The year before this, in 1802, Richard Trevithick, in England, had been more successful in his attempts at producing a locomotive. He produced a steam locomotive that operated on the streets of London and the public highways, hauling a wagonload of people. But the unevenness of the roads proved disastrous to his engine, and as it could make no better time than a slow horse, it was soon abandoned. But Trevithick had learned from this failure that a good roadbed was quite as essential to the success of a locomotive as the machine itself, and two years later he produced what is usually regarded as the first railway locomotive. This was built for the Merthyr-Tydvil Railway in South Wales, and on several occasions hauled loads of ten tons of iron at a fair rate of speed. It was not considered a success financially, however, and was finally abandoned. At this time a curious belief had become current among the inventors to the effect that if a smooth-surface rail and a smooth-surface wheel were used, there would not be sufficient friction between the two to make it possible to haul loads, or more than barely move the locomotive itself. Learned mathematicians proved conclusively on paper by endless hair-splitting calculations that the thing was impossible,--that any locomotive strong enough to propel itself along a smooth iron rail would be heavy enough to break the strongest rail, and smash the roadbed. In the face of these arguments the idea of smooth rails and smooth wheels was abandoned for the time. Trevithick himself was convinced, and turned his attention to the perfecting of an engine with toothed drive-wheels running on a track with rack-rails. But this engine soon jolted itself and its track into the junk-heap without doing anything to solve the problem of locomotion. Shortly after this, a man named Chapman, of Newcastle, built a road and stretched a chain from one end to the other, this chain being arranged to pass around a barrel-wheel on the locomotive, which thus pulled itself along, just as some of the boats on the Rhine do at the present time. But the machinery for operating this engine was clumsy and unsatisfactory, and the road proved a complete failure.