In 1791, John Cox Stevens, of New York, commenced improvements on steam navigation; but it was not until 1804 that any of these were carried into practice; and even after an expenditure, as he states, of “twenty thousand dollars,” and the constant devotion “of thirteen years of the best period of his life” to the project, he admits that his attempts were on the whole unsuccessful. These consisted of a plan for propelling a boat 25 feet long and 5 feet wide, by a rotatory engine, on the axle of which revolved a wheel, like a windmill or smoke-jack, worked at the stern, but he found it impossible to preserve a sufficient degree of tightness in the packing of the engine. A second modification of his rotatory apparatus proving on trial no better than the first, he had recourse to Watt’s engine, omitting the beam, and having a cylinder 4½ inches diameter with a nine-inch stroke; the boiler, which was only 2 feet long, 15 inches wide, and 12 inches high, consisting of no less than forty-one copper tubes, each an inch in diameter. This boat (which is interesting as the first in which we have a direct account of the use of tubular boilers) was tried in May 1804, and attained a velocity of 4 miles an hour.[78] After having made repeated trials with her, his son undertook to cross from Hoboken to New York, when, unfortunately, as she approached the wharf, the steam-pipe gave way. The boiler having also been damaged, he constructed another with the tubes placed vertically, and for this, perhaps the only portion of his invention worth securing, he, in the year 1805, obtained a patent in England,[79] where he then resided. [Sidenote: Oliver Evans.] While Fitch and Stevens were employed in the manner I have described, another American citizen, Oliver Evans, an ingenious mechanic, was endeavouring to mature a plan for using steam of a very high pressure, to be employed in propelling waggons on common roads, and in an account of his plans which he published in 1786,[80] he suggests a mode of propelling vessels by steam. From this circumstance he has been regarded by some authors as the contriver of a practicable steam-boat: his pretensions, however, rest solely on his own allegations. He states that, in 1785, he placed his engine, used to cleanse docks, in a boat upon wheels, the combined weight being equal to 200 barrels of flour, which he transported down to the water, and, when it was launched, he fixed a paddle-wheel to the stern, and drove it down the Schuylkill to Delaware and up the Delaware to the city, “leaving all the vessels going up behind me at least half-way, the wind being ahead.” In 1794, one Samuel Morey, of Connecticut, is said to have built a steam-boat which he propelled at 5 miles an hour on the Connecticut River, and, in 1797, he built another, with side wheels, at Bordentown, New Jersey, which was publicly exhibited and made a passage to Philadelphia, but which does not appear to have been afterwards employed.