footwear, and diet-related items left by the crew and passengers. The following is a brief summary of discoveries in each of the four categories (more detailed reports are currently in preparation).
The hull of Heroine was extensively recorded throughout the excavation campaigns to permit an analysis of its design and assembly, and to prepare a set of lines and construction plans. Overall, the steamboat's hull was in good condition, with intact structure preserved up to the level of the main deck at the bow, amidships, and at the stern. The two sections of hull between these areas, the forward hold and the after hold, were in rougher condition, with much of the main deck gone between midships and the stern, and the entire deck and part of the sides missing between midships and the bow. Not surprisingly, very little was left of the steamboat's lightly built superstructure, but the lower ends of support posts provided some evidence of its dimensions and assembly. Notwithstanding the good state of preservation, the high costs and logistical complexity of recovery, conservation, and display ruled out salvage of the entire vessel (Crisman, 2011).
Several patterns were evident in the shipwright's selection of building materials. First, nearly all of the hull and main-deck timbers sampled for tree species were fashioned from white oak (Quercus alba); the exceptions to this were the pine (genus Pinus) deck planking and black locust (Robinea pseudoacacia L.) bitt posts. The quality of the oak timber appeared to be uniformly good, which we might expect in light of the immense timber resources available to mid-continent shipwrights in the 1830s.
Second, almost no compass timber was employed in the construction of the vessel. For the greater part of its length the hull had near-flat floors and vertical sides, a form well suited to the use of straight-grained pieces of wood. The turn of the bilge was created by cutting the first and second futtocks from straight-grained pieces of oak to match the required curve (it is probably not a coincidence that many of the frames were cracked or broken at this location). The beams of the main deck lacked lodging, hanging, or dagger knees. The bow assembly featured naturally curved stem, apron, and breast hooks, the stern assembly had a large stern knee, and the frames at the extreme ends contained some naturally curved pieces, but that was the extent of the compass timber used in the vessel.
The third pattern evident in the hull's materials was the builder's reliance on square-sectioned chisel-point iron spikes and round-sectioned through or clench bolts to fasten all of the timbers together. No other types of wood or metal fasteners were observed on the wreck. The extensive iron industry that existed in the Ohio River Valley by the 1830s made such fasteners both widely available and inexpensive.
It is clear that when Heroine was assembled in 1832, western river steamboat builders were taking every measure possible to keep the hull's draft to a minimum. They did this by employing a full design, with a gracefully swept but short entrance, a short, relatively full run aft, and a boxy section with a shallow hold over most of the vessel's length (the depth of hold was only 6 ft (1.82 m). The builders further minimized the draft of the hull by incorporating a shallow keel with a moulded height of only 4 inches (10.16 cm). Finally, weight was minimized by reducing the quantity and dimensions of timbers in the hull. The frames and deck beams were considerably smaller and more widely spaced than similar elements in contemporary lake-going steamships of this length (Robinson, 1999; Schwarz, 2012). A draft mark carved into the starboard of Heroine's stem, ‘V’ for 5 ft (1.52 m), showed what the steamboat was expected to draw when fully laden.
Steamboats on the western rivers were rarely subjected to the stresses caused by high waves, but the combination of shallow draft, light construction, a high length-to-beam ratio, and uneven loading of the boilers, machinery and cargo meant that their hulls were prone to hogging and sagging. This problem was partly rectified by the 1840s with the widespread adoption of truss systems employing wrought-iron rods, turnbuckles, and sampson posts. The longitudinal supports were known as ‘hog chains’ and the transverse systems as ‘cross chains’ (Hunter, 1969: 94–100; Custer, 1991; Kane, 2004: 63–64, 94–96). Heroine had at least one cross chain immediately forward of the paddle wheels to support the main deck guards and side wheels. That was it, however: the well-preserved bow and stern showed absolutely no evidence of longitudinal hog chains. Instead, the wreck revealed an earlier and heretofore unknown approach to the hogging problem on western steamboats: the principal longitudinal members were notched down over the frames and secured with heavy bolts that were headed over clinch rings. Timbers fitted in this manner included the keelson, bilge keelsons (the stringers at the turn of the bilge on each side of the vessel), and upper and lower clamps (Fig. 8). Interlocking the frames and longitudinal members in this manner undoubtedly stiffened the hull, but measuring, sawing, and chiselling 500 or more notches also added greatly to the labour and time required to complete a new steamboat. Notched longitudinal timbers have not been observed in the hulls of later steamboat wrecks, suggesting that this practice was abandoned once hog chains came into use (Kane, 2004: 105–107).
