Note 359 — Chapter XXIX

These angles were taken both at sunrise and sunset, and with an excellent theodolite, and were repeated after two considerable intervals. The telescopes were reversed after each observation, and every precaution used to insure accuracy; nevertheless the mean of one set of observations of angular height often varied 1 degree from that of another set. This is probably much due to atmospheric refraction, whose effect and amount it is impossible to estimate accurately in such cases. Here the objects are not only viewed through 160 miles of atmosphere, but through belts from between 6000 to 20,000 feet of vertical height, varying in humidity and transparency at different parts of the interval. If we divide this column of atmosphere into sections parallel to those of latitude, we have first a belt fifteen miles broad, hanging over the Khasia, 2000 to 4000 feet above the sea; beyond it, a second belt, seventy miles broad, hangs over the Assam valley, which is hardly 300 feet above the level of the sea; and thirdly, the northern part of the column, which reposes on 60 to 100 miles of the Bhotan lower Himalaya: each of these belts has probably a different refractive power.