204 ON THE EFFECTING OF 



strongly illuminated, apparently rising from the surface 

 of the sea' to the upper edge of the mist, at an angle of 

 about twenty degrees from the horizon. In this ellipsis 

 the iridescent colours were not distinguishable. The inner 

 edge was pearly white, with the faintest tinge of blue ; the 

 middle, yellowish, deepening into brown and purple ; the 

 outer edge a blackish blue ; beyond that, a brighter line ; 

 outside of which again lay the cirrostratus mist in its 

 peculiar brown. Within, the ellipsis was bounded by a 

 deep blue line, and the inner space filled with mist of the 

 same colour and illumination as the exterior. 



In one centre of the ellipsis my shadow appeared de- 

 picted, the head surrounded with a circle of the liveliest 

 iridescence. Beyond this was another with similar iri- 

 descence ; but the colours were reversed in order, and more 

 faint ; the belts were also broader. One circumstance 

 surprised me much : whilst the ellipsis rose at an angle 

 from the horizon, the iris circle appeared depicted on the 

 surface of the sea. No account of such a phenomenon 

 having in my recollection been recorded, I thought it 

 might be deemed worthy of consideration. 



The Linnaean Islands run in a curve, bending westward 

 and northward, from the Greenland side across Davis's 

 Straits, and by their resistance prevent the descent of that 

 amazing accumulation of ice to which the name of icy 

 continent is given. In the open spaces between the 



