Fired by the hot-tube ignition system, the force of the explosion would drive both pistons down, forcing the outer one tight against the head of the smaller one, and at the end of the stroke the longer wall of the outer piston would strike an arm projecting into the cylinder near the open end, moving forward the exhaust valve rod to which the arm was attached, thus pushing open the valve in the head.[13] On the exhaust stroke the unrestrained outer piston moved all the way to the head, expelling all of the products of combustion and pushing the exhaust valve shut again. With a bore of four inches or less, this engine, Charles believed, should develop about three horsepower and run at a speed between 350 to 400 revolutions per minute.[14]

As no ignition system had yet been provided, they prepared a 4½-inch length of one-quarter inch iron pipe, closed at one end, and screwed the open end into the head. Heating this tube with an alcohol burner would cause ignition of the mixture when a portion of it was forced into the heated tube toward the end of the compression stroke. No attempt was made at this time to use the electrical make-and-break circuit used in their second engine, as the free piston would have wrecked the igniter parts on the exhaust stroke, and the push rod located on the end of the piston would have prevented the piston from closing the exhaust valve.

After keying the flywheel to the lower end of the crankshaft, Charles and Frank decided to make an attempt to run the engine. Carrying it into a back room, probably during July or August, 1892, they blocked it up on horses. A carburetor had not yet been constructed, so they attempted to start the engine by spinning the flywheel by hand, at the same time spraying gasoline through the intake valve with a perfume atomizer previously purchased at a drugstore in the Massasoit House. Repeated efforts of the two men to start the engine resulted in failure.

Figure 15.—Conjectural drawing of the free-piston engine used in the Museum vehicle prior to the present engine. (Drawing by A. A. Balunek.)

In the belief that the defects, whatever they might be, could be remedied after the engine was installed, the Duryeas went ahead and mounted the engine in the carriage. To do this they shortened the original reach of the carriage, allowing the engine itself to become the rear continuation of the reach. The four ears on the front, or open end of the engine, were bolted to the centrally located frame, with the bearing blocks in between. This frame, the same one now in the vehicle, was constructed of two pieces of angle iron, riveted and brazed together. Greater rigidity was obtained by a number of half-inch iron rods running from the frame to both front and rear axles. Because of the absence of any mounting brackets on the engine casting itself, a wooden block with a trough on top to receive the body of the engine was fitted between the engine and the axle, while two U-shaped rods secured it with clip bars and nuts underneath.

Beneath the flywheel was mounted the friction transmission of Charles' design. This consisted of a large drum, perhaps 12 inches in diameter, equal in length to the diameter of the flywheel and keyed to a shaft directly under the center of the crankshaft and parallel to the axles. (Diameter of drum estimated by examination of existing features.) In view of the four projections of the frame extending downward and just in front of the jackshaft position, it is likely that these supported the four jackshaft bearings. Being a bicycle manufacturer, Charles saw the need for a differential or balance gear. Accordingly, he purchased from the Pope Manufacturing Company a very light unit of the type formerly used on Columbia tricycles, and installed it somewhere on the jackshaft. A small sprocket on each end of the shaft carried a chain from the larger sprockets clamped to the spokes of each rear wheel. The lower surface of the flywheel had been machined so as to form a friction disc, with a one-quarter inch depression 3 inches in diameter turned in the center. The drum was positioned so that its upper surface was one-quarter inch below the face of the flywheel. Hanging loosely around the drum was an endless belt, one and one-half inches wide, first made of rather soft rubber packing material. The belt lay on the drum surface between the fingers of a shipper fork. While it lay under the 3-inch depression in the center of the flywheel, the belt and the drum were at rest, but when it was moved away from that depression the belt wedged itself tightly between the drum and flywheel, the resulting friction causing the drum to turn and setting the vehicle into motion. The farther the belt was moved toward the outer edge of the wheel, the faster the drum and the vehicle moved.

In September 1892, Charles, who had contracted with a Peoria, Illinois, firm to have bicycle parts manufactured, decided to move to that city. Departing on the 22d of September, he did not return to Springfield for over two years, and thus was not able to participate in the completion and testing of the carriage. At the time of his departure several units on the carriage were incomplete. A carburetor had not been built, nor had a satisfactory burner or belt-shifting device. Charles had experimented with various shifting levers just before leaving Springfield: however, as he reported later, he did not succeed in designing a workable mechanism.[15] Frank Duryea, now left to finish the work unassisted, continued the experiments with the belt shifter. He finally worked out a fork mounted on a carriage that was supported by two rods, each of which slid in two bearings. Although the short distance between the two bearings caused the shifter carriage to bind occasionally, the device was thought to be sufficient and was installed just in front of the frame. Connected to a system of cables, arms, and rods, possibly similar to the present cam-bar shifter, the shipper-fork carriage was moved from side to side by raising or lowering the tiller.

Figure 16.—Drawing of the carburetor used on both Duryea engines, 1893-1894,
showing sight feed on left and choke mechanism on right. (Smithsonian photo 13455.)

Turning now to an efficient burner for heating the ignition tube, Frank started with an ordinary wick-type kerosene lamp with a small metal tank. Wishing to use gasoline in the lamp, he found it necessary to fabricate a number of burner units before he found a type that gave him a clean blue flame. He then found the flame to be very sensitive to drafts and easily extinguished, and devised a small shield or chimney to afford it some protection.

Early in October, while still working with the burner, Frank developed a severe headache. He felt the fumes of the lamp had probably caused it, and went to his room in the home of a Mr. and Mrs. Patrick on Front Street in Chicopee. After he noticed no improvement, a doctor's examination showed he had typhoid fever, and on October 5 he was