it, however inconsiderable, must be susceptible of varying its energy and speed within very wide limits. This constitutes one of the greatest practical difficulties which the railroad system has to encounter.
Upon common turnpike roads or paved streets, this inconvenience is less than on railroads. The power of traction necessary on these roads is very variable, owing to the want of uniformity in their surfaces; but on a level Macadamised road it is estimated, on an average, by Mr. Gurney, as a 12th part of the weight of the load. Thus a carriage, weighing 12 cwt., would require a power of traction of 1 cwt.; a carriage weighing 6 tons requires a power of traction amounting to half a ton, and so on. The increased power of traction required by an ascent on a turnpike road is estimated exactly in the same manner as for railroads. An ascent of 1 foot in 12 will add to the power of traction necessary on a level an increased power amounting to one-twelfth of the load, and thus such an ascent would require the power of traction to be doubled; but all ascents less abrupt than 1 in 12 would not require the power of traction to be increased in so great a degree as double its amount on the level. It therefore follows, that so great a susceptibility of increase is not necessary in the powers of traction used on common roads in cases of ascent, as in those used on railroads. This arises not from any advantage possessed by common roads compared with railroads, but from the very reverse. The increase to the power of traction required by an ascent on a common road, is exactly the same in amount as that which would be required by an ascent of the same elevation on a railroad. But the power of traction necessary on a level common road is so great, that the increase caused by an ele-