Figure 8. Perspective view of Heroine's hull construction, showing how the keelson, bilge stringers, and clamps were notched down over the frames and secured with bolts headed over clinch rings. (Drawing: Kevin Crisman, INA/TAMU)
Download figure to PowerPoint
Despite the extensive hogging and twisting of the hull and loss of the sides and main deck at certain locations, enough of the structure has survived to permit reconstructions of Heroine's lines and assembly. Lines prepared from the archaeological data show that Heroine had an overall length of 136 ft, 8 in (41.65 m) between perpendiculars, a moulded breadth of 20 ft, 4 in (6.19 m), a maximum breadth across the main deck of 36 ft (10.97 m), and a depth of hold of 6 ft (1.82 m). Heroine's registered tonnage was variously listed as 146 tons (Lyford, 1837), 150 tons (Collins, 1836: 62) and 160 tons (Hall, 1836). Comparison of these tonnage figures with a listing of tonnages for 183 of the 220 steamboats operating on the western rivers in 1832 show that Heroine fits into the medium-to-small range of contemporary river steamers (Otis, 1832: 127–130; Chevalier, 1961: 209). All lines of evidence, archaeological and historical, suggest that in terms of its design and construction Heroine was probably a very typical western steamboat of the 1830s.
Plans, specifications, or preserved examples of early western river steamboat propulsions systems (the boilers, pistons, and side-wheel machinery) are all relatively rare; the earliest set of detailed plans we have for this style of equipment dates to 1840, nearly three decades after the introduction of steamboats on western rivers (Hodge, 1840). Recovery and study of surviving boiler and machinery elements and thorough documentation of their supporting structure was therefore a high priority of the excavation project. Heroine's propulsion system was in much the same condition as the hull: although not complete, more than enough was preserved to provide us with a good idea of its dimensions, layout, and operational parameters. We know that the single-cylinder engine was salvaged immediately after the sinking (New Orleans Daily Picayune, 23 May 1838). The boilers, chimneys, and wood-and-iron connecting rod (or ‘pitman’) may have been recovered around this time as well, since they were not found on the site. Aside from these items, tangible evidence of the remaining elements of the propulsion system was encountered on the wreck.
Heroine had a direct-acting high-pressure steam engine, as did most 19th-century western rivers steamboats. This type of engine was relatively easy to make, cheap to purchase, simple to operate and maintain, and for its size and weight had a greater horsepower than the low-pressure engines that were widely used on contemporary European and eastern North American steamboats. They were well suited to meet the rigours of river navigation. The big drawback of high-pressure systems of the era was safety: they had a tendency to blow up or unexpectedly release steam, killing or injuring nearby crew and passengers. Heroine's propulsion system consisted of three elements: the boilers that generated the steam, the piston engine that used the steam to produce reciprocal motion (converted into rotary motion by the crank), and the drive train assemblies made up of fly wheels, shafts, and paddle wheels.
Surviving elements from Heroine's boiler assembly included a feedwater pipe fitted with a one-way valve and four pieces of the cast-iron mounting frame and firebox front, which together indicated the height of the boilers above the deck, their diameters, and the general appearance and assembly of the firebox. Also found were three whole and many pieces of cast-iron firebox grate bars, whole and fragmentary fire bricks that lined the interior of the firebox, and large and small sections of the sheet-iron cowling that covered the lower halves of the boilers and contained the heat and smoke of the fires. The cast-iron pipe with the flange at one end and 45-degree elbow at the other was likely used to vent excess steam from the boilers to one of the chimneys (Hodge, 1840: Plates XXXII and XXXIII).
The engine cylinder was missing, but there were numerous clues to its location on the vessel and to its general size and configuration. The heavy wooden support structure for the engine and flywheels on the main deck, the cylinder timbers, was nearly complete; an arrangement of bolt holes and notches near the forward end of the cylinder timbers showed where the engine was seated. Also found here was one of a pair of rectangular cast-iron bed plates that provided a solid mounting for the cylinder. The length of the cylinder's stroke (and thus the approximate length of the cylinder itself) could be determined from the turning radius of the crank pin on the flywheels, 57 in (1.44 m).
