Introduction
Ship Construction
  Hull
  Main Deck
  Guard
Main Deck Features
  Bulwarks
  Windlass
  Hatches and Deck Openings
  Cabin Sole Plate
  Rudder
  Aft Cargo Hold and Lower Cargo Deck
  Engineering
  Boiler
  Engine
  A-frame
  Walking Beam and Connecting Rod
  Paddle Shaft and Paddle Wheels
  Architectural Window Glass
  Production
  Decoration
  Glazing

Findings
  Undecorated Glass
  Colored Glass
  Acid Etched
  Enameled/Etched Glass

Findings

Introduction

Natural processes and salvage attempts have stripped nearly every structural feature rising above the main deck from the Maple Leaf. Very few disarticulated timbers were found in the overburden and few if any can be traced to the superstructure. The main deck is intact except for damage near the bow and around the engineering spaces. The preserved portions of the vessel are the lower hull, internal arrangements covered by the main deck, parts of the steam power plant, and propulsion machinery. The following discussion is separated into five parts. First is a discussion of the Maple Leaf's construction including the wooden hull and other parts of the naval architecture. Second is a presentation of the ship’s equipment and features found on the main deck. Third is a brief discussion of the internal features found in the aft cargo hold. Fourth is a presentation dealing with the remaining steam engine and the mechanical drive train components. Last is a discussion of decorated and undecorated architectural window glass.


Ship Construction

Hull

The registered dimensions of the Maple Leaf are length 173.2 feet from the inner part of the stem to the fore part of the stern, beam 24.7 feet amidships and depth of hold 10.6 feet (Certificate of Ownership). In the field, overall hull length measured 184.6 feet from the front of the stem to the aft edge of the stern counter; the beam was 24.8 feet for the hull itself and 44.3 feet across the guards.

The hull is deeply buried so examination of structural components was generally limited to those found at main deck level. However, a centerline keelson was tentatively examined in 1994 through a hole in the lower cargo deck. It measured 12 inches sided and rose 25 inches above the ceiling planks.

At the bow, the stem rises 44 inches above the deck and has been badly damaged at the top by toredo worms. It is molded 1 foot 6¼ inches and sided 1 foot 1¼ inches. A horizontal 1 inch wide rabbet is cut into both sides and the back of the stem to create a ledge to support the bow cap rail. A wooden cutwater is fastened to the stem’s forward edge with iron drift pins. The cutwater's sided dimension is 9 ½ inches but the forward edge is so badly eroded precise measurement of the molded dimension is impossible.

The bow was reinforced by at least one breast hook located under the deck (Figure 17). Missing and damaged deck planking partially exposed the breast hook but not enough to permit measurements. Additional reinforcement was provided by two hawse pieces fitted to each side of the stem and attached to the breast hook below. They extend aft 8 feet 3 inches along the bow at deck level. Each is sided 13 inches, molded 11 inches and pierced by a fairlead and cast iron hawse pipe. The tops of the hawse pieces are mortised to hold the bow cap rail stanchions. The oval hawse pipes have an inside diameter of 7 ½ inches by 4 inches. The fairleads are located 2 feet 9 inches aft of the Fig 17 hawse pipes and measure 1 foot 2 ½ inches by 5 ½ inches.

Most of the hull has standard double framing. There are triple frames and one set of quadruple frames in the engineering space. This reinforcement appears limited to the engineering space to support heavy machinery. Futtocks are molded 5 ½ inches with a sided dimension ranging between 4 and 8 inches. The average sided dimension is 6 inches. Room and space dimensions varied depending on double or triple framing configuration. Space averaged 11 inches while room measured 13 inches for double frames and 16 to 19 inches for triple frames. Fasteners used to hold the futtocks together were not observed.

All frame tops end at main deck level in an arrangement where each pair of futtocks have staggered heights. This arrangement provides support for deck beams at the side of the hull. Generally, one futtock is cut 7 inches lower so a deck beam can rest on its top. The other futtock is cut even with the top of the deck beam so the deck planking is level. Aft of the paddle box, deck beams rest on top of the aft futtock. The arrangement was observed but not fully recorded forward of the paddle box.

Deck beams are sided 5 ½ inches, molded 7 inches, and spaced 18 inches apart on 24 inch centers. Deck stanchions along the keelson provide the beams with centerline support. One stanchion, removed for examination, was located at 137.5 feet on the baseline. It is 9 feet 2 ½ inches long. The lower half is square, with a 5 inch square base, while the upper 4 feet 11 inches is round and machine turned. The top fits into a rectangular mortise on the underside of a deck beam. Stanchions were not found under every deck beam but no interval was determined.

The sheer strake is notched to allow deck beams to pass through the side of the hull, extending to the guard beam on the sponson edge. This arrangement extended the main deck to the maximum width of the paddle wheels and created the characteristic sponson hull. Notches in the sheer strake get progressively deeper moving aft along the hull. From 1 inch forward of the paddle wheel to 7 inches aft of the wheel. All deck beams examined on the starboard side have broken at the side of the hull.

In addition to the support provided by the futtocks, the beams rest on top of a large shelf clamp inside the hull. The shelf clamp is 5 inches molded and 27 inches sided. It is made of two thick planks, one on top of the other, measuring 12 and 15 inches wide respectively. Ceiling planks on the interior are 2 ½ inches molded and butt against the lower edge of the clamp. Outer hull planks measured 2 to 2 ½ inches thick but width and fastening patterns were not recorded.

The knee pattern observed inside and outside the hull further supports deck beams as well as providing longitudinal stiffening to the hull. A hanging knee was observed under a deck beam on the exterior hull just aft of the starboard paddle wheel. This may indicate each deck beam is supported by a similar hanging knee.

Dagger knees are fastened to every other deck beam inside the hull (Figure 18). Their pattern suggests they are part of the longitudinal reinforcing system. Forward of the paddle box, the top of the dagger knee leans aft and fastens to the forward side of the deck beam. The lower leg of the knee butts up under the adjacent deck beam. Aft of the paddle box, the pattern is reversed so the knees angle forward. One disarticulated dagger knee examined on the surface was 4 ½ inches wide, 27 inches long on the beveled top fig 18 edge and 41 ½ inches long on the lower portion. Square headed iron bolts, 1 inch in diameter, were used to fasten the knee to the ship; two on the deck beam and three on the shelf clamp.

Four hanging knees are located amidships in front of the boiler. Damage to the engine room caused by demolition destroyed the deck, deck beams, and many knees in this area. The remaining hanging knees are spaced on 24 inch centers. There is a gap between this group of hanging knees and the dagger knees located forward making it impossible to determine exactly where the kneeing pattern changed.


