PROCEEDINGS OF THE POLYTECHNIC ASSOCIATION. 573 



of tlic locomotive from side to side strains the rails and they must be 

 secured to crossties or they would be torn apart. Ag-ain, the locomotive 

 cannot ascend steep grad(>s, so that moderate hills must be tunneled and 

 gentle valleys bridged. Nor is it free from danger. It is liable to explode, 

 to run off the track, to be thrown from the track by stray cattle, fallen 

 trees, drift from rain, snow, &c. It is also liable to cause accidents by 

 frightening horses, and its smoke, ashes and dust are offensive in cities 

 and intolerable in long tunnels or underground roads. 



The stationary engine possessing so many advantages above the loco- 

 motive, it was to be expected that great efforts would be made to substi- 

 tute the one for the other. The great difficulty was to convey its power to 

 a train of cars in motion, and of all the means proposed for accomplishing 

 that end, the oidy one now approved by engineers is the use of atmospheric 

 air in tubes. 



An endless band has been used on the London and Blackwall, and some 

 other railroads. A rotating endless screw has been proposed ; so has a 

 tube of india-rubber to be filled with water behind the driving wheel or 

 roller; and other plans which were simply abortions. 



The employment of atmospheric air in different ways has been more 

 successful. Between Dublin and Dalkey, in Ireland, an atmospheric rail- 

 road was used for a short time. A tube having a slit along its upper side 

 was laid between the rails. A piston to travel in this pipe was fitted with 

 a bent arm, which, passing through the slit in the top of the tube, was 

 attached to the car. The slit was closed by a valve of leather strengthened 

 with metal strips across it, which was fast along one side and was sealed 

 to the other side with pitch so that the bent arm in passing along tore up 

 the sealed side of the valve, and a hot roller following resealed it. Motion 

 was given to the car by the pressure of the air acting on the piston in the 

 tube, when a partial vacuum was produced in front by air-pumps worked 

 by a stationary engine. The valve was very imperfect, and whilst travel- 

 ing in the car, I have heard it hissing throughout its length as it leaked 

 in air. Some ingenious improvements in the valve were proposed, but the 

 great defect in this plan was that the lateral strain on the tube, as the car 

 oscillated from side to side, distorted the tube so that the piston could not 

 fit it and the leakage was too great. 



To dispense with the slit along the tube it was proposed to have a piston 

 travel in a perfectly enclosed tube, and by means of a toothed rack on top, 

 give motion to a double set of rollers, half within the tube and half with- 

 out, those within being connected with those without by pins passing 

 through air-tight holes, so that the outside rollers should grasp a rack on 

 the bottom of the car and thus give motion to it. This was very ingenious, 

 but that the car and piston should travel at the exact same rate, was not 

 under complete control. This plan was consequently abandoned. 



It was next proposed to have holes at intervals along the tube, and to 

 have these closed by valves which could be operated by a traveling slide 

 valve, so as to make a communication, at the moment of passing, between 

 the inside of the tube and an air engine in the car, the car being, in fact, 

 an air locomotive. This plan has been lately revived, but it seems too 



