Astronomical refraction at low altitude varies with temperature, pressure, humidity, temperature gradient, azimuth, colour of light, wind velocity, irradiation, etc., and on the approximate values and nature of them, reached at on some observations and assumptions, I reported at the ninth meeting of the Japan Nautical Society. Now it is impossible either in theory or in practice to correct the refractions for all these factors. So, this time, we have determined to examine the reliability of the Radau's Refraction Table (that is used in our nautical table) and its effect on the observational position line at low altitude. According to our observations, the values of the Radau's Refraction Table (1882) are greater than ours by 1.08% in arithmetical mean, and its variations in probable error are as shown in fig.2. It has been our custom to avoid low altitude observations at less than 10 degrees on account of their low accuracy, but the fall in accuracy is mainly due to great variations of refraction, and we think the graph shows clearly that low altitude observations, as observations at less than 10 degrees are here called according to the custom, are fairly reliable and it is not necessary, in altitudes of more than 3 degrees at least, to expect errors twice as large as those in ordinary observations. Moreover, the variations are greater, though by a small amount, during 40 or 50 minutes after sunrise than they are during the corresponding time before sunset.
抄録全体を表示