Main Deck

As the Maple Leaf's superstructure disintegrated, the main deck became the top surface of the site. Deck planks measure 2 ½ inches thick and average 6 inches wide. They are fastened to deck beams with two iron spikes placed in a diagonal pattern. Each ¼ inch square spike is set in a counter sunk hole and capped with a wooden bung. Although the deck was built for strength and much of it is intact, several areas of extensive damage have ruined the deck's original structural integrity. Only the sediment filling the interior spaces keeps the deck from collapsing.

The forward deck is extensively damaged along the starboard side approximately 15 feet aft of the bow (Figure 14). This area roughly corresponds to the main impact of the torpedo explosion concentrated on the starboard side, about thirty feet from the stem. Reportedly, "the hog frame was broken and the whole side of the vessel was stove in" (Farnham 1864).

Another damaged area is located between the forecastle hatch and the forward cargo hatch. Deck planking has been stripped off and several deck beams are broken. Barnacles growing on material recovered from the hold indicate planking was removed before the vessel filled with sediment.

The largest damaged area is over the engineering spaces. Between 70 feet and 105 feet on the baseline, much of the deck is missing or severely damaged (Figure 15). The open space is littered with broken timbers, including the forward paddle beam, and the starboard boiler lay exposed. Damage continues aft of the paddle shaft to approximately 120 feet on the baseline. The aft paddle beam and many deck beams are broken while some planking is loose or missing. This destruction probably occurred in 1883 when Ross removed the crank frame and gallows frame, and allowed the walking beam to fall through the deck causing severe damage along the ship's centerline Aft of 120 feet, the main deck is intact to the stern. The hole cut through the deck in 1988 to gain access to the aft cargo hold is centered at 139 feet on the baseline. It begins on the centerline and extends to the port side, measuring approximately 8 feet athwart ship and 42 inches bow to stern. It is covered with an aluminum frame and Plexiglas panels.


Guard

The longitudinal sponson hull is a distinguishing characteristic on nineteenth century eastern North American steamboats. The structure is an evolutionary development that started as a box frame extending from each side of the ship to house the paddle wheels and support the ends of the paddle shaft. Eventually the frame’s outer edge was lengthened fore and aft and tapered inboard to meet the stem and stern. The overhanging sponson, or guard, often added twenty feet or more to the vessel's main deck beam depending on the width of the paddle wheels (Cuthbertson 1931:243-244; Russell 1861:108,113).

On the Maple Leaf, the frame can be viewed as a massive structure that overlies the hull. The fore and aft ends are composed of two paddle beams while the sides are framed by the guard beams. The hull passes through the middle of the frame and also forms the interior side of each paddle wheel opening. The guard beam is 9 inches molded and 12 inches sided aft of the paddle wheel. A wooden rub rail is fastened to the exterior side. The aft paddle beam is made of two timbers fastened together one in front of the other. The larger forward timber, 16 inches sided and 10 inches molded, extends completely across the vessel to both guard beams. The smaller timber, sided 8 inches and molded 10 inches, ends inside the hull and rests on the shelf clamp. Hanging knees on the interior and exterior support the large beam as it passes through the hull. The exterior knee is extremely large. The top extends horizontally from the hull to the guard beam.

Wooden trusses built on the outer edge of the guards supported the ends of the iron paddle wheel shaft. The starboard truss was built as a symmetrical A-frame (Figure 19). It rises approximately 6 feet above main deck level on timber legs measuring 10 ½ inches high and 8 ½ inches wide. The top is 6 feet 8 inches long and 13¼ inches wide. A large square notch centered in the top, 34 inches long and 5 ½ inches deep, once held a pillow block to secure the end of the paddle shaft. The pillow block is no longer present fig 19 and the paddle shaft has been dislodged from its original position. The port pillow block is still in place on the port side shaft support and is discussed below with the paddle shaft. The ship’s main deck is expanded over the entire guard by extending the deck beams out of the hull to the guard beam. Deck construction is identical to the rest of the main deck. Planks measure 2 ½ inches thick, average 6 inches wide, and are fastened in a diagonal pattern. Figure 16 shows a narrow opening, 22 inches wide, behind the paddle beam just before the deck planking begins. The purpose of this opening is not clear but it may represent a way for water picked up by the paddle wheel to drain out of the paddle box.

Most of the starboard guard was examined except the forward end. It extends a maximum of 9.6 feet from the side of the hull to the outer edge at the paddle wheel. The rectangular paddle wheel opening in the guard is located amidships, centered 102 feet from the stem. It extends from 86 feet to 118 feet aft of the stem with an overall length of 32 feet. The inside width of the opening is 8 feet 6 inches.

The guard's general condition is very good and most of the deck planking is still intact. However, those deck beams that once extended from the side of the ship to support the guard have all broken where they pass through the hull. As a result, the guard sags at a 10 to 15 degree angle (Figure 20). Also, late nineteenth century demolition caused a fragment of the walking beam to fall on the aft paddle beam, breaking a section out of the center. The beam is also broken at the edge of the hull.


Hogging Truss fig 20 Portions of the hogging truss found on the starboard side represent one major element of the longitudinal reinforcing system. One of two built into the ship, the truss's overall length was 131 feet extending from 24 feet to 155 feet aft of the stem. Although both lower ends of the truss are present, the upper portion is missing. In discussing the Maple Leaf's loss, Second Officer Charles Farnham stated the explosion broke one of the hogging trusses (Farnham 1864). Additional damage probably resulted when Roderick Ross cleared wreckage for safe navigation. Inclement weather and bad visibility made detailed recording of the truss’s forward end impossible. The aft truss section came loose as supporting sediment was removed. It was subsequently recovered for documentation (Figure 21).

The aft truss segment is built on a sole timber, or base, which measures 10 ½ inches wide by 10 inches high and 17 feet long. It is badly eroded. The sole timber was fastened over the frame tops at the top of the hull. The bottom has notches to fit over frame tops and deck beams as they pass through the hull. Iron drift pins fastened the sole timber to the shelf clamp and deck planking butts against the side. Wooden chocks fitted between the deck beams and resting on the shelf clamp distribute the weight of the truss more evenly. One of these chocks is still fastened to the bottom of the sole timber by a drift pin. Several similar blocks of wood were found unsecured in the space between the deck beams.

The arching chord timber is joined to the sole timber with scarphs and iron drift pins. The attachment is strengthened on top by a large knee and below by a small wooden chock located under the chord. The chord fragment measures 8½ inches wide by fig 21 12 inches high. Originally, the chord was composed of many segments joined together to form an arch. A notch measuring 3 inches wide and 50 inches long on the upper end of the fragment is the mortise joint used to attach the next section of the arch. The mortise is filled by a tenon fragment from the next chord section.

