TOL. LXI.] PHILOSOPHICAL TRANSACTIONS. idl 



swift current of air is caused to flow from the middle of the cylinder towards the 

 end: this will easily appear by considering, that the retlundant fluid is collected 

 in much greater quantity near the ends of the cyhnders, than near the middle. 

 The same thing may be said, but he believes in a less degree, of the edges of thin 

 plates. '""» 



What has been just said concerning the current of air, serves to explain the 

 reason of the revolving motion of Dr. Hamilton's and Mr. Kinnersley's bent 

 pointed wires, vide Phil. Trans, vol. 51, p. 905, and vol. 53, p. 86; also 

 Priestley, p. 429: for the same repulsion which impels the air from the thick part 

 of the wire towards the point, tends to impel the wire in the contrary direction. 



It is well known, that if a body b is positively electrified, and another body a, 

 communicating with the ground, be then brought near it, the electric fluid will 

 escape faster from b, at that part of it which is turned towards a, than before. 

 This is plainly conformable to theory ; for as A is thus rendered undercharged, 

 B will in its turn be made more overcharged, in that part of it which is turned 

 towards a, than it was before. But it is also well known that the fluid will 

 escape faster from b, if a be pointed, than if it be blunt ; though b will be less 

 overcharged in this case than in the other; for the broader the surface of a, 

 which is turned towards b, the more effect will it have in increasing the over- 

 charge of B. The cause of this phenomenon is as follows: 



If a is pointed, and the pointed end turned towards b, the air close to the 

 point will be very much undercharged, and therefore will be strongly repelled by 

 A, and attracted by b, which will cause a swift current of air to flow from it 

 towards b, by which means a constant supply of undercharged air will be brought 

 in contact with b, which will accelerate the discharge of electricity from it in a 

 very great degree: and moreover, the more pointed a is, the swifter will be this 

 current. If, on the other hand, that end of a which is turned towards b, is so 

 blunt, that the electricity is not disposed to run into A faster than it is to run 

 out of B, the air adjoining to b may be as much overcharged as that adjoining 

 to A is undercharged ; and therefore may, by the joint repulsion of b and attrac- 

 tion of A, be impelled from b to a, with as much or more force than the air 

 adjoining to a is impelled in the contrary direction ; so that what little current 

 of air there is may flow in the contrary direction. 



It is easy applying what has been here said to the case in which b is negatively 

 electrified. 



^5. In the paper of Mr. Canton's, quoted in the 2d section, and in a paper 

 of Dr. Franklin's (Phil. Trans., 1755, p. 300, and Franklin's letters, p. 155) are 

 some remarkable experiments, showing that when an overcharged body is 

 brought near another body, some fluid is driven to the farther end of this body, 

 and also some driven out of it, if it is not insulated. The experiments are all 



