THE determination of position from land or air observations in high latitudes is rendered difficult, apart from conditions of temperature, by the unreliability of the magnetic compass and gyroscopic compass. On aircraft the latter is not yet a practicable instrument. Hence, astronomical methods must be utilised. In the consideration of any method of air navigation to the poles it must first be made clear that, even with assumed perfectly accurate instruments and accurate observations with them, astronomical observations from an aeroplane can give an exact position in two cases only: (1) by the observation of two stars (2) by choosing the time of flight so as to render possible simultaneous observations of both sun and moon. The first case implies that the flight takes place at the season of perpetual night, and need not therefore be considered. The limitations of the second case are increased by the known inaccuracy of any form of sextant used in the air. Broadly, while sextant observations on the ground may be relied upon to give position within, say, one to three miles, aerial observations can only be relied upon to give position within about ten miles, and it is usually necessary to take the mean of a number of readings to secure this result. The use of automatically stabilised aircraft may render a somewhat more accurate location possible, but the utility of such aircraft under polar conditions can scarcely be regarded as established.