The Maple Leaf presumably had a transverse support system to keep the guards from sagging and the trusses from swaying. The system probably used horizontal tie rods running between the top of each hogging truss with diagonal rods linking the trusses to the outer edge of the guard. Using turnbuckles on the tie rods to create tension, the system would stiffen the vessel transversely. A short tie rod, measuring 4 feet long and 1 ½ inches in diameter, is attached to an eye bolt on the paddle beam and is probably the diagonal rod linking the guard to the hogging truss. A broken turn buckle is fastened to the end of the rod. Another section of rod laying on the deck nearby may also be part of this system. One end is attached to an eye on a metal plate that bolted the rod to a timber, possibly the hogging truss (Bates 1968:23).


Main Deck Features

Bulwarks

A low bulwark remains intact along the aft starboard guard extending from the paddle wheel to the stern. Originally, this structure stood 32 inches high and marked the edge of an exterior walkway providing access to cabins on the aft main deck. A similar bulwark was not found forward of the paddle wheel with the exception of a cap rail at the bow.

The aft bulwark is intact except for damage incurred by natural deterioration and possibly by demolition of the superstructure. The facing planks nailed to the bulwark stanchions are falling off as the iron fasteners corrode. Most of the starboard cap rail has been pulled off and now lays in the mud above the deck. The cap rail at the stern and parts of the bulwark were removed by SJAEI in 1991.

The bulwark is constructed with stanchions planked on both sides and topped with a cap rail (Figure 22). Stanchions are fitted through the waterway into the edge of the guard with an average spacing of 36 inches. Each stanchion is roughly 32.5 inches tall, 3 ½ inches square at the base, tapering slightly towards the top. A tenon at the top of the stanchion fits into a mortise on the underside of the cap rail. The assembly is secured by a ½ inch diameter iron drift pin driven through the joint at an angle. The cap rail is 6 inches wide and 1½ inches thick. The underside is grooved along the outboard edge to accept horizontal planking used to panel the exterior side of the bulwark. Horizontal tongue and groove planking, 4⅞ inches wide and ½ inch thick, covers both sides of the bulwark. The top of the cap rail is mortised to accept stiles that helped support the upper deck. Although no stiles were found, mortises in the cap rail indicate they were placed at irregular intervals, more than 36 inches apart.

One wooden fairlead is located in the bulwark, at 155 feet on the baseline. It is made from a solid block of wood measuring 43¾ inches wide, 10 inches high and 8 inches thick. The oval opening through the center is 17 inches wide and 4 inches high.

A gangway opening to board the ship is located in the bulwark just aft of the paddle wheel (Figure 16). Originally the 55 inch wide opening was framed by a fig 22 stanchion post on each side. The forward stanchion is missing but its location is marked by a mortise in the waterway. The aft stanchion is standing but the top is eroded off 70 inches above the deck. The lower portion of the stanchion is rectangular, measuring 6 by 3 inches, with a rabbet on one corner. This appears to be a stop for the gate, or door, used to close the gangway opening. The rectangular section rises 36 inches above deck, indicating the height of the door, and becomes round for the rest of its length.

The cap rail found at the bow in 1989 was removed for conservation and display at the Jacksonville Museum of Science and History. Documentation of this piece took place at the museum (Figure 23). The V-shaped structure is composed of two cap rails fastened together at the apex by a horizontal lodging knee. This configuration forms a square mortise that fits around the stem and rests on a narrow 1 inch wide ledge cut into the stem. Each arm of the V is a rail constructed of two timbers laid on top of each other and fastened together with iron bolts secured on the underside with square nuts. Crude chocks are cut into the top timber of each rail. Damage by wood boring mollusks is evident on the forward ends of each rail (Cantelas 1992).


Windlass

The windlass is located 8 feet aft of the stem and is the largest feature on the forward deck (Figure 24). The wooden barrel is mounted on two carrick bitts held in place by cheeks and braced forward by large knees. A strongback made of 1 inch diameter iron rod braced the carrick bitts laterally against each other but is now broken. The pawl post is located just forward of the windlass and rises 5 feet 2 3/4 inches above fig 23 fig 24 the deck. This massive timber measures 18 inches square and is likely stepped into the bow deadwood (Chapelle 1973:602). The pawl, a ratchet used to stop the windlass from turning in reverse, is missing from the aft face of the pawl post (Paasch 1908:121-122). The windlass operated manually by inserting hand spikes into square holes along the barrel and turning the barrel by hand.

The windlass provided lifting power for a number of tasks but its primary purpose was to raise the anchors carried at the bow. As anchor chain was brought aboard it passed over the windlass barrel and through the deck into the chain locker below. Normally, anchor chain passed through the deck by way of an iron chain pipe. Chain pipes are absent on the Maple Leaf and no other openings were noted near the windlass. However, two unusually wide deck planks located just aft of the windlass whelps may have been removable to provide access to a chain locker below.


Hatches and Deck Openings

Four hatches are located on the main deck. These allowed entry to the forecastle cabin, two cargo holds, and the engine room. The forecastle hatch is a small opening 23 feet 7 inches aft of the stem which provides passage to the forecastle cabin. The opening measures 4 feet athwart ship and 2 feet 3 ½ inches fore and aft. The coaming rises 6 inches above the deck and is 4¼ inches wide. The 1856 Maple Leaf ambrotype shows a small square structure on the weather deck in the area of this hatch that probably covered a companionway leading from the upper deck to the forecastle (Figure 25). Just aft of the forecastle hatch on the starboard side is a 6 inch circular hole passing through the deck. Fig 25 The edge is cleanly cut with no sign of wear and no evidence of a deck fitting mounted to the hole.

The badly damaged forward cargo hatch is located 43 feet 10 inches aft of the stem. Three sides of the coaming remained in place and the missing aft coaming was found on the deck a short distance away. The deck beam supporting the forward coaming is broken at the hatch. Originally, the opening measured 3 feet 7 inches athwart ship and 3 feet 7 inches fore and aft.

Part of the hatch was temporarily recovered to examine construction details (Figure 26). The same type of general construction methods are used on all the vessel's hatches. First, a section of one deck beam is removed along the ship's centerline to create a large opening. Carlings placed along the side of the opening support the cut ends of the deck beam. Next, the timbers forming the hatch are secured in place. Coaming timbers rest on top of the carlings along the side of the opening and head ledge timbers sit on deck beams at each end. The component timbers are joined at the corners with a lapped rabbet joint and fastened to the underlying deck beams and carlings with iron drift pins. Each coaming has three mortises on the interior face to fit strongback timbers that supported the hatch cover.