The third element in Heroine's propulsion system, the flywheel-and-paddle-wheel drive train, was the most complete of the three sub-systems (Fig. 9). Mounted at the after end of the cylinder timbers on the centreline of the boat were two 14 ft (4.26 m) diameter flywheels that smoothed the motion of the crank and reduced jarring or vibration on the moving boat. Each consisted of a large cast-iron flange, eight oaken arms, and a cast-iron rim made up of 16 overlapping sections; the wheel assemblies weighed about a ton apiece. Both flywheels were in damaged condition on the wreck, with the central flanges cracked in several places and the top half of each rim broken off; whether this damage occurred at the time of the wrecking or thereafter was not certain, but most of the detached rim sections were recovered during the excavations. The main shafts, bolted to the centres of the flywheels, extended outboard to the sides of the hull, and each was supported by a pair of heavy cast-iron pillow blocks fitted with copper-alloy liners to reduce friction. Both main shafts and all four of their pillow blocks were found on the wreck.
Figure 9. Heroine's surviving cast-iron drive-train elements, including flywheel and paddle flanges, shafts, and bearings. (CAD drawing by Glenn Grieco, CMAC)
Download figure to PowerPoint
The port drive train was nearly complete, with all of its elements mounted on their support timbers (Fig. 10). The cast-iron coupling that connected the outboard end of the main shaft with inboard end of the paddle shaft was still in place, along with the Y-yoke and lever that allowed the crew to disconnect the paddle wheel when required. Heroine's port paddle-wheel assembly consisted of the paddle shaft mounted on two pillow blocks, with a pair of flanges wedged in place on the shaft. The flanges were missing part or all of their arm pockets and were clearly in need of replacement at the time of the boat's sinking. Most of the port wheel's 12 pairs of oaken arms and about half of the paddle blades (‘buckets’) were also in place. The starboard paddle wheel may have been partly salvaged in 1838, for we found only the shaft and outboard flange, and one half of the inboard flange and wooden arm assembly. Missing from the starboard assembly were the two pillow blocks that supported the paddle shaft, as well as the coupling that engaged or disengaged the paddle and main shafts.
Figure 10. View of the surviving portside propulsion machinery on the wreck of Heroine, as re-created by CMAC modeller Glenn Grieco. (Photo: Wayne Smith, INA/TAMU)
Download figure to PowerPoint
Cargo-related finds from Heroine's hold included tools for loading and unloading cargoes, shipping containers (barrels and boxes), and remnants of the commodities themselves. The after hold yielded two hand-trucks used to lift and roll heavy items on and off the deck. The trucks were composed of wooden handles and cross pieces, cast-iron wheels, and wrought-iron axles and noses. One truck was nearly complete, and had the maker's name, ‘J. Walter’ branded into one of its cross pieces and stamped on a small copper-alloy plate tacked on to another cross piece. This truck was probably made by the same J. Walter who operated a foundry and machine shop near the riverfront in Louisville, Kentucky (Collins, 1838: 110). The second truck was missing the upper handle ends. Both had seen hard use, for the complete truck had a cracked handle repaired with a splint, and the second had a broken wheel that was made whole (though wobbly) by heat-shrinking a wrought-iron rim around its circumference. Other cargo-loading tools included a pair of broad-bladed can hooks for lifting barrels (found with their connecting rope and a pair of thimbles in the stern compartment), and a pair of two-pronged bale tongs for lifting bales of cotton or hay (found in the forward hold). Spare or discarded block sheaves, a half-dozen iron hooks of various sizes, several fragments of rope, and an intact double block found on the after deck may all have been used for shifting cargoes during Heroine's career.
