Two notable improvements were made in the engines in the meantime. Robert L. Stevens designed a light wrought-iron "walking beam" to replace the heavy cast-iron beam that was previously used. Then in 1824 James P. Allaire, who had made a reputation as an engine builder, brought out a "compound" engine. A compound engine has two or more cylinders of different diameters coupled to the one shaft. In the Watt engines the steam was conducted from the boiler to one end of the cylinder and allowed to flow in until the piston was driven almost to the opposite end. Then it was shut off and the way to the condenser was opened. Later it was found that if the flow of steam was cut off at half the length of stroke and the steam already in the cylinder was allowed to work by expansion, the total power of the engine was reduced only a little, while the saving of coal was very great. While thinking of this fact, it occurred to Allaire that the principle involved might be used to better advantage if the steam was taken at full pressure into a small cylinder and exhausted thence into a second and larger one to work by its expansive power. The _Henry Eckford_, the _Sun_, and a number of others were supplied with engines on this principle. The _Eckford_ had a small cylinder 12 inches in diameter and a large one of 24. Both had the same stroke, of course. But, curiously enough, it was not until about 1870 that compound engines became the fashion. The racing era, as one may call it, began on the Hudson about 1835, the year that Daniel Drew became interested in the Hudson River steamers. Racing had been done before the consolidation of 1832, but it was mild in comparison with that under Drew's initiation. An unreasoning mania for speed took possession of the public as well as of the owners of steamers. Though no practical end was to be served by the saving of an hour in the time required in the trip from New York to Albany, every sacrifice was made and every risk was taken to secure it. Instead of going to a pier or dock to land way-passengers at the towns along the river, the unfortunates were dumped into a small skiff that was towed at the end of a long rope. Then, as the steamer ran in close, but not too close, to the landing, the man in charge of the skiff, using a steering oar, sheered it within a few feet of the landing, and the passengers were told to jump. And jump they did. Of course a few fell into the water at every trip; perhaps some were drowned thereby. But the steamer's reputation for making quick passages was maintained, and for a long time the public protested in vain. When legislation was invoked (1842) the act was opposed as an unwarranted attempt to interfere with private business which could much better regulate itself!