A large hatch used to enter the engine room is located 64.5 feet aft of the stem (Figure 15). The opening measures 5 feet fore and aft and 6 feet athwart ship. It was used to load wood into the bunkers to fuel the boilers when the vessel operated on the Great Lakes. The wood coaming is badly worn from loading fuel despite a metal strap placed around the top edge for protection. Two deck beams are sistered together on the fig 26 forward side of the hatch and may represent a repair or reinforcement. The forward beam continues across the ship, supporting the forward hatch coaming, while the aft beam is cut short and butts against the side of the hatch. Beams forward and aft of the pair still retain the normal 18 inch spacing typical for this vessel.

A large circular cast iron through-deck fitting is located on the starboard side of the engine room hatch. The interior diameter is 1.3 feet. This appears to be part of the engine room ventilator system. The 1856 ambrotype shows a large ventilator funnel over the engine spaces on the starboard side.

Remains of the aft cargo hatch were found approximately 115 feet on the baseline. This area was badly damaged when the walking beam crashed through the deck during demolition work in 1883 (Figure 16). Only one disarticulated hatch coaming was found in the damaged area. It is 32 inches long and notched to accept three strong backs. The hatch was located between the crank frame timbers just forward of the paddle beam.

A bulkhead below the main deck, at 116 feet, separates the aft cargo hold from the engine room. The bulkhead stops on each side at the crank frame timbers creating a gap in the center. The location of the bulkhead further suggests the hatch location between the crank frame timbers at the forward end of the cargo hold.

Two small through deck openings were found on the aft deck. A small circular hole is located 141 feet aft and 8 feet starboard of the centerline. The 6 inch diameter hole has a cleanly cut edge with no sign of wear and no evidence of a deck fitting mounted through. A sandstone through-deck fitting is positioned at 165 feet on the baseline and 1.5 feet to starboard. The fitting is flush to the deck, 18 inches square with a 7 inch diameter hole through the center. The purpose of this feature is not known. The use of stone suggests it may have insulated the wooden deck from a hot steam or water pipe used to heat the upper cabins.


Cabin Sole Plate

Evidence of the superstructure and cabins is nearly absent over most of the site. However, on the aft deck several cabin sole plates were located, marking the position of cabin walls. Figure 16 shows their locations. Two different types of plates are present. Both are roughly 2 inches high by 4 inches wide fastened to the deck with nails. One type is a plain unmodified rectangular board. The other type has a concave surface on each side.

Except for one curving plate near the stern, all plates lie on the ship's longitudinal axis. Both interior and exterior walls appear to be represented. The curving plate at the stern is an exterior wall and demonstrates the shape of the after most cabin. The straight plate branching off near the starboard end is an interior wall. Another short segment of plate near the starboard side at 160 feet on the baseline is probably for an exterior wall and is notched to accept wall studs on 24 inch centers. A 12.5 foot long section of unmodified plate is located near the forward end of the aft deck. It is 8 feet starboard of the ship's centerline suggesting it is an interior wall.


Rudder

The rudder is still attached to the wreck indicated by the rudder post passing through the deck near the stern. The post is 1 foot in diameter and broken off 7.2 feet above the deck. It rose through the saloon deck above, and was probably rigged with emergency steering tackle.

On the main deck, two short, wide boards fit around the rudder post. The purpose of the boards is unknown but removing them may provide access to steering gear below deck. Aft of the rudder a small rectangular wooden block is attached to the two planks. It has a shallow rectangular mortise in the top but its function is unknown.


Aft Cargo Hold and Lower Cargo Deck

The internal architecture of the aft hold appears undamaged. However, the small area open for examination, centered 139 feet aft of the stem, allowed only a tentative description of interior construction. The deck and deck beams covering the aft cargo hold are intact but have suffered severe structural damage. All deck beams observed in the confined space are broken approximately 3 feet from the port side causing the central portion of the deck to sag 2.5 feet at the break. Removing sediment from the cargo space during excavation may have caused the sag to increase.

The main feature inside the hold is the lower cargo deck built approximately 4 feet above the bilge ceiling and 2 feet above the keelson (Figure 27). The cargo is packed on this lower deck. The deck is lightly built and planked with tongue and groove boards 4¼ inches wide and⅞ inch thick laid longitudinally. The planking is nailed to 2 by 3 inch deck beams spaced on 24 inch centers. Short centerline deck stanchions, resting on the keelson, support the beams. Each stanchion is rabbeted on the ends to fit around the fig 27 corners of the keelson and deck beams. These precarious joints are secured with a few small nails. The lower deck construction on the Maple Leaf’s is almost identical to the steamboat Commonwealth (Figure 28).

Main deck stanchions pass through the lower deck and are mortised to main deck beams. The lower deck planks are fitted around the stanchions and not attached. One stanchion recovered for examination exhibited rope wear marks, presumably caused by securing or moving cargo in the hold.


Engineering

Typical of side wheel steamers, the Maple Leaf's engine room is located amidships between the paddle wheels. Evidence from the site indicates the A-frame that supported the walking beam was located forward of the paddle shaft. It follows that the engine, including the condenser, air pump, and hot well, was placed in front of the A-frame. A boiler is located on each side of the A-frame and extended aft under the paddle shaft. Fuel bunkers are probably forward near the engineering hatch along the side of the hull. For reference in the following discussion, Figure 29 illustrates the general configuration of a marine walking beam engine.

The 1993 investigation in the engineering spaces located one boiler, the paddle shaft, the paddle shaft connecting rod and walking beam fragment, A-frame timbers, and fragments of the air pump. The engine is missing, including the engine cylinder, condenser, hot well, and valve chest. Timbers and other debris fill all open spaces of the engine room restricting access to lower areas. This material was left in place limiting fig 28 fig 29 documentation to the upper levels of the engine spaces. Each major component will be discussed separately.


Boiler

Historical records and the 1856 ambrotype indicate the Maple Leaf had two cylindrical boilers placed on either side of the A-frame (Figure 25). These boilers were new when installed at the Kingston Marine Railway Yard in 1851, and repaired in Toronto in 1858 (DBW March 26, 1851; Girvin 1993:89). Although only the starboard boiler was examined, the port boiler is considered to be identical in design. The 7 foot diameter starboard boiler begins at 86.5 feet on the baseline. The aft end of the port boiler was found at 114 feet on the baseline, indicating both boilers are 27.5 feet long and pass under the paddle shaft. They are return, fire tube boilers with the firebox facing forward.

The remains of the starboard boiler smoke stack are 2.5 feet in diameter and centered 6 feet aft of the boiler face. The stack is encircled by a steam dome on top of the boiler. The steam dome is badly damaged suggesting it may have suffered an explosion. Shredded metal plating flares out away from the dome's center indicating a strong outward blast. This blast probably caused the boiler face to separate and lean forward, creating a 5 inch gap. The fire box doors on the boiler face were found open, offering further evidence of an explosion.

