M. Arago's Biographical Memoir of James Watt. 249 



able portion and the internal surface of the cylinder. To the 

 very great surprise of the makers, one of their machines one 

 day commenced working with most unwonted speed. After 

 numerous observations, it was ascertained that on that occasion 

 the piston had been pierced, and that the cold water had found 

 its way directly into the cylinder in small drops, — these, in 

 traversing the steam, quickly condensed it. From this for- 

 tuitous observation is dated the complete suppression of the ex- 

 ternal refrigeration, and the adoption of an injection to produce 

 a shower of cold water throughout the whole cylinder at the 

 moment proper for the descent of the piston. The oscilla- 

 tions thus acquired all the velocity that coidd be desired. 



And once more we must remark that, on a different occa- 

 sion, an accidental circumstance seems to have had a share in 

 an improvement equally important. The first machine of New- 

 comen required the most unremitting attention on the part 

 of the individual who unceasingly opened and closed cer- 

 tain stopcocks, first for the introduction of the steam into the 

 cylinder, and then for injecting the cold shower for its con- 

 densation. It happened on one occasion that the person 

 .so employed was a boy named Henry Potter. His young 

 companions at their sports, uttered cries of delight, which 

 vexed him beyond endurance. He was all impatience to join 

 in their sport, but his required duties did not allow him 

 half a minute's absence. His anxiety excited his ingenuity, 

 and led him to observe relations he had never before thought 

 of. Of the two stopcocks, the one required to be opened 

 at the moment that the beam (which Newcomen first and 

 so usefully introduced into his machines) terminated the 

 descending oscillation, and required to be closed precisely 

 at the termination of the opposite one. The management 

 of the other stopcock was precisely the reverse. The posi- 

 tions, then, of the beam, and of the stopcocks, had a neces- 

 sary dependence upon each other. Potter seized upon this 

 fact ; he perceived that the beam might serve to impart to 

 the other parts of the machine all the required movements ; 

 and on the spiu* of the moment he realized his conceptions. 

 He attached a number of cords to the stopcocks ; some to the 

 one end of the handle, and some to the other, and these he 