Historical records tell us that on its final voyage Heroine was transporting the annual supply of provisions to a US Army cantonment on the upper Red River, and that part of the cargo was salvaged after the sinking. Evidence of the cargo was found in abundance, however, principally in the form of whole barrels and parts of barrels. A total of 170 individual oaken staves were recovered, along with 13 heads representing at least ten complete barrels, and ten half heads. Three intact barrels were recovered, complete with their pickled-pork contents. Researcher Nina Chick identified three types of barrels in the cargo, based on the length and thickness of the individual pieces, their assembly, and markings on the barrel heads: 1) thickly staved and heavily hooped tight barrels intended for pickled pork, 2) medium-staved semi-tight barrels that held flour, and 3) light-staved and minimally hooped slack barrels that probably held dried beans. Pork and flour barrel heads had stencils or brands that identified contents, packers, or inspectors. Some flour barrel heads were stencilled ‘USA’ (then an acronym for the US Army) (Fig. 11a), branded ‘S. Fine’ (for super-fine flour), and stencilled with the name ‘Armstrong’, indicating that the contents had been inspected by Cincinnati Flour Inspector James Armstrong (Deming, 1834: 11). One intact pork barrel recovered in 2005 from the midships pork-and-flour-barrel feature was stencilled ‘A. S. Reeder Packer Cin't’ identifying the meat as the product of Cincinnati packer Alfred S. Reeder (Fig. 11b) (Woodruff, 1836: 142). A complete wooden box found in the stern compartment was stencilled ‘No. 1 SOAP’ and labelled with the painted script ‘Vicksburg’; this may have belonged to the boat or the crew, rather than the cargo (Chick, 2011).
Figure 11. a) Barrel head stencilling provided evidence for the destination of Heroine's final cargo, the US Army garrison at Fort Towson in the ‘Indian Territory’. (Photo: Rebecca Sager, INA/TAMU) b) Barrel head stencilling provided evidence for the origin of the pickled pork in Heroine's final cargo: ‘A. S. Reeder Packer Cin't’. (Photo: Wayne Smith, INA/TAMU)
Download figure to PowerPoint
Personal possessions or artefacts related to food storage or preparation were present on the wreck, but in relatively modest quantities, which is perhaps not surprising when we consider that nearly all of the space below Heroine's main deck was given over to stowage of cargo (Jones, 2011). On western river steamboats of the 1830s and later, all passengers and crew typically slept, ate, worked, and socialized on the main or upper decks, not in the hold. The majority of the personal possessions were found in the stern compartment or scattered inside the hold. The bow compartment held the cast-iron pieces awaiting recycling, a few tools and odds and ends, but the only obvious personal item here was an octagonal light-green glass bottle embossed with the words ‘Millers Tonic’.
The stern compartment apparently served as a catch-all for spare, worn, or broken ship's equipment; a considerable quantity of wood shavings and spilled pine tar suggest that wood-working or other types of repairs took place there as well. The run contained a curious selection of personal effects: a complete though worn leather boot and several discarded shoes, a stirrup, half of a crystalline geode, the clasp to a coin purse, and the silver handle of a fork or spoon with possible owner's marks scratched on one surface. There was not much of use or value, suggesting that the crew normally did not stow their personal possessions in this space, or that they were able to salvage the interior of the run after Heroine sank.
The steamboat Heroine
- Top of page
- Surveys and test excavations, 1999–2002
- Summary of finds
- The steamboat Heroine
Historical research on the Red River wreck initially focused on identifying the steamboat and determining the date of its loss. When the wreck was first examined by INA and OHS in 1999, two features helped to narrow down the likely time of the sinking. The first was the relatively large dimensions of the hull. Until 1838 steam navigation on the upper Red River was obstructed by a 257 km (160 mile) logjam in north-western Louisiana known as the Great Raft (Bagur, 2001: 60–116); a handful of small steamers slipped through the bayous around the raft prior to 1838, but it is unlikely that a vessel of this size could have ascended to Fort Towson until a channel was opened through the logjam. The second datable feature was the flywheels, which indicated that this vessel had the single-piston configuration of the earliest western steamers; most boats after 1840 were equipped with twin pistons, which did not require flywheels (Norman, 1942: 403–404). Researchers thus began to search historical records for boats sunk in the upper Red River during the late 1830s or early 1840s.
Evidence soon pointed to one likely candidate for wreck 34Ch280. Correspondence from the US Army commander at Fort Towson reported the snagging of a steamboat laden with supplies for the fort in early May 1838 (Vose, 1838; Foreman, 1968: 89). Unfortunately, he did not name the boat. The memoirs of a Red River steamboat captain, written in the 1870s, recounted the sinking of a steamer not far from Fort Towson in the late 1830s, but misidentified the boat as the New York (Wittenbury, 1871). The 23 May 1838 edition of the New Orleans Daily Picayune ultimately provided the correct name in a brief note: ‘The steamboat Heroine, laden with stores for U.S. troops, struck a snag about two weeks since while ascending Red River, two miles above Jonesborough, and sunk. Her cargo was saved’. Jonesborough, Texas was abandoned not long after Heroine sank, but its location is well-known today, and the town was indeed about 3.21 km (2 miles) downriver of the wreck.