The starboard boiler is badly corroded and large debris fills the open space in front. These obstacles limited examination to the upper port side of the boiler face. Figure 30 is a reconstruction of the face combining data gathered in the field with historical information on similar boilers. Dashed lines on the starboard side and lower portions are a hypothetical reconstruction.

A glass water gauge and four gauge cocks, on the upper right side of the face, indicated water level in the boiler. Water level had to be maintained above the fire tubes while the boiler operated. Otherwise, fresh feed water touching the hot tubes flashed into steam, dramatically increasing boiler pressure and possibly causing an explosion (Ward 1860:23).

The Maple Leaf's glass water gauge is a ball float type designed with a lower vertical pressure chamber and an upper sight glass (Figure 31). The pressure chamber is open to the boiler at the top and bottom through short brass fittings, allowing boiler water to maintain an equal height in both vessels. Inside the pressure chamber, a hollow brass ball floats on the water surface. A brass wire soldered to the ball passes through a leather gasket at the top of the pressure chamber into a thick sight glass tube. Water level in the boiler is indicated by the wire height on a scale mounted behind the sight glass. The thick glass tube is mounted in a brass holder and held in place by screw tension. A scale mounted on the holder, behind the tube, is marked in one inch increments beginning with one on the bottom and nine on top. A cock on the bottom of the pressure chamber drained excess water.

Many nineteenth century glass water gauges omit the pressure chamber completely. The sight glass is directly connected to the boiler at the top and bottom by metal fittings. For safety, these fittings have stop cocks to turn off pressure to the sight fig 30 fig 31 glass if it breaks (Lardner 1848:116). The Maple Leaf's glass water gauge does not have stop cocks to isolate the pressure chamber. It is only open to the atmosphere at the top, through the sight glass. The pressure chamber is designed to seal itself if the glass breaks. The solder holding the wire to the brass ball float is shaped to fit a conical depression at the top of the pressure chamber. If the glass broke, boiler pressure forced the brass ball against the top of the pressure chamber causing the solder to seal the opening.

While a glass water gauge gave an accurate measure of water level, it was prone to fouling and breakage. Gauge cocks provided a reliable method of checking water level with slightly less accuracy (Figure 32). Four brass gauge cocks were mounted diagonally on the port side of the glass water gauge. The top and bottom cocks are 8½ inches apart vertically and align horizontally with the top and bottom openings on the water gauge's pressure chamber.

Placement of the cocks indicates the highest and lowest water levels allowable to permit safe boiler operation. "The highest cock is at a point above which the water cannot be permitted to rise without encroaching upon the room provided as a reservoir for steam; and the lowest cock is at a point below which the water cannot be permitted to fall, without endangering exposure of the flues" (Ward 1860:23). The two intermediate cocks show water between high and low levels. In operation, if one of these cocks is opened and emits steam, water level is below that point. If the cock emits water, water level is above that point (Ward 1860:23). The gauge cock shown in Figure 32 is in the closed position. Fig 32 Large debris hindered investigation of the boiler face below the gauge cocks. Except for the port side fire door, much of the lower area could only be examined by feeling around the debris. Openings for two fire doors are spaced 28 inches apart. Each door hinged outward away from the center of the boiler and was fully open when examined. A heat baffle on the inside of the port door indicates the fire box opening measured 16 by 12 inches. Mud filled the fire box so the interior was not examined.

An opening detected 30 inches below the fire door is probably an open ash door. Obstacles made further examination impossible. The overall height of the fire box was not measured because the bottom of the fire box could not be found.


Engine

The field investigation did not locate any intact components of the engine suggesting it was removed or destroyed. Unfortunately, historical sources only provide a brief description. The piston had an 11 foot stroke and a diameter of 52 inches (Heyl 1967:171). The stroke length is verified by the crank which measures 5.5 feet between the centers of the crank pin and paddle shaft.

Pieces of the engine were found in a deep excavation located in the center of the ship, 87 feet from the stem. The excavation uncovered two cast iron cylinder fragments eight feet below the baseline. One fragment was removed for conservation and analysis. This fragment is part of the air pump cylinder and includes a flange and brass compression ring. The compression ring indicates the fragment is from the top or bottom of the cylinder. It attached either to the bed plate at the base or to the hot well at the top. The cylinder diameter was approximately 24¼ inches and wall thickness 1 inch. Deep vertical striations mark the interior of the cylinder wall. These striations pass through an unidentified filler metal used to fill small voids and imperfections in the cast iron. This indicates the striations occurred during use. They apparently resulted when particles in the water pumped from the condenser became trapped between the piston and cylinder wall.


A-frame

Located on the ship’s centerline, the A-frame, or gallows frame, functioned to support the walking beam, linking the engine to the paddle shaft. It also incorporated the crank frame to support the paddle shaft. On the Maple Leaf, the massive structure was composed of large wooden timbers held together with a complex system of iron tie rods. It was removed by Roderick Ross in 1883.

Many twisted and broken tie rods lie between the paddle shaft and the front of the boiler to mark the A-frame's position. One timber forming the forward starboard leg is located in front of the boiler and measures 16 by 12 inches in cross section. It is broken off below the river bottom and slopes down towards the bow.

Two timbers mark the crank frame at 114.8 feet on the baseline, just behind the boilers. They measure 9 by 14 inches and are spaced 6.7 feet apart. The tops are sawn or broken off flat just below the river bottom mud line. Both timbers pass aft of the bulkhead separating the engine room from the aft cargo hold. They slope down into the hold, towards the stern, at a 45 degree angle.


Walking Beam and Connecting Rod

A cast iron walking beam, pivoting on trunnions atop the A-frame, transferred power from the engine to the paddle shaft. The piston rod imparted a reciprocating, or up and down, motion to the forward end of the beam. This motion was transferred to the paddle shaft connecting rod on the aft end of the beam, which in turn rotated the cranks on the paddle shaft. This rotary motion caused the paddle wheels to turn.

A fragment of the walking beam is connected to the paddle shaft connecting rod and paddle cranks (Figure 33). Apparently, Ross's demolition caused the beam to break and fall aft, pivoting on the crank pin, carrying the connecting rod with it and smashing through the main deck. Now, the connecting rod rests horizontally and the walking beam fragment hangs vertically below it. The fragment is 8.7 feet long.

