156 PHILOSOPHICAL TRANSACTIONS. [aNNO J720. 



or flashes from thence to the zenith, expanded at every flash, like a broad, thin, 

 white cloud, of which some faint view could be seen after each explosion for 

 some time; and after all the explosions were over, there remained a thin duskish 

 vapour in and near the zenith, and all along the arch from east to west, from 

 14 to 20 degrees broad, which undulated and moved like a stormy sea, the 

 motion coming from the s. s. e. and so lessened till it appeared no brighter than 

 the milky way, but more like a very thin cloud or mist, through which the 

 stars appeared. At the same time there was another thin cloud, having the 

 same appearance, arch ways, to the southward, at about the height of 40 

 degrees, which Mr. D. supposed had been another, which had been over, and 

 had moved thither from the northward before he went out; and during the 

 whole time there were lesser lights towards the north, but dispersed here and 

 there, and not forming any large body of light. During the whole time, 

 the hemisphere was clear, except a few very small clouds near the horizon ; and 

 when any moved into the enlightened arch they broke the connection, so that 

 the light was above them, at the same time it froze hard each night. 



From these observations, Mr. Dobbs supposes, that the aurora borealis is a 

 thin nitro-sulphureous vapour, raised in our atmosphere considerably higher 

 than the clouds, which is discontinued in several places by the interspersed air, 

 and which by pressure and motion is kindled ; and perhaps the explosion of one 

 may by its shock and motion contribute to kindle the next; by which means 

 they go off one after another, till the whole vapour within their influence is 

 discharged, and then the light disappears, and the thin smoke appears, and 

 undulates, according to the motion in that part of the atmosphere. Hence he 

 thinks most of the appearances may be solved; for first, as to the continued 

 light near the horizon, they being at a great distance from us, and nearly in a 

 line, all these explosions may seem as a continued light; when these approach 

 nearer to us, and so appear higher in our hemisphere, we observe the motion in 

 each flash, and still seeing them laterally, yet somewhat breaking the conti- 

 nuity of the light; they, by the reflection of the vapour floating in the atmos- 

 phere, and being not reflected, where the air between them is free of those 

 vapours, may appear as columns; and as the flash below and beyond them 

 moves, as it kindles and expands, so they seem to move, and perhaps are 

 shocked at the same time by the motion; but afterwards, when they are nearer 

 and raised to the altitude of 40 degrees, we get somewhat under them, and see 

 the expansion of the explosion, which appearing somewhat globular, gives the 

 faint colours observed above, the light not being intense enough to make them 

 vivid; and afterwards when they rise to, or near the zenith, they are nearest 

 to us, and then they expand very wide at each flash, like little clouds; and the 



