Observations upon Auroral and Optical Phenomena. 223 



Fiist. The identity of the arch seen by Prof. Olmsted with that 

 seen by myself is proved by several coincidences : by the coinci- 

 dence in the first place, as to the time of its formation, — it being, ac- 

 cording to Prof. Olmsted, at 9h. 51m. P. M., and according to my 

 own minutes at 9h. 55m. P. M. Again, by the coincidence in the 

 time of disappearance, which was put by Prof. Olmsted at 10^. 

 25m., and by myself at lO/t. 30m. Of course, the disappearance 

 being gradual, there would naturally be some difference as to the 

 exact moment, even if the watches of the different observers were 

 just alike, which is not to be expected to a single minute, or perhaps 

 more. My own time was carefully compared, on the following morn- 

 ing, with that of the clocks in New York, and all due allowances 

 made to reduce it to time at New Haven. The faint reappearance 

 of the arch after its first vanishing was not noticed by Prof. Olmsted. 

 Further, there is a coincidence in the azimuthal position of the arches 

 seen by both observers. In their first position, the meridian of both, 

 it may be seen by a comparison upon the globe, cut the horizon 

 at the same point — 15 degrees east of north ;* but in their final po- 

 sition this meridian, according to both observations, was shifted to a 

 point several degrees nearer to the north, — yet all the while the en- 

 tire arch was seen (as it ought to have been) in a more southerly 

 position by the more northerly observer. Again, there was a coin- 

 cidence in the remarkable phenomenon presented by the arch when 

 it broke into parallel fleeces, as described by me, or ridges like par- 

 allel snow drifts, as described by Prof. Olmsted. Lastly, the arch 

 being nearly over head to both observers, and the distance of the 

 stations being only twenty two miles asunder, it is evident that if there 

 had been two similar objects, and those even as low as the region of 

 the ordinary clouds, both would have been visible to both observers. 

 In fact however only one was seen by either. The combined force 

 of these coincidences amounts to a demonstration of identity. 



Secondly. The base line, although only twenty two miles long, 

 was fortunately so situated as to subtend nearly the greatest possible 

 angle and give nearly the greatest possible parallax. The arch was 

 vertical within three degrees to its southern extremity, and the base 

 itself varied in direction but a few degrees from the meridian of the 



* This determination was made, as before mentioned, from the western limb of 

 the arch, the eastern would show its meridian nearer the north point, but still east 

 of it by several degrees — say 11°, 