The walking beam is solid cast iron and had a long, narrow elliptical shape. Cross section views of the beam in Figure 33 show a center web with a flange around the circumference and another longitudinally down the center. The flange stiffens and strengthens the central web. A wrought iron band placed around the circumference of the beam provided additional strength and reinforcement. The band has been removed but four U-shaped wrought iron band guides are cast into the beam to hold the band in place. They are paired on opposite sides of the beam.

Two pins on the beam fragment are attachment points for connecting rods. These pins are centered 7.2 feet apart. The small pin near the broken end of the beam powered auxiliary equipment, possibly a boiler feed water pump. The connecting rod strap has been dismantled and the rod removed leaving the 2½ inch diameter pin intact. fig 33 The second pin attaches the large paddle shaft connecting rod to the end of the walking beam. This pin is 5 inches in diameter and flares on the end to hold the bearing blocks in place. The journal assembly connecting the rod and the beam is intact on the starboard side but has been taken apart on the port side. Apparently, there was an unsuccessful attempt to disconnect the rod before the beam was broken and fell into the river.

The intact journal on the starboard side is a typical nineteenth century example. The connecting rod forms a yoke to meet the pin extending from each side of the walking beam. On both sides of the walking beam a two piece brass bearing fits over the end of the pin and rests on top of the yoke. Next, a pin strap fits over the pin and bearing. Slots on each end of the strap align with a slot in the connecting rod below the bearing. The assembly is secured by a placing a tapered key and gib through the aligned slots. A locking bolt holds the key in place. An oiling cup, normally found on top of the journal to lubricate the bearing, is missing.

The paddle shaft connecting rod is wrought iron and lies on the centerline of the ship approximately 3 feet below the river bottom. It measures 18 feet 5 inches between the crank pin and the walking beam pin and is 6 inches in diameter. On the lower end, the rod is still connected to the crank pin which is 7¼ inches in diameter. The rod is fastened to the crank pin with a strap in the same fashion it is attached to the walking beam.


Paddle Shaft and Paddle Wheels

The paddle shaft is located slightly aft of amidships and is the largest feature on the site regularly exposed above the river bottom. The structure is actually two shafts, one on the port and one on the starboard side, joined in the middle by two cranks and a crank pin. The baseline passes through the cranks at 102 feet. These three sections will be dealt with separately.

Each wrought iron crank was attached to the paddle shafts by heating and shrinking over the ends of the shafts (Ward 1861:105). The overall length of each crank measured 6 feet 10 inches with a maximum thickness of 11 inches. The distance between the center of the paddle shaft and the crank pin center is 5 feet 6 inches, indicating a piston stroke of 11 feet. This stroke length confirms historical documentation (Heyl 1967:171).

The cranks were originally connected together by a crank pin. However, Ross’s demolition work pulled the starboard crank off the crank pin. The crank pin is normally attached to only one crank and rides free in the other. This arrangement is called a drag link and allows the paddle shaft to shift as the guards settle without straining the cranks (Ward 1860:105-106). On the Maple Leaf, the crank pin attached to the port crank is 7¼ inches in diameter.

On the port side paddle shaft two narrow straps encircle the shaft adjacent to the crank and probably mark the location of one pillow block mount. The straps are 1½ inches wide and are spaced 12 inches apart. This is the only feature noted on the port shaft.

Port side shaft diameter decreases in stages from the cranks to the outboard end. The shaft diameter is 11 inches between the cranks and paddle wheel flange. Diameter between the flanges is 10 inches and further reduces to 8 inches on the outboard side. The end of the shaft secured by the outside pillow block is 7 inches in diameter. In 1883, Ross's workers removed the pillow blocks holding the two shafts in place and moved them from their mountings (Russell 1883). They pushed the starboard end well aft while the port shaft shifted only slightly aft of its original alignment. The outside pillow block remains on the port guard but all others are missing. It contains a two piece brass bearing to fit the end of the 7 inch diameter paddle shaft.

The port shaft is broken 14 feet from the crank, near the outboard end. The short broken section is 5 feet 10¾ inches long and located slightly aft of the main shaft. This piece carried the radial paddle wheel and includes two paddle wheel flanges, spaced 30 inches apart, with wooden paddle arms. The upper portion of both flanges are broken off along with the paddle arms. The lower half, with paddle arms attached, projects into the bottom.

The starboard shaft is complete and undamaged despite being pushed aft on the outboard end. This shaft is actually two pieces joined by a large coupling located 4 feet from the crank (Figure 15). The coupling consists of two flanges that fit over the end of each shaft segment. They butt together and are fastened around the perimeter by large bolts. The Maple Leaf broke its paddle shaft shortly after leaving the port of Charlotte, New York on July 16, 1859. The local Rochester, New York newspaper mentions the installation of a new shaft (Union and Advertiser July 16, August 1, 1859). It seems instead, the old shaft was fixed or a new section added with the use of the coupling. Both the port and starboard shafts were manufactured at the same time by the Kingston Foundry when the Maple Leaf was built. The port shaft was forged as one section indicating the starboard shaft was also originally one piece. The coupling is evidence of repair work.

Two eccentric bands hanging loose from the starboard paddle shaft on each side of the coupling are the only remaining parts of the engine valve gear. They are badly bent and the eccentric arms are broken off. Their condition made it impossible to determine the throw, or horizontal distance, traveled by the eccentrics. Each band is slightly different in size, possibly due to corrosion or concretion formation. The outboard band is 1¾ inches wide and 1¼ inches thick. The inboard band is 2 inches wide and 1 inch thick. No eccentrics or eccentric bands were found on the port paddle shaft.

Two cast iron paddle wheel flanges are attached near the end of the shaft to hold the paddle arms of the starboard wheel. They are spaced 50½ inches apart on center. The inside flange is broken in half leaving the lower portion in place with paddle arms still attached. The lower ends of all the arms are deeply embedded in the bottom. The outside flange is extensively damaged. All of the perimeter has been broken away along with the paddle arms, leaving only the central hub.

Comparing the spacing between the flanges on the port and starboard sides reveals a large discrepancy. Spacing on the port flanges is 30 inches and on the starboard flanges 50½ inches. The difference in spacing is due to damage incurred during demolition work. All flanges are badly damaged and appear to have shifted on the paddle shaft.

The Maple Leaf had simple radial paddle wheels. The intact lower portions of both paddle wheels were inaccessible. However, disarticulated fragments found during excavation allow a hypothetical reconstruction of the port paddle wheel (Figures 34 and 35).

