VOL. XXX.] PHILOSOPHICAL TRANSACTIONS. 4J3 



well be supposed to have excited an extraordinary quantity of vapour of all 

 sorts; of which the aqueous and most others, soon condensed by cold, and 

 wanting a certain degree of specific gravity in the air to buoy them up, ascend 

 but to a small height, and are quickly returned in rain, dews, &c. whereas the 

 inflammable sulphureous vapours, by an innate levity, have a sort of vis cen- 

 trifuga, and not only have no need of the air to support them, but being agi- 

 tated by heat, will ascend in vacuo Boileano, and sublime to the top of the 

 receiver, when most other fumes fall instantly down, and lie like water at the 

 bottom. By this we may coujprehend how the matter of the meteor might 

 have been raised from a large tract of the earth's surface, and ascend far above 

 the reputed limits of the atmosphere; where, being disengaged from all other 

 particles, by that principle of nature that congregates homogenia visible in so 

 many instances, its atoms might in length of time coalesce and run together, 

 as we see salts shoot in water, and gradually contracting themselves into a nar- 

 rower compass, might lie like a train of gunpowder in the ether, till catching 

 fire by some internal ferment, as we find the damps in mines frequently do, 

 the flame would be communicated to its continued parts, and so run on like a 

 train fired. 



This may explain how it came to move with so inconceivable a velocity: for 

 if a continued train of powder were no larger than a barrel, it is not easy to 

 say how very fast the fire would fly along it; much less can we imagine the 

 rapidity of the accension of these more inflammable vapours, lying in a train 

 of so vast a thickness. If this were the case, as it is highly probable, it was 

 not a globe of fire that ran along, but a successive kindling of new matter: 

 and as some parts of the earth might emit these vapours more copiously than 

 others, this train might in some parts thereof, be much denser and larger than 

 in others, which might occasion several smaller explosions, as the fire ran along 

 it, besides the great ones which were like the blowing up of magazines. Thus 

 we may account for the rattling noise like small-arms, heard after the great 

 bounce on the explosion over Tiverton ; the continuance of which for some 

 time, argues that its sound came from distances that increased. 



What may be said to the propagation of the sound through a medium, ac- 

 cording to the received theory of the air above 300000 times rarer than what 

 we breath, and next to a vacuum, I must confess I know not. Hitherto we 

 have concluded the air to be the vehicle of sound ; and in our artificial vacuum 

 we find it greatly diminished: but we have this only instance of the effect of 

 an explosion of a mile or two diameter, the immensity of which may perhaps 

 compensate for the extreme tenuity of the medium.* 



* See the note on Dr. Halley's former article on such meteors, p. 100, &c. of this volume. 



