steamer service was rather irregular. The Sirius was withdrawn after making a few trips, and though the Great Western continued running, she lost money for the company that owned her. Other steamers made a few trips, but they also, like the Sirius, were withdrawn. In 1839 Samuel Cunard, of Halifax, N. S., went to England and succeeded in forming the celebrated Cunard Company. Assisted by a liberal Government mail subsidy, it was commercially successful from the start. The first vessels put in service were the paddle steamers Britannia, Acadia, Columbia, and Caledonia. They were of 1154 tons burden and their machinery was of 740 horse power. The Britannia, the first to sail, left Liverpool on July 4, 1840, and made the passage to Halifax in 12 days 10 hours; on her return voyage she did better, the time being but little over ten days. The Cunard Company enjoyed a practical monopoly of the transatlantic service until 1850, when the Collins (American) and Inman lines were started. The Collins Line lost two of its four steamers and was discontinued in 1858. Two other American steamship lines were started in 1850, the New York and Havre Steamship Company and the Vanderbilt Line. Both ceased running at the beginning of the Civil War.
The Inman Line was more successful. It began its career with steamers built of iron and propelled by screws. Though no faster than the wooden paddle-wheel Cunarders, they were cheaper to operate. The screw propeller (q.v.) now began rapidly to displace the paddle wheel, though the Cunard Company launched the Heotia, their last and finest paddle steamer, in 18G1. The rapid increase in size of ocean steamships led to the production of the Great Eastern (q.v.), but she was half a century ahead of the demands of ocean traffic and the adequate development of marine steam engineering. She was fitted with both screw and paddle engines, as it was thought impossible for either separately to deliver sufficient propulsive effect. The combination was not an economical one, and was a leading cause of her failure as a commercial venture. By 1860, in the fight for supremacy, the screw had become the unquestioned victor over the paddle wheel so far as ocean navigation was concerned, both in the merchant marine and in naval construction. Its advantages for war vessels were numerous, but the greatest of these was the possibility of placing all the propelling machinery of a screw steamer below the water line; and this alone was decisive. In merchant steamers the advantages of the screw were of a different kind. The efficiency of the paddle wheel depends upon the depth of immersion of the paddles: if too great or too little, the losses from slip, drag, and churning of the water are serious. The variation of draught consistent with economical propulsion was therefore very small—too small to admit of heavy loading. Very large wheels and feathering wheels reduced the losses somewhat, but introduced troubles of another type, while the rolling of paddle steamers in heavy seas greatly interfered with their speed and economy no matter what the character of the wheels. The draught and condition of lading of screw steamers was of much less importance and could be varied within much wider limits without perceptible loss of efficiency; rolling produced little effect, and though pitching might be serious in short vessels in which the screw was not deeply immersed, yet, owing to the small ordinary angle of pitch, the screw rarely rose high enough above the surface to give trouble.
Up to this time boilers were of the box type and the pressure of steam carried rarely exceeded 25 pounds per square inch—in many of the early steamers 10 pounds or less was the common practice. But the displacing of box boilers by cylindrical permitted a higher steam pressure, and this in turn demanded another form of engine to utilize it economically. The compound engine, which was built and patented by Hornblower in 1781 and revived by Woolf in 1804, had not been much used, because the conditions had not demanded it; but now it became a necessity. It consisted at first of two cylinders—and many compound engines are still so built—in which the steam was expanded in two stages, the first expansion taking place in the high-pressure cylinder, by which the pressure was reduced one-half, more or less, and the second expansion in the low-pressure cylinder, where the pressure was carried down to the atmospheric line or below it.
The demand for increased speed led to higher steam pressure and greater engine speed. The range of economical expansion in one cylinder being limited, the tri-compound or triple-expansion engine was designed to utilize the increased boiler pressures. The gain was two-fold. The new engines, using a higher pressure of steam, were lighter than their predecessors of equal power and they were also more economical. The first large vessel to be fitted with them was probably the Propontis, which, in 1874, was supplied with engines designed by Mr. A. C. Kirk. By 1880 the use of triple-expansion engines became common, though compound engines were largely used for another decade and they are still fitted in certain steamers where the conditions favor their economical working.
The continued demand for increased power, particularly in small vessels (torpedo boats and the like), naturally pushed up the steam pressure again, and, although the locomotive boiler was used to some extent, the advantages of the water-tube boiler soon became apparent. (See section on Boilers below.) Its capability to furnish very high pressure steam reacted upon engine design and produced the quadruple-expansion engine. The water-tube boiler is not yet much used in the mercantile marine, but is rapidly displacing the cylindrical boiler in naval construction. It has not yet brought about the extended use of quadruple-expansion engines in large vessels, but this may follow in the course of time.
The length of the voyage and the vast amount of traffic has caused the transatlantic trade to be the principal field of steamship development. While the gain in size and speed of the vessels in this trade has been continuous from the start, a great impulse was given by the building of the Britannic and Germanic for the White Star Line in 1874. They at once reduced the average passage from Queenstown to New York to about eight days. They were followed in 1870 by the Arizona of the Guion Line, confessedly built to outstrip all competitors, and her success was the begin-
