Sir Henry Holland thus alludes to the phenomenon:—“I am tempted to notice a spectacle, having a certain association with this science, which I do not remember to have seen recorded either in prose or poetry, though well meriting description in either way. This spectacle requires, however, a combination of circumstances rarely occurring—a perfectly clear Eastern and Western horizon, and an entirely level intervening surface, such as that of the sea or the African desert—the former rendering the illusion, if such it may be called, most complete to the eye. The view I seek to describe embraces the orb of the setting Sun, and that of the full Moon rising in the East—_both above the horizon at the same time_. The spectator on the sea between, if he can discard from mental vision the vessel on which he stands, and regard only these two great globes of Heaven and the sea-horizon circling unbroken around him, gains a conception through this spectacle clearer than any other conjunction can give, of those wonderful relations which it is the triumph of astronomy to disclose. All objects are excluded save the Sun, the Moon, and our own Globe between, but these objects are such in themselves that their very simplicity and paucity of number enhances the sense of the sublime. Only twice or thrice, however, have I witnessed the sight in its completeness—once on a Mediterranean voyage between Minorca and Sardinia—once in crossing the desert from Suez to Cairo, when the same full Moon showed me, a few hours later, the very different but picturesque sight of one of the annual caravans of Mecca pilgrims, with a long train of camels making their night march towards the Red Sea.”[3] It is due to the same cause that the Sun and the Moon when very near the horizon may often be noticed to exhibit a distorted oval outline. The fact simply is, that the upper and the lower limbs undergo a different degree of refraction. The lower limb being nearer the horizon is more affected and is consequently raised to a greater extent than the upper limb, the resulting effect being that the two limbs are seemingly squeezed closer together by the difference of the two refractions. The vertical diameter is compressed and the circular outline becomes thereby an oval outline with the lesser axis vertical and the greater axis horizontal.