356 -Water-fronts. [Book IV. 



It was inti mated that water- fpouts were among the 

 phenomena, which fome philosophers have attempted 

 to explain on electrical principles. A water-fpout is a 



generally repeated many times in a minute. They appear much 

 broader, more diffofe, and of a weaker light than the beams : they 

 grow gradually fainter till they difappear ; and fometrmes con- 

 tinue for hours, flafliing at intervals. 4thly, Arches, nearly in the 

 form of a rainbow : thefe, whu>n complete, go quite acrofs the 

 heavens, from one point of the horizon to the oppofhe point. 



When an aurora happens, thefe appearances fcem to fucceed 

 each other in the following order : i. the faint rainbow-lik.fi 

 arches : 2. the beams ; and, 3. the fiafhes. As for the northern 

 horizontal light, it appears to confilt of an abundance of fiafhes, or 

 beams, blended together by the fuuation of the obferver. 



' The beams of the aurora borea'is appear at all places to be 

 arches of great circles of the fphere, with the eye in the center ; 

 and thefe arches, if prolonged upwards, would all meet in one 

 point. 



' The rainbow-like arches all crofs the magnetic meridian at 

 right-angles. When two or more appear at once, they are con- 

 entric, and trnd to the call and weft ; alfo the broad arch cf the 

 horizontal light tends to the magnetic eaft and wed, and is bi- 

 .fecled by the magnetic meridian ; and when the aurora extends 

 ver any part of the hemifphere, whether great or final!, the line 

 fcparating the illuminr.ted part of the herqifphere from the clear 

 part, is half the circumference, of a great circle crofting the mag- 

 netic meridian at right-angles, and terminating in the eaft ar.d 

 weft : moreover, the beams, perpendicular to the hori/.on, are only 

 . thofe on the magnetic meridian. 



' That point in the heavens to which the beams of the aurora 

 appear to converge, at any place, is the fame as that to which the 

 fouth pole of the dipping needle points at that place. 



' The beams appear to rife above each other in fucceffion ; fo 

 ..that of any two beams, that Which has the higher bafe. has alfo the 

 higher fummir. 



Every beam appears broadcft at or near the bafe, and to grow 

 narrower as i afcends j fa that the continuation of the bounding 

 lines would meet in the common center to which the beam tends. 



' The height of the rainbow-like arches of the aurora are ef!i- 

 mated by Mr. Dahon to be above the earth's furfacc about 150 

 milis.' ^(iarm's Le3ttr,s, vol. iv. p. 542. 



