VOL. XLIII.] PHILOSOPHICAL TRAKSACTIONS. 115 



tube turn about its axis, and nriove on towards the fire in such a manner, as made 

 him still less inclined to think either of the motion . owing to the draught of the 

 fire, and certainly not to the whole weight of the ' loving tube ; a fine spirit-level 

 informing them that the supporting tubes leaned irom the fire; so that the motion 

 was a little up-hill. 



This success determined him to go on further; and, furnishing himself with 

 tubes of several lengths and thicknesses, he made several trials; and found, that 

 with a moderate fire the experiment succeeded best, when the supported tube was 

 about 20 or 11 inches long, the diameter about ~ of an inch, and had in each 

 end a pretty strong pin, fixed in cork, for an axe to roll with upon the supporting 

 tubes; which, to lessen the contact, had nearly the same diameter with the 

 moving one. Under these circumstances the tube would begin to move at 18 

 inches distance from the fire; and continue to do so, with little intervals, till it 

 touched the bars ; and moved much in the same manner, when a little ball of 

 cork, an inch or more in diameter, was fixed in the middle of it. But what sur- 

 prised him still more, and seemed to take off the objection of the draught of the 

 chimney, was, letting it once stay a little while against the bars, its still continuing 

 its motion about its axis in the same direction. 



This put him on making little rings of wire, to fix on and move along the sup- 

 porting tubes, so as to stop the moving tube at any distance from the fire. Stopped 

 with these, the motion of the tube about its axis still continued. 



Desirous to try what would be the effect in or near an upright position, he 

 made the pin at one end of the tube rest on a China plate, that at the other turn 

 in a silver socket (that carried his pencil) fixed in a horizontal arm of wood, but so 

 as he could slip it up and down, to adapt it to the length of the tube. Here he 

 found, that if the tubes leaned to his right hand, the motion was from east to west; 

 but if they leaned to his left, the motion was from west to east; and the nearer he 

 could get to the perfectly upright position, the less the motion seemed to be 

 either way. 



He now placed the tube horizontally on a glass plane, a large fragment of a 

 coach-side window glass. The tube, instead of moving towards the fire, moved 

 from it, and about its axis, in a contrary- direction to what it had done before. 

 Observing that this glass plane was broader at one end than the other, and that 

 the rotation backwards was more sensible when the narrower end was towards the 

 fire, he placed a triangular piece of the same glass with its vertex towards the fire 

 nearly horizontal, but rather rising from the fire; so that its base was a little 

 higher than its vertex; and on it a tube of glass, aliout 11 inches long, and \ of 

 an inch diameter, near the vertex and the fire. This tube receded from the fire, 

 moving about its axis till it came to the distance of 8 inches; which is 4 inches 



a2 



