which may have been indicated, rather shows a diminution of resistance with increased speed; but for practical purposes the resistance may be regarded as constant, and quite independent of the velocity. I therefore infer, that the power of traction necessary on level roads, whether they be roads of stone or roads of iron, will always be in proportion to the load, and independent of the speed.
This becomes one of the most striking features of difference between the effects in favour of roads and against canals. In the latter, every increase of speed renders a proportionate increased power of traction necessary; while in the former no increased power of traction whatever is needed. If a carriage be propelled on a road ten miles in five hours, or ten miles in one hour, the power of traction must be precisely the same in both cases; but if a boat be propelled on a canal ten miles in one hour, the power of traction must be more than ten times that which would be necessary to carry it ten miles in five hours. This observation will be equally applicable to turnpike roads and railroads, as compared with canals; and it will lead to the inference that there is a limiting speed, at which the effect of canals must equal the effect of a hard level road travelled by a carriage; and that below this limit the canal has the advantage, while, above it, the advantage lies with the road. As the resistance to the boat in the water has an immediate dependence on its rate of motion, it follows, that by reducing that rate of motion without limit, the resistance may be also reduced in a proportionate limit; while, on the other hand, the resistance to a carriage moving on a railroad, being independent of the speed, the reduction of speed can cause no dimi-
