Mr White's Dcsi^ifor a Rotatory Steam-Engine. 267 



from the valve. A fly-wheel being supposed to be attached to 

 the engine, will continue the motion, cause the quadrants ort 

 the piston-plate to open the valve, and allow the piston-plate to 

 pass. It has now made a complete revolution. When the 

 valve f begins to open, it shuts the steam off, consequently 

 there is no expence of steam when the power has to be perform- 

 ed by the fly-wheel. The re-shutting of the valve is also per- 

 formed by steam, from the small lever on the spindle of the 

 valve connected with the piston, working in the small steam cy- 

 linder T : when the valve begins to open, it will raise the pis- 

 ton in the cylinder T ; but as soon as the piston-plate P has 

 passed the valve, the steam acting on the piston of the small cy- 

 linder T, will force it down again, and shut the valve. Thus 

 may we consider it ready for a second revolution. Before ex- 

 plaining further, let us suppose that the engine only consisted 

 of one steam-passage, in place of three, and let it be that which 

 I have already described : when the piston-plate P has only 

 made a half revolution, it is evident that the steam would then be 

 entirely on the under-side of the revolving cylinder A, and force 

 it up against that side where there is no steam, and xjause such 

 a degree of friction on its bearings, that the force on the piston- 

 plate would be little more than sufficient to turn the revolving 

 cylinder round. To overcome this great obstacle, I have di- 

 vided the [outer cylinder into three divisions, shown in Plate V. 

 Fig. J^. by the division plates marked <r, oc, &c. These plates 

 do not revolve. The first division contains the cylinder A, be- 

 ing that which I have already described as having made a re- 

 volution ; the mid division, the cylinder E, and the third divi- 

 sion the cylinder D. The two end divisions containing the cylin- 

 ders A and D, are in every respect similar to one another, and 

 the valves j^^ are on the same spindle. The mid division con- 

 taining the cylinder E is in principle exactly similar to the other 

 two, and only has this difference, that it is equal in length to 

 both. 



Fig. 1. Let the steam be admitted by the steam-pipe s, and 

 pass downwards through the two end valves f, f; at the samp 

 time it will pass upwards, through the pipe W, over the outer 

 cylinder, and through the mid valve V, act upon the piston- 

 plate F of the mid division, shewn in Fig. 2. and equahze the 



