116 ANNUAL OF SCIENTIFIC DISCOVERY. 



of evening- twilight, in the form of a northern light, resembling the 

 dawn ; that it usually arrives at its maximum at all places, however 

 differing- in longitude, at the same part of the night, namely, from ten 

 to eleven o'clock, but more frequently a little before eleven ; and that 

 auroras of the highest order frequently continue all night, while those 

 of an ordinary character commonly end before midnight. 2. That a 

 great aurora is usually preceded by a large bank or cloud of a peculiar 

 vapor, differing in its nature from ordinary clouds, commonly exhibit- 

 ing a milky appearance, but sometimes of a smoky hue, or the two 

 mixed together ; and that the extent and density of this auroral vapor, 

 resting upon the northern horizon, form the best prognostic we have 

 of the probable intensity of the exhibition which is to follow, com- 

 prising the material of which the successive forms of the aurora are 

 constituted. 3. That the auroral waves, when peculiarly grand, make 

 their appearance later than the streams and arches, and usually later 

 than the corona, continue to a later hour of the night, appear at a 

 lower level than the streamers, and roll upwards, in the direction of 

 the streamers, toward the point of general concourse. 4. That auroral 

 exhibitions of the higher order are commonly of great extent, spread- 

 ing over no inconsiderable part of the earth's surface, and reaching to 

 a great but variable height. 5. The auroras of the first class have 

 three distinct forms of periodicity ; a diurnal periodicity, commen- 

 cing, arriving at the maximum, and ending, at different hours of the 

 night, as already asserted ; an annual periodicity, rarely or never oc- 

 curring in June, and the greatest number of the highest order cluster- 

 ing about November, these last bearing a striking resemblance to each 

 other ; and a secular periodicity, the most remarkable of all, recurring 

 in great series, which we have denominated " auroral visitations." 

 That the visitations most marked and best defined occur at intervals 

 of about sixty-five years, recurring from the middle of one period to 

 the middle of the next period, and last from twenty to twenty-two 

 years, making the interval from the end of one to the beginning of the 

 next about forty-five years. 6. That, while the forms of the aurora 

 usually appear to be under the control of magnetic forces, yet this is 

 not always the case, since the arches do not always culminate in the 

 magnetic meridian, nor do they always place themselves at right 

 angles to that meridian, nor does the effect on the needle correspond 

 to the different states of intensity of the aurora. 7. That the aurora 

 has remarkable geographical relations, belonging chiefly to the higher 

 latitudes, and only in the great visitations descending below the lati- 

 tude of 40 ; but descending lower on the western than on the eastern 

 continent, and prevailing more in the northern than in the southern 

 hemisphere. 



Professor Olmsted then gave his reasons for thinking that the aurora 

 is not produced by electricity or by magnetism, though it apparently 

 has some relation to the latter in the forms and positions of its arches, 

 and in other respects, after which he stated his grounds for seeking 

 for the explanation of the phenomena in the planetary spaces, and 

 argued in favor of its cosmical origin ; first, from the extent of the ex- 

 hibitions, which is greater than could arise from any terrestrial ema- 