Finding this name raised a new complication, for researchers soon discovered that there were two steamboats named Heroine operating on the Mississippi and Ohio Rivers at the same time. The only 1830s-era Heroine on the Lytle-Holdcamper List (1975) (a standard reference of documented early American steamboats) was a 97-ton vessel built at Bridgeport, Pennsylvania in 1832 and reported sunk in the upper Mississippi River in 1837. Neither the tonnage or date and place of loss matched the Red River steamboat wreck. A contemporary steamboat inventory solved the mystery, for it listed a second, 146-ton Heroine, also built in 1832 but at New Albany, Indiana (Lyford, 1837: 464); this was the steamboat sunk near Fort Towson in 1838.
Copies of government enrolment papers for the Red River Heroine cannot be found in the US National Archives (many enrolments from this period are unfortunately missing), but details of the vessel's five-and-a-half-year career have been culled from other sources, principally newspapers. The historical evidence suggests that throughout most of its career Heroine was owned, either entirely or in part, by Jeremiah Diller of Louisville, Kentucky. A cabinetmaker who changed career in his early 40s to become the owner and captain of a series of steamboats, Diller exhibited many of the traits described as typical of river captains of his era:
These … boats were often commanded by men of marked individuality, great force of character and courage, though as a rule, they were men without cultivation or early education. Such men were best adapted to the business, as it was a wild and hazardous life in those days, they having to deal often with desperate characters, both on board of their boats and on shore (Fulkerson, 1885: 18).
A good example of Diller's single-mindedness was his decision to proceed with Heroine's maiden voyage of 1832 at the height of the Mississippi Valley's first outbreak of Asiatic cholera, an epidemic that resulted in a horrific death toll at New Orleans (Fig. 12) (Clapp, 1863: 115–141). Diller was surely aware of the severity of the sickness at his destination, but continued Heroine's passage from St Louis to New Orleans nonetheless (New Orleans Bee, 26 November 1832).
Figure 12. An advertisement for Heroine's inaugural voyage, which was begun during the height of the cholera epidemic in the American west. (Louisville Public Advertiser, October 26, 1832)
Download figure to PowerPoint
What do we know about the design and appearance of Captain Diller's steamboat from historical sources? Newspaper advertisements for Heroine described it with the usual steamboat superlatives of the day (‘staunch’, ‘for accommodation and speed … not surpassed’, ‘new and fast running’, ‘substantial’), but also revealed some of the vessel's characteristics. During July and August 1834, a time when low water levels obstructed the lower Ohio River, New Orleans Bee advertisements reassured potential passengers that Heroine was a ‘light draught’ boat and therefore likely to reach Louisville without running aground. Heroine was promoted as an ‘upper cabin’ steamboat in another newspaper, a reference to the late 1820s and early 1830s trend of shifting the living quarters of higher-fare cabin passengers from the main deck or hold to the upper deck (Missouri Republican, 14 April 1835; White, 2009: 62–63). The upper deck was found to be an altogether quieter and more comfortable place to reside on a steamboat, and was also safer in the event of catastrophic boiler failure.
Documentary and pictorial evidence suggests that Heroine and most other western rivers steamers of the time were built and outfitted with a standardized arrangement of hold, decks, interior spaces, and propulsion system (Fig. 13). The shallow draft imposed by the rivers and the long, narrow design favoured for speed meant that hulls had limited internal capacity; the hold was reserved for cargo, with spaces at the ends for ship stores and the crew's personal effects. The broad main deck with its overhanging ‘guards’, on the other hand, served multiple functions. The boilers, engine, and side wheels all resided here, spread out over the middle of the boat to distribute their weight (side wheels were standard at this time, only later in the 19th century would the stern wheel become commonplace). The main deck was often used for carrying cargoes, especially bulky items such as cotton bales and livestock. Also occupying this deck were a capstan at the bow, stacks of firewood for the boilers, sleeping quarters for the crew, the galley, crew and passenger toilets, storage lockers, and stairs for ascending to the upper deck. Finally, but hardly least, the main deck was occupied by second-class or ‘deck’ passengers, who generally constituted the majority of the fare-paying travellers on board. These people, often the poorer classes of workers, immigrants, or blacks (both slave and free) slept in hammocks, wooden bunks, or on the deck, and cooked on a communal stove in the midst of the main deck's noise and crowding. All accounts agree it was a hard way to travel, but deck passage had the advantage of being only one third the cost of cabin passage (Logan, 1838: 112–115).
