Published online by Cambridge University Press: 18 January 2010
Every navigator is familiar with the necessity for correcting sextant observations of altitude for the effect of atmospheric refraction. He does this by means of tables which the great majority of navigators are compelled to take on trust, as they would a table of haversines. Unfortunately it is much easier to guarantee the accuracy of a table of a straightforward mathematical function than it is to tabulate accurately an optical effect occurring in a notoriously variable atmosphere. Recent American work based on large numbers of marine observations has to some extent called into question the accuracy of the low-altitude portions of the usual tables, and it therefore seems worth while to give a brief account of refraction theory and also to consider how far it is confirmed by observation. The treatment will be restricted to refraction as it affects marine navigation. This is a considerable simplification. Aerial observations may be made at any level between the ground and the lower stratosphere, so that air navigation involves a much greater range of pressures and temperatures at the position of the observer. Moreover, it involves some occurrence of much more strongly negative altitudes than can be observed from bridge height on a ship, and at such negative altitudes (i.e. zenith distances over 90°) the intrinsic variability of the refraction is large enough to make any table rather unreliable.