H. A, Newton on the height of Auroral Arches. 287 
This curve is not caused by mists in the atmosphere obscuring 
_ and revealing parts of an indefinite cloud. For the arch has 
little or no relation to the horizon, and cutsit at all angles. 
_ Itis not a straight line, for the arch does not cut the horizon 
at points 180° from each other. 
_ The arch resembles the projection of a portion of a circle, or 
_ asphero-conic. The venerable Hansteen has in two instances 
‘en at Christiana nearly the whole ellipse. Prof. Twining has 
observed, at Middlebury, Vt., in one instance at least, an arch in 
which the extremities of the major axis of the ellipse were vis- 
above the horizon. 
jing into space beyond. The bank of auroral light is a similar 
broader or more distant ring. 
The results of Prof. Loomis’s investigations respecting the 
ee cal distribution of the aurora” confirm and modify this 
i 
‘form surrounding the magnetic and astronomical poles of the 
ein, an should be regarded as part of a circle whose cen- 
Académie Royale de Bruxelles, tome xx, p.118. See also 
Marsh, this Journal [2], ah 811. es 
al [2], xxx, 89, 
