1840.] On Lightning Conductors to Powder Magazines. 299 



But let Mr. Daniell try this simple experiment. Let a be a Ley- 

 den jar, b a rod and balbconnected with the 

 inner coating, c c sl rod and ball connect- 

 ed with the outer coating, b represents the 

 excited cloud, c c the excited surface of 

 the earth. To the ball of c apply one 

 branch of a pointed discharging rod, and 

 then rapidly approach the other point to 

 the ball b. A hiss is heard for an instant 

 and a loud explosion then ensues. If the point had been 

 brought very slowly towards the ball b 9 there would have 

 been nothing more than a silent, or at most a hissing, dis- 

 charge. Mr. Daniell will perhaps admit that it is the same 

 thing, that the earth carrying the conductor should approach 

 to the cloud, or the cloud approach to the earth ; and he will 

 see in this experiment that it is simply the degree of velocity 

 of the approach that governs the nature of the discharge. The 

 electric cloud in a calm atmosphere will give off a constant and 

 quiet stream to the rod — but let the cloud be driven onwards 

 before the wind, or drawn within the vortex of mechanical 

 electrical attraction, and then explosion will inevitably ensue* 



18. So much for the cause and nature of the direct lightning 

 flash to a single pointed conductor. Let me beg the Board to 

 honour me with their attention to this distinction, as it is 

 essential that no misconception should arise. 



19. An explosion then, it must be admitted, may occur to a 

 pointed conductor. I have next to shew that this explosion, 

 or the flash, or the amount of the electricity passing (which I 

 use as convertible terms) may be so much greater in quantity 

 than the single conductor can convey in the time of the dis- 

 charge, that a considerable part, nay, the whole of the excess, 

 must pass to the adjacent objects. To make my meaning 

 clearer ; — Let us suppose the cloud to be charged with 1000 

 parts of active electric matter — let us assume the conducting 



2q 



i 



