| 15 0 | 3 17 || 50 | 0 45 || 85 | 0 5 | +-------+---------++----+---------++----+---------+ | 16 0 | 3 4 || 51 | 0 44 || 86 | 0 4 | | 17 0 | 2 53 || 52 | 0 42 || 87 | 0 3 | | 18 0 | 2 43 || 53 | 0 40 || 88 | 0 2 | | 19 0 | 2 34 || 54 | 0 39 || 89 | 1 1 | | 20 0 | 2 26 || 55 | 0 38 || 90 | 0 0 | +-------+---------++----+---------++----+---------+ [Sidenote: PLATE II. The inconstancy of Refractions. A very remarkable case concerning refraction.] 183. In all observations, to have the true altitude of the Sun, Moon, or Stars, the refraction must be subtracted from the observed altitude. But the quantity of refraction is not always the same at the same altitude; because heat diminishes the air’s refractive power and density, and cold increases both; and therefore no one table can serve precisely for the same place at all seasons, nor even at all times of the same day; much less for different climates: it having been observed that the horizontal refractions are near a third part less at the Equator than at _Paris_, as mentioned by Dr. SMITH in the 370th remark on his Optics, where the following account is given of an extraordinary refraction of the sun-beams by cold. “There is a famous observation of this kind made by some _Hollanders_ that wintered in _Nova Zembla_ in the year 1596, who were surprised to find, that after a continual night of three months, the Sun began to rise seventeen days sooner than according to computation, deduced from the Altitude of the Pole observed to be 76°: which cannot otherwise be accounted for, than by an extraordinary quantity of refraction of the Sun’s rays, passing thro’ the cold dense air in that climate. KEPLER computes that the Sun was almost five degrees below the Horizon when he first appeared; and consequently the refraction of his rays was about nine times greater than it is with us.” 184. The Sun and Moon appear of an oval figure as _FCGD_, just after their rising, and before their setting: the reason is, that the refraction being greater in the Horizon than at any distance above it, the lowermost limb _G_ appears more elevated than the uppermost. But although the refraction shortens the vertical Diameter _FG_, it has no sensible effect on the horizontal Diameter _CD_, which is all equally elevated. When the refraction is so small as to be imperceptible, the Sun and Moon appear perfectly round, as _AEBF_. [Sidenote: Our imagination cannot judge rightly of the distance of inaccessible objects.]