1892.] VEEDER THUNDERSTORMS. I35 



a single year, and would be very plainly perceptible in tables such as 

 those which have been constructed. Thus there is no liability what- 

 ever to confusion with the time of a synodic revolution of the moon 

 which differs more than two days from the period of auroral recur- 

 rence. On the other hand the conformity with the rotation period of 

 the sun is exact, and is capable of verification in ways that will 

 appear in the further course of the discussion. 



The evidence being so clear that there is recurrence of auroras 

 at the interval of a synodic revolution of the sun, it becomes of inter- 

 est to compare their daily prevalence with the coincident condition of 

 the sun's surface. For this purpose suppose the entire surface of the 

 sun to have been divided meridionally into as many lunes as there are 

 days in the synodic interval, and make for each of these divisions a 

 list of the sizes of all sun-spots observed each day throughout the 

 year as given in the Greenwich Records of Photographic Results. 

 From these records also it is known where each lunewas located upon 

 any given day or series of days. Upon the first day of each auroral 

 or synodic period for example, each lune returns to a particular part 

 of the sun's disc, which in turn is occupied on the next day by the 

 lune following and so on. By this method it is learned that upon 

 series of dates characterized by recurrence of the aurora, disturbed 

 portions of the sun are always at the eastern limb appearing by rota- 

 tion. It is true that for many years past instances have been numer- 

 ous in which at the time of an aurora a disturbance has been located 

 just west of the meridian, but this latter relation is adventitious and 

 depends upon the fact that the chief seats of solar activity recently 

 have been three in number and about nine days apart so that when 

 one is at the eastern limb another is apt to be just beyond the meridian. 

 In years when there is cessation of activity at one or more of these 

 centers it becomes evident that it is the eastern limb effect that 

 persists, that assumed to proceed from the meridian disappearing when 

 the disturbances are far enough apart to avoid any possibility of 

 confusing the one with the other. It is to be noted also that at times 

 when all these centers are active at the proper interval from each 

 other there is increased liability to confuse the direct effect of a distur- 

 bance at the eastern limb with that of the one next preceding because 

 of the fact that when an aurora appears there seems to be a reactionary 

 inductive effect from the earth itself towards the sun which tends to 

 increase the size of any sun spots near the meridian in a manner that 

 will be more fully explained in a subsequent paragraph in which sun 

 spots are shown to depend upon a reflux of magnetic induction toward 

 the sun whenever its surroundings become more highly charged. 