Figure 13. A watercolour painting of steamboat Ouishita c.1836, showing the general appearance and layout of vessels like Heroine. (Painting by Richard G. A. Levinge, courtesy of the Amon Carter Museum, Fort Worth, TX, cat. no. 1968–271)
Download figure to PowerPoint
Upper-cabin steamboats such as Heroine reserved the upper deck for the first-class or ‘cabin’ passengers. Although three or four times the cost of deck passage ($25 versus $6 for a Louisville-New Orleans passage in the early 1830s), a cabin passage meant that the traveller was provided with a heated internal space in cold weather, tables and chairs for dining and relaxing, carpeted floors, three meals per day, and a bed to sleep on at night (steamboats commonly over-booked their beds, so cabin passengers could find themselves sleeping on a mattress on the cabin floor instead) (Haites et al., 1975: 31–32). The largest class of steamboats featured individual staterooms, but a second-class boat such as Heroine likely offered only gender-segregated common spaces for men and women, called the gentlemen's and ladies' cabins, each with a row of bunk beds (curtained for privacy) extending along either side of the compartment. At the forward end of the gentlemen's cabin was a separate smaller public space with a bar, the men's washroom, the clerk's office, and sleeping quarters for the captain and other officers (Maximillian, 2008: 199).
Western steamboats of Heroine's time had only a main and upper deck (a third deck with enclosed crew quarters, the ‘Texas’, appeared later in the century). The top of the superstructure had a canvas-and-tar-covered roof that was crowned to shed water. This ‘hurricane deck’ had a number of protruding features: at the forward end was the pilot house, a roofed but windowless structure that contained the wheel; raised, box-like skylights over the men's and women's cabins allowed natural light to enter during daylight hours; and spread along the length of the roof were the sheet-metal chimneys for cabin and cook stoves, as well as a heavier ‘scape pipe’ that vented the engine's used steam. The hurricane deck was the favourite location for cabin passengers on fair-weather days, when it offered cooling breezes, vistas of passing scenery, and opportunities for sportsmen to take pot-shots at alligators, birds, turtles, and other unfortunate wildlife (Levinge, 1836; Martineau, 1838: 8; Levinge, 1846: 34).
Contemporary travel accounts and steamboat records tell us about the different crew occupations on a western steamboat. The number of people employed in each category depended on the size of the boat, the preferences and resources of the captain and owners, and probably fluctuated from passage to passage. At the top of the chain of command was the captain, the decision maker whose role in the day-to-day operations of the boat was largely supervisory. The position was a revolving one: Jeremiah Diller was captain for substantial part of every year between 1832 and 1837, but every few weeks or months he left the boat and turned command over to another. Heroine steamed under at least 12 captains over the course of its career, one of whom was Diller's son-in-law Christopher Castleman. Execution of the captain's commands and responsibility for the ‘hands-on’ management of the boat and crew fell to the steamboat's mate. Other officers included a clerk who kept the financial records, pilots who navigated the winding and hazardous river channels, and engineers who kept the boilers, engine, and wheels in operation. Any or all of these individuals might have assistants or apprentices. Also labouring on the boat were several deck-hands and the firemen who stoked the boilers; the latter job was often described as the most physically demanding work to be found on a steamboat (Gerstäcker, 1854: 95–96). Staff attending to the care and feeding of the crew and passengers included the cook and his assistants, the cabin steward, steward's assistants, and a chamber maid who attended to passengers in the ladies' cabin. Altogether, Heroine likely had a crew of about 20–25 when steaming up and down the rivers (Hunter, 1969: 442–451).
