An introduction to astronomy: designed as a text-book for the use of students in college. By Denison Olmsted ...

PARALLAX. 33
through its densest part, the refraction is greatest. The following numbers, taken at different altitudes, will show how
rapidly refraction diminishes from the horizon upward. The
amount of refraction at the horizon is 34' 00". At different
elevations it is as follows.
Elevation. Refraction. Elevation. Refraction.
0~ 10' 32' 00" 30~ 1' 40"
0 20 30 00 40 1 09
1 00 24 25 45 0 58
5 00 10 00 60 0 33
10 00 5 20 80 0 10
20 00 2 39 90 0 00
From this table it appears, that while refraction at the horizon
is 34 minutes, at so small an elevation as only 10 minutes above
the horizon it loses 2 minutes, more than the entire change
from the elevation of 30~ to the zenith. From the horizon to
1~ above, the refraction is diminished nearly 10 minutes. The
amount at the horizon, at 45~, and at 90~ respectively, is 34', 55",
and 0. In finding the altitude of a heavenly body, the effect
of parallax must be added, but that of refraction subtracted.
90. Let us now learn the method by which the amount of
refraction at different elevations is ascertained. To take the
simplest case, we will suppose ourselves in a high latitude,
where some of the stars within the circle of perpetual apparition
pass through the zenith of the place. We measure the distance
of such a star from the pole when on the meridian above the
pole, that is, in the zenith, where it is not at all affected by
refraction, and again its distance from the pole in its lower
culmination. Were it not for refraction, these two polar distances would be equal, since, in the diurnal revolution of a
star, it is in fact always at the same distance from the pole;
but, on account of refraction, the lower distance will be less
than the upper, and the difference between the two will show
the amount of refraction at the lower culmination, the latitude
of the place being known.
xcamarnple. At Paris, latitude 48~ 50', a star was observed to
pass the meridian 6' north of the zenith, and consequently
3