VOL. LXIV.] PHILOSOPHICAL TRANSACTIONS. 513 



B remained a full inch asunder. From this experiment, it seems evident how 

 much better adapted a sharp point is to draw off lightning, than a knob of 4. inch, 

 in diameter ; and consequently how much more likely to cause it to pass in that 

 conductor, to which it is affixed, rather than in any other part of the building, 

 where it might occasion much damage, as well as endanger the lives of those 

 who might happen to be in it. The following experiments seem to make still 

 more strongly in favour of the same conclusion. 



Exp. 1. He affixed to the top of a glass- stand a wire, -§• inch in diameter, 

 terminated at one end by a ball, -f. inch in diameter; and at the other end by a 

 very sharp point ; see fig. 3. Round the middle of this wire was hung a chain 

 12 inches long. He then charged a bottle, containing 100 square inches of 

 coated surface, and connecting the chain with the coating of the bottle, brought 

 the knob of it very gently towards the ball on the insulated wire, that he might 

 observe precisely at what distance it would be discharged on it; which he found 

 to happen constantly at the distance of half an inch, with a loud and full explo- 

 sion. Then re-charging the bottle, he brought the knob, in the same gradual 

 manner, towards the point of the insulated wire, to try also at what distance 

 that would be struck; but this in many trials never happened at all. The point, 

 being approached in this gradual manner, always drew off the charge impercep- 

 tibly, leaving scarcely a spark in the bottle. ,, 

 Exp. 3. Mr. H. had now recourse to the apparatus known to electricians by 

 the name of the thunder-house, which he thought a nearer resemblance of the 

 operations of nature on these occasions. Having connected a jar, containing 

 509 square inches of coated surface, with the prime conductor, see fig. 4 ; he 

 observed, that if it was so charged as to raise the index of the electrometer to 60 

 degrees, by bringing the ball on the wire of the thunder-house to half an inch 

 distance from that connected with the prime conductor, the jar would be dis- 

 charged, and the piece in the thunder-house thrown out to a considerable dis- 

 tance. Using a pointed wire for a conductor to the thunder-house, instead of 

 the knob, as in the former experiment, the charge being the same, the jar was 

 discharged silently, though suddenly : and the piece was not thrown out of the 

 thunder-house. 



Exp. 4. Having made a double circuit to the thunder-house, fig. 5, the 1st 

 by the knob, the ad by a sharp-pointed wire, at l^inch distance from eachotli^r, 

 but of exactly the same height, the charge being the same; though the knob 

 was brought first undei- that connected with the prime conductor, which was 

 raised half an inch above it, and followed by the point, at 14- inch distance, yei 

 no explosion could fall on the knob; the point drew off all the charge silently.; 

 and the piece in the thunder-house remained unmoved. 



Exp. 5. Having insulated the jar, and connected, by chains, with the external 



VOL. XIII. 3 U 