The various captains of Heroine were listed in the newspapers, but the names of other crew members are thus far unknown. We can get a glimpse of them, however, through the eyes of Virginia businessman William Fairfax Gray, who took passage on board from Vicksburg, Mississippi to Louisville in June, 1836. Gray found the officers ‘very attentive’, but was particularly impressed with the supervisor of Heroine's upper cabin: ‘The steward, an old black man, is the best steward I have seen on the western waters’ (Gray, 1965: 186–187).
It is clear from the newspaper ‘marine news’ columns that Heroine was one of the commonplace ‘transient boats’ of its day, steamers defined by a 20th-century historian as ‘free lances roving from trade to trade wherever business beckoned, without any fixed field of operations and without schedule or regularity’ (Hunter, 1969: 317). The vessel mostly ran between the lower Ohio and Mississippi River ports of Louisville, St Louis, and New Orleans, although it made occasional runs into the Missouri, Red, and Yazoo Rivers. Trade tended to follow seasonal patterns that corresponded to the harvest and shipment of bulk and processed agricultural commodities, although livestock and finished goods (both domestic and imported) were also part of the cargoes. The transportation of cotton from southern river landings to New Orleans was profitable business during the late fall and early winter months; the ability of steamboats to load great stacks of cotton bales never failed to astonish visitors to the west (Power, 1836: 100–103, Maximillian, 2008: 297–298). We know that on one occasion Heroine hauled 970 bales of cotton down from the Yazoo River (New Orleans Bee, 4 February 1832). The boat was part of both the mundane and the unusual events of its decade. Heroine played a contributing role in the Texas Revolution of 1836 by carrying volunteer troops and supplies to Red River ports and to New Orleans (Louisville Daily Journal, 12 May 1836; New Orleans Bee, 6 July 1832). Heroine and the hundreds of other steamboats operating on the western rivers in the 1830s were at the vanguard of a transportation revolution that was utterly transforming US society and the North American landscape, and they were widely celebrated for their many contributions to the economic development and population growth of the frontier states and territories (Hunter, 1969: 27–60; Howe, 2007: 203–242).
The social and economic benefits came at a high price, however. The 1830s western river steamboats had a justifiably bad reputation, as evidenced by their longevity and safety record. Lightweight construction and the rigours of river navigation resulted in most boats wearing out within five years of launching, but only if they first survived a gamut of man-made and natural hazards (US Treasury Department, 1838: 309–311). The greatest threat was posed by semi-submerged logs, called ‘snags’; these lurked like giant spears, with one end on the bottom of the river and the other at or just below the surface, waiting to tear open the bottom of a steamboat as it ascended against the current. Narrow river channels and lack of fixed rules regarding right-of-way and navigation lights meant that collisions between boats were frequent. The frail, heavily painted and tarred wooden superstructures of steamboats easily caught fire and then burned with incredible rapidity; two boats commanded by Diller's son-in-law Christopher Castleman, New Brunswick and Ben Sherrod, burned in 1833 and 1837, the latter with the loss of an estimated 60–70 lives, mostly passengers (New Orleans Bee, 28 October 1832; Louisville Daily Journal, 8 May 1871; Lloyd, 1856: 95–101).
Most frightening of all to the travelling public was a hazard both new and unique to steam navigation: catastrophic failure of the boiler system. This took two principal forms. The more common was a sudden and unexpected release of steam, caused by the collapse of a boiler tube or ‘flue’, the breaking of a cast-iron steam pipe, or the cracking of a cylinder head. The resulting release of high-pressure steam could fatally scald crew and passengers caught in the vicinity of the leak. More rare, but generally more horrific was a full boiler explosion, when one or more boilers over-pressurized and blew up. The effect was that of a giant bombshell: the blast and steam killed, cooked and maimed people, and left boats mangled or sunk (Hunter, 1969: 282–304). One of the most infamous explosions of the decade was that of Moselle at Cincinnati in April 1838, a disaster that killed an estimated 150 people, scattered bodies and limbs along the city waterfront, and sank the vessel (Fig. 14) (Fox et al., 1838). Moselle was but one of a series of dreadful steamboat accidents in early 1838 that spurred the passage by the US Government of the first federal transportation safety legislation later that summer (Brockmann, 2002).
