THE STEAM-ENGINE. 563 



dependent of the speed with which the train was moved ; and it was accord- 

 ingly assumed in all calculations respecting the power of locomotive-engiues 

 that the resistance would be practically the same, whatever might be the speed 

 of the train. It had been well understood that, so far as the atmosphere might 

 offer resistance to the moving power, this would be dependent on the speed, 

 and would increase in a very high ratio with the speed ; but it was consid- 

 ered that the part of the resistance due to this cause formed a fraction of the 

 whole amount so insignificant that it might be fairly disregarded in practice, 

 or considered as a part of the actual computed resistance taken at an average 

 speed. 



It has been, until a late period, accordingly assumed that the total amount 

 of resistance to railway-trains which the locomotive-engines have had to over- 

 come was about the two hundred and fiftieth part of the gross weight of the 

 load drawn : some engineers estimated it at a two hundred and twentieth ; oth- 

 ers at a two hundred and fiftieth ; others at a three hundred and thirtieth part 

 of the load ; and the two hundred and fiftieth part of the gross load drawn may 

 perhaps be considered as a mean between these much-varying estimates. 

 What the experiments were, if any, on which these rough estimates were 

 based, has never appeared. Each engineer formed his own valuation of this 

 effect, but none produced the experimental grounds of their opinion. It has 

 been said that the trains run down the engine, or that the drawing-chains con- 

 necting the engine slacken in descending an inclination of sixteen feet in a 

 mile, or 3^. Numerous experiments, however, made by myself, as well as 

 the constant experience now daily obtained on railways, show that this is a 

 fallacious opinion, except at velocities so low as are never practised on rail- 

 ways. 



In the autumn of 1838 a course of experiments was commenced at the sug- 

 gestion of some of the proprietors of the Great Western railway company, with 

 a view to determine various points connected with the structure and the work- 

 ing of railways. A part of these experiments were intended to determine the 

 mean amount of the resisting force opposed to the moving power, and this part 

 was conducted by me. After having tried various expedients for determining 

 the mean amount of resistance to the moving power, I found that no method 

 gave satisfactory results except one founded on observing the motion of trains 

 by gravity down steep inclined planes. When a train of wagons or coaches 

 is placed upon an inclined plane so steep that it shall descend by its gravity 

 without any moving power, its motion, when it proceeds from a state of rest, 

 will be gradually accelerated, and if the resistance to that motion was, as it 

 has been commonly supposed to be, uniform and independent of the speed, the 

 descent would be uniformly accelerated ; in other words, the increase of speed 

 would be proportional to the time of the motion. Whatever velocity the train 

 would gain in the first minute, it would acquire twice that velocity at the end 

 of the second minute, three times that velocity at the end of the third minute, 

 and so on ; and this increase of velocity would continue to follow the same 

 law. however extended the plane might be. That such would be the law 

 which the descending motion of a train would follow had always been sup- 

 posed, up to the time of the experiments now referred to ; and it was even 

 maintained by some that such a law was in strict conformity with experiments 

 made upon railways and duly reported. The first experiments instituted by > 

 me at the time just referred to afforded a complete refutation of this doctrine. 

 It was found that the acceleration was not uniform, but that with every in- 

 crease of speed the acceleration was lessened. Thus if a certain speed were 

 gained by a train in one second when moving at five miles an hour, a much 

 less speed was gained in one second when moving ten miles an hour, and a 