Originally, the cast iron flanges were 6 feet in diameter with 24 arm pockets to hold the paddle arms. The method used to attach the flanges to the shaft was not determined. Two paddle arms and a bucket were examined on the port side aft of the shaft. The 3 by 7 inch arms, or spokes, are broken off the flange but had an overall length of approximately 14 feet. The proximal ends are tapered to fit the arm pockets on the flange and are attached with two bolts. Notches on the distal ends of the arms probably held a stiffening or reinforcing band around the circumference of the wheel. Two additional reinforcing straps once circled the wheel as indicated by iron strap fragments and fasteners on the paddle arms. Although some sort of cross bracing certainly existed between the paddle arms to provide transverse stiffening, no evidence of this bracing was found.

The paddle bucket attached to the end of the paddle arms. One badly eroded bucket recovered for examination measured roughly 5 feet 2 inches across the blade and 18 inches wide. It is made from two planks attached to the paddle arms by iron fasteners. The fasteners were missing but the holes measured 1 inch in diameter. The fasteners indicate the paddle arms were attached near each side of the bucket and spaced 52 inches apart.

The bucket’s width indicates the overall width of the paddle wheel was 5 feet 2 inches. The wheel fit into an 8 foot 6 inch wide opening in the guard with 20 inches of fig 34 fig 35 clearance on each side. The spacing of the paddle arms, 52 inches apart, is also the original spacing of the paddle flanges on the paddle shaft. This measurement implies the flanges shifted as a result of demolition work. The combined measurements of flange diameter and length of the paddle arms suggests the paddle wheel diameter was approximately 29 feet.
A wooden piling rests over the entire length of the port shaft extending north 35 feet from the crank. It is probably a piling that once marked the site as a navigation hazard but was later pulled out of the bottom and dropped on the site. It has been left out of the site map (Figure 13) to provide an unobstructed view of the shaft.


Architectural Window Glass

When built in 1851, the Maple Leaf incorporated many features common to steamboats on Lake Ontario and the Great Lakes. Besides gross components, such as the hull and steam machinery, many delicate decorative elements added to the ship's personality, including decorative window glass. A contemporary source praised the glass artist and designs he used on the Maple Leaf. “We are particularly pleased with the profusion of stained glass, tastefully and elaborately painted by our friend Mr. E. C. Bull, whose skill has covered every glass door and window with pretty little sketches enwreathed with maple leaves, which would form quite a study for the youthful artist” (DBW 26 March 1851).

This analysis includes both decorated and undecorated window glass fragments recovered from the steamboat's aft deck in 1994. Initially, window glass was differentiated from other types of artifact glass, such as bottles, glassware, and lighting devises, which represented the ship’s cargo. Disqualifying features included glass with mold marks, curvature, or embossing. Characteristics ascribed to window glass include flatness, relatively uniform thickness, glazing marks, specific seed bubbles patterns, and decorative treatments. This analysis looked for specific evidence of window glass manufacture, decoration, and use as an architectural component. Before discussing the findings of this subject a summary of glass production, decoration, and installation is presented.


Production

Glass is defined as a supercooled liquid: "in vitreous bodies there is a gradual stiffening of the liquid until the viscosity or stickiness becomes so great that the behavior is that of a solid" (Armitage 1959:75).

Silica is the primary component in glass and generally derived from sand, quartz pebbles, or flint. Prior to the twentieth century, the melting point of silica, 1,750°C, was too high for most furnaces to reach and maintain. A lower melting point was achieved by adding an alkali flux in the form of plant ashes. Sodium and potash were the most common fluxes (Newton et al. 1989:54-56). For example, adding twenty-five percent sodium to silica lowers the melting temperature to 795°C (Armitage 1959:75). This soda glass is unstable and soluble in water but by adding lime a more durable product, called soda-lime glass, is produced (Newton et al. 1989:60).

Ideally, the ingredients used in glass making yield a colorless product. Nevertheless, impurities in the silica or flux often give glass an unwanted color. Iron, the most common impurity, imparts a greenish tint. Adding a decolorant such as manganese or antimony lessens or eliminates the color (Newton et al. 1989:59).

Glass production begins by placing the ingredients in a pot or crucible made of refractory material and melting the batch in a furnace or kiln. The molten material is called metal. From the early Roman period to the nineteenth century window glass was produced by two different glass blowing methods. Crown glass is made by spinning a gather of metal on the end of a pontil to make a disk of desired thickness and size (Figure 36). The crown's center is marked by a boss, or bull's eye, where the pontil attached to the crown. Fragments of crown glass can be identified by circular ripples in the surface or curving lines of small bubbles, called seed (Newton et al. 1989:55, 91-93).

The second type, cylinder glass, persisted as a manufacturing technique into the twentieth century (Figure 36). The glass blower produces the cylinder by blowing the gather into a globe. The gather is then suspended vertically while the worker alternately blows, swings, and spins the growing cylinder. The weight of the swinging molten glass pulls itself into a cylinder while the centrifugal force of spinning gives it an even shape and thickness. When finished, both ends are cut off and the cylinder is split longitudinally with a hot iron. The glass is then reheated in a kiln and flattened into a sheet (Ericson 1926:112; Newton et al. 1989:91). Cylinder glass can be identified by the elongated shape of the seed bubbles which form straight parallel lines consistent with the longitudinal shape of the cylinder (Newton et al. 1989:91-92).

Once a glass product is completed, whether window glass or any other glass fig 36 artifact, it must be annealed to relieve stresses which build up during rapid and differential cooling experienced while working the metal. Glass that is not annealed is weak, brittle, and prone to shatter during use. Annealing involves reheating the object in a furnace and slowly cooling it off over a period of several hours (Newton et al. 1989:63-64).


Decoration

While most of the window glass found on the Maple Leaf is undecorated, some has been embellished with different types of decorative treatments. These include coloring the metal during the manufacturing process, or painting and etching the finished panes. Two techniques are used to color glass during manufacture and both employ metals and metallic oxides as coloring agents. Pot colored glass is made by adding metal or metallic oxide to the molten batch in the furnace. This produces a uniform color throughout the glass. The second technique, called flash, is effected by dipping a gather of uncolored glass into a batch of colored glass. The gather is then blown into a cylinder or crown leaving a thin covering, or flash, of colored glass on one side (Newton et al. 1989:95-96).

In both techniques, colors are dependent on many variables in the glass making process; impurities in the batch, a reducing or oxidizing atmosphere in the furnace, furnace temperature, concentration of coloring agent, and mixing different coloring agents. Gold and copper produce ruby glass. Copper, as well as manganese and cobalt, can make blue. Copper oxide combined with lead oxide produces greens and manganese makes purple (Newton et al. 1989:57-59).

Standard panes of uncolored window glass can be embellished a number of ways, including painting, staining, enameling, and etching, depending on the desired effect. Paint is generally used to make dark opaque lines for drawing outlines much like an ink drawing on paper. Paint consists of ground glass, iron oxide, flux, and a binding medium, such as gum arabic, to make the paint flow and adhere to the glass. Flux is necessary to lower the melting point of the paint below the window glass. When the drawing is completed the glass is fired to fuse the paint to the surface (Newton et al. 1989:96; Reyntiens 1967:57).