Figure 14. The explosion of the Moselle at Cincinnati, Ohio on April 25, 1838, one of many steamboat boiler disasters of the 1830s. (Print courtesy of the Ohio Historical Society)
Download figure to PowerPoint
Despite its ‘very attentive’ officers Heroine was not immune to the dangers of boiler failure. On 4 October 1835, while the boat was working up the Mississippi River between Cairo, Illinois and St Louis, one of the flues collapsed, killing the engineer instantly, blasting three deck-hands overboard, and badly scalding three other people (at least one of whom later died). Among the dead was the captain's nephew, who was also named Jeremiah Diller (Daily Evening Herald, 6 October 1835; Ringwalt, 1877: 44). Damage to Heroine appears to have been minor, for the boat was back in operation before the end of the month (Daily Evening Herald, 29 October 1835).
By 1837 Heroine had been in service for five years, and was no doubt showing its advanced age (in June 1836 William Fairfax Gray already thought it ‘a sorry old boat’) (Gray, 1965: 186). Captain Diller was preparing his new boat Worden Pope for service and apparently sold Heroine to a man named J. R. Hord in late 1837 or early 1838. Hord specialized in the lower Mississippi River trade, and in January began running the steamboat on a schedule between Vicksburg and Natchez, Mississippi and Natchitoches, Louisiana, advertising Heroine as a ‘Red River Packet’ in the local newspapers (Natchez Daily Journal, 12 January 1838; Rodney Standard, 20 February 1838; Nardini, 1963: 155).
Heroine had not been in Hord's employ for very long before it was hired by the agent working for Columbus, Ohio businessmen-turned-government-contractors James Nisewanger and William Sullivant. The boat was to carry the annual supply of provisions to the US Army's Fort Towson in the ‘Indian Territory’ (present-day Oklahoma). The supplies included 240 barrels of salt pork, 500 barrels of flour, dried beans, salt, candles and soap (Nisewanger & Sullivant Contract, 1838). The Great Raft logjam had blocked the upper Red River prior to 1838, but that spring, after five years of steady clearing, Captain Henry Shreve finally opened a channel that allowed steamboats to proceed up the river as far as Fort Towson on the Kiamichi River (McCall, 1984). Heroine was one of the first boats to pass through the former barrier, and the boat successfully made a lengthy passage up the unknown and hazard-filled river as far as Jonesborough, Texas. From here the vessel attempted to navigate the last four miles (6.43 km) to the Fort Towson landing, but ran afoul of a snag on 6 May, sinking in shallow water until the main deck was awash (Nisewanger, 14 June 1838). After the sinking the crew and a detachment of soldiers from the fort were able to salvage the candles, soap, and some of the pork, but Nisewanger and Sullivant's uninsured cargo was largely ruined (Chick, 2011: 13). The crew also took off Heroine's engine and someone removed the boilers, but the hull was simply abandoned in the channel. It remained visible during low water for several more years, until a flood in 1843 shifted the course of the river and buried the hull under 7.62 m (25 ft) of sediment (Wittenbury, 9 January 1871; Stroud, 1997: 31–33). There it would lie forgotten until re-exposed by another flood in 1990 (Fig. 15).
Figure 15. The wreck of the Heroine, re-exposed by the Red River after being buried for 147 years under a cattle pasture. (Photo by Carrie Sowden, INA/TAMU)
Download figure to PowerPoint
Marine steam technology radically accelerated the pace of early 19th-century trade and communication and marked the beginning of a worldwide transportation revolution that continues to this day. Steamboats were found to be particularly useful on the vast Mississippi River system that drains the interior of North America. The shallow, fast-moving nature of western rivers forced steamboat inventors and builders to rapidly develop lightweight, shallow hulls and high-pressure steam propulsion systems capable of meeting the rigorous navigational demands of these waterways. The expanding fleet of cheap, fast, capacious steamers contributed greatly to the profound cultural and environmental transformations that took place in the mid-continent region.
The earliest years of steam on North America's western rivers are poorly documented in many respects, particularly as they pertain to the evolution of designs, material selections, assembly techniques, and operational parameters of both the hulls and the propulsions systems. The fortuitous circumstances that buried and subsequently re-exposed Heroine provided us with a rare example—the earliest thus far examined by archaeologists—of a western river steamboat. The wreck of Heroine constitutes a new benchmark for understanding what boat builders and engine makers accomplished during the first quarter-century of hull and machinery development. The wreck also provides us with a look at the human aspects of river travel by steamboat in the 1830s, at its living and working conditions, its advantages, and for many, its attendant dangers.