M, Arago's Biographical Memoir of James Watt. 235 



charged it, and it is upon the shoulder therefore that the gun- 

 stock acts with violence. To change the direction of the re- 

 coil, all that is necessary is to change the direction in which 

 the gas issues. If the barrel, closed at its extremity, were 

 pierced by a lateral opening at right angles to its direction 

 and horizontal, it would be laterally and horizontally that the 

 gas of the powder would escape, — it would be at right angles 

 that the barrel would produce its recoil, — and it would be 

 upon the arm, and not upon the shoulder, that it would be 

 felt. In the former instance, the recoil operates upon the in- 

 dividual who has fired the piece, from before backwards, with 

 a tendency to throw him down ; in this latter instance, it 

 would have a tendency to make him whirl round upon him- 

 self Were we then to attach the barrel, constantly and in a 

 horizontal position, to a moveable vertical axis, at the moment 

 of being discharged it would more or less change its direc- 

 tion, and would cause that axis to revolve upon itself. Still 

 maintaining the same arrangement; suppose now, that the 

 vertical rotative axis is hollow, but closed at its upper ex- 

 tremity, that it abuts at its lower extremity, like a sort of 

 chimney in a boiler producing steam, and that, moreover, 

 there exists a free lateral communication between the interior 

 of the axis and the interior of the gun-barrel, so that the 

 steam, after having filled the axis, penetrates into the barrel, 

 and issues from it at the side by its horizontal opening, — 

 then this steam, in escaping, will act, except in the degree 

 of its intensity, in the same manner as the gas disengaged 

 from the powder in the gun-barrel which was stopped at its 

 extremity and pierced laterally ; only that, in this latter in- 

 stance, we shall not have a simple shock merely, as happens in 

 the violent and instantaneous explosion of the musket, but, 

 on the contrary, the rotatory motion will be uniform and con- 

 tinuous, as the cause which produces it : and finally, were we, 

 instead of having a single musket, or rather a single horizon- 

 tal pipe, to adapt a number of them to a rotatory vertical 

 tube, then we should have before us, with some unessential 

 differences, the ingenious apparatus of Hero of Alexandria. 



Here then, beyond doubt, is a machine in which steam pro- 

 duces motion, and may cause mechanical eifects of some im- 



