[ 76 ] 



though I was well fat'isfied with the accuracy of this mode of 

 trial, I conftru»aed another that might appear to others lefs liable 

 to objedion. 



I SCREWED a circle of iron three feet three inches diameter 

 upon a folid floor ; in the centre of this circle I ereded an 

 upright axis or roller upon two pivots, one refting in a focket 

 of brafs upon the floor, the other in a bridge which was raifed 

 acrofs the machine. This axis or roller had a fmall filk cord 

 wound round its circumference, which pafllng into an adjoining 

 ftaircafe, had a fcale and weights tied to it, which turned the 

 roller with the required velocity. From the roller a horizontal 

 arm of wood extended to the circumference of this iron circle, 

 and to its extremity was faftened a piece of fteel in the form 

 of an axle-tree of a carriage, and upon this was placed a 

 wheel, which by thefe means was carried round upon the brafs 

 circle, as the ftone of tanners bark-mill moves round the trough 

 which belongs to it. This arm was permitted to move up and 

 down by means of a hinge, fo as to let the wheel rife over any 

 obftacle which was placed in its way. Befides this another arm 

 was placed above that which carried the wheel, at the extremity 

 of which was faftened a piece of tin, forming a vanCj which, by 

 its refiftance to the air, regulated the motion of the machine. 

 The roller was now made to turn by putting weights into the 

 fcale, and it was let to revolve until its motion became uni- 

 form. After eight or ten turns it revolved with an equable velo- 

 city ; and during every fet of experiments the fame velocity was 

 preferved, and whatever renftance the carriage was expofed to 

 was overcome by the addition of weight. The additional weight 



became 



