346 PHILOSOPHICAL TRANSACTIONS. [aNNO 1778. 



permitting SO great an accumulation to take place as was experienced in the IQth 

 experiment, when the separated substitutes struck, at the greatest distance. 



Having now gone through the experiments where points were introduced, we 

 shall next relate the several experiments where other terminations were used. 

 By this method of proceeding, we shall be able to form a proper judgment of 

 what kind of conductors are the most advantageous for securing buildings, &c. 



Exp. 23. On repeating the 18th experiment, but with a rounded conductor 

 on the model, every other circumstance continuing the same, and the model at 

 rest, the greatest distance at which it was struck, by 10 or It turns, was not 

 more than \ of an inch. Now if we compare this distance witii that at which 

 the point was struck in the 1 8tii experiment, the proportion will be found to be 

 less than l to 5. — Exp. 24. When the 23d experiment was repeated, and the 

 rounded end struck at 4 of an inch, the same kind of spark appeared between 

 the 2 cramps as in the 20th experiment when the point was used ; but in this 

 experiment with the rounded end the spark at the cramps appeared considerably 

 less to every observer. — Exp. 25. On repeating the 21st experiment where the 

 1 substitutes were brought into contact, every other circumstance remaining the 

 same, the rounded end was struck at the same distance nearly as when a spark 

 was taken by a larger metal ball, suppose 3 inches in diameter, from any part of 

 the great cylinder when equally charged : for in this case the 2 substitutes, being 

 in contact, made in reality but 1 great substitute. — Exp. 26. I now repeated the 

 22d experiment, where motion was introduced; and without any other change 

 of circumstances than putting in the place of the point the rounded end. On 

 this occasion, as well as on the former, the same person moved the stand with 

 the rounded end on it, and with the same velocity, but not before the connected 

 substitutes were fully charged by an equal number of turns. The instant that 

 the rounded end approached within a certain distance of the ball at the end of 

 the less substitute, it struck ; but the explosion seemed inferior to that which the 

 point occasioned at the distance of Qi inches. In this experiment the distance 

 between the rounded end and ball was not more than 6i^ inches. From which 

 it appears, that even in those circumstances the point was struck at a greater 

 distance than the rounded end in the proportion of gj- to 6i. 



Exp. 27. It has been observed in a former part of this paper, that a great 

 quantity of rain fell when the accident happened at Purfleet ; as this circumstance 

 seemed to be material, it was proper to put the model into a similar situation ; 

 and therefore, after removing the tin foil on the parapet, I washed the model all 

 over with a sponge ; and, while it continued in this state, the machine was put 

 into motion : after 10 turns of the wheel, the point was struck at 5 inches dis- 

 tance. In consequence of this a small explosion appeared, not only between the 

 cramps ; but also another was seen, still more vivid, at the inner corner of the 