Staining and enameling also involve firing to produce colors on the glass surface. In staining, silver nitrate is applied to a glass pane to produce a range of yellow colors when it is fired. This technique was developed in the fourteenth century to provide a way to incorporate more than one color on a single piece of glass (Newton et al. 1989:99-100). Translucent enamels were developed during the sixteenth century and became popular for the wide range of colors available. Enamels are made by mixing finely ground glass powder with a metallic oxide colorant. The ingredients are combined in a liquid medium and painted on the window pane. When fired, the enamel fuses to the surface and becomes a thin coat of glass on the window pane (Newton et al. 1989:97; Reyntiens 1967:87).

Most of the enameled glass from the Maple Leaf is also acid etched to create subtle surface decorations. These decorations appear as a dull mat finish on the glossy glass surface. Etching is done with hydrofluoric acid which dissolves silica. The window pane is prepared for treatment by applying a protective coat of bee's wax to the surface. Next, a decorative design is scratched through the protective wax to expose the glass below. The glass is then placed in a bath of hydrofluoric acid to etch the exposed surfaces (Newton et al. 1989:73).


Glazing

When the finished crown or cylinder glass comes out of the annealing furnace, it is cut to size before shipment to market. These glass panes are called lights (Ericson 1926:11). In 1817 lights were advertised in sizes 8 by 10 inches, 10 by 12 inches, and larger. Advertisements for 1847 mention lights measuring 21 by 25 inches and 24 by 30 inches. Generally size became larger throughout the nineteenth century as cylinder glass replaced crown glass (Roenke 1978:35).

Cutting and installing window glass is done by a glazer. The glazer prepares for glass installation by first measuring the opening for the light in the window frame. The wooden frame is rabbeted to create a recess to hold the glass (Ericson 1926:53). Glass is cut by scoring the surface using a steel wheel cutter or diamond tool. Once scored, the glass is broken by bending or tapping along the cut. Using a dull cutting tool or pressing too hard with a cutter leaves a trail of tiny splinters along the score and is considered a bad cut (Ericson 1926:26-27,46-47; Reyntiens 1967:48-51). Once properly sized, the pane is fitted in the frame and secured in place with glazing putty. Putty is made from linseed oil and whiting with white lead or zinc added for durability. A bead of putty is placed around the edge of the window pane and smoothed with a putty knife to fill the rabbet (Ericson 1926:57-60).


Findings

The glass discussed in this paper was found in a thick layer of shell hash covering the aft main deck. The layer contained a high artifact concentration including decorated and undecorated window glass. In the analysis, weight was used instead of individual artifact counts. Given the fragmentary nature of the glass, a minimum count of panes was not attempted. Hopefully, by using weight a more accurate and quantifiable analysis can be completed. Table 1 lists the glass types by provenience. In the following discussion, undecorated glass will be examined first.


Undecorated Glass

Undecorated window glass makes up 79.8% (2462.0 grams) of the total glass examined and is more prevalent than any other type. It all appears to be the same type with a greenish tint caused by iron impurities. Thickness ranges from 0.064 to 0.12 inches with an average of 0.64 inches. Another characteristic is the small seed bubbles which indicate the method of manufacture. The seed bubbles are elongate and run in parallel lines showing the panes were made from cylinder glass.

One unbroken pane from Recovery 109 can be used to extrapolate the approximate size of other panes used on the vessel. The unbroken piece, part of artifact number 05077, weighs 106.2 grams, measures 4.04 inches wide by 9.61 inches long, and averages 0.07 inches thick. Two other broken pieces provide additional width measurements. A fragment in artifact number 05151 is 4.10 inches wide, nearly the same as the first. A narrower fragment is part of artifact number 05077 and measures 3.28 inches.

Table 1. Distribution of window glass by type and provenience is given in weight. Overall percentage by weight of each type is presented at the bottom.

Colored Glass

Small amounts of pot and flash colored glass are represented in the sample (Table 1). There are three shades of pot colored glass; green, blue, and purple. In addition, a very small amount of flash green was found. Green is the only color found in both pot and flashed glass. Red, the fourth color represented, is only used as a flash.

Seed bubbles on the pot colored green glass suggest it is cylinder glass. Evidence of the manufacturing process used to make the flashed glass is inconclusive. Most of the fragments are too small to hold an adequate number of seed bubbles. Larger pieces contain seed bubbles elongated in random directions. This may be a function of the flash process.


Acid Etched

A small amount of colorless glass has been acid etched and two patterns have been defined. The snow flake pattern (Figure 37) is represented by one example from recovery 108, artifact 05075. It is a repeating octagonal pattern. The size of the original window could not be determined from the fragment. Seed bubbles proved inconclusive for determining the manufacturing process.

The pinwheel pattern is part of recovery 112, artifact number 05134. Upon recovery the glass was noted as being painted black on the rough etched surface (Figure 37). Subsequent conservation treatment washed the pigment off. This indicates the pigment was not fused to the glass by firing and was not very durable. This is the only example of this decorative technique found on the site.


Enameled/Etched Glass

The majority of decorative glass work found on the Maple Leaf falls under this category (Table 1). This technique uses acid etching to create a design or pattern on the Fig 37 glass. Colored enamel is applied to the opposite side to highlight the patten and enhance the piece.

Besides the painted glass previously mentioned, all secondary, or post manufacture, colored glass is enameled. Colors are primarily earth tones dominated by browns, yellows, and pale greens. Two patterns are represented well enough to recreate. The ivy pattern consists of vines and leaves (Figure 37). Intact edges on one window pane fragment give a width measurement of 3.21 inches, nearly the same width as the uncolored fragment from artifact number 05077. Scratch marks on the enameled side run parallel to the finished edges and were probably caused when the glass was cleaned with an abrasive cloth or newspaper. More importantly, the scratches indicate the width of the glazing putty used to hold the glass in the window frame. The narrow strip overlain by the putty measured 0.22 to 0.25 inches wide. Seed bubbles indicate the pane was made from cylinder glass.

Figure 37 illustrates the floral border pattern. This pattern was used to create a border around a fairly large glass pane, leaving the center open for additional decoration. Although the central decoration is very fragmentary, it might represent a maple leaf. The example illustrated shows a corner fragment. Cleaning marks are not present but a score mark from a bad cutting attempt is present along one edge of the enameled side.

Several other patterns are present but are too fragmentary to reconstruct. The etched surfaces appear to have floral motifs and different patterns can be discerned by the motif style and enamel design.