His theory goes to show that the Moon does not act directly upon the weather, but by a tidal action upon the Earth’s atmospheric envelope, which accordingly depends upon the time of the Moon’s syzygies and quadratures. It is this time-factor that has been overlooked and neglected. Reduced to a single sentence, the lunar theory may be stated as follows:—The nearer to midnight any phase of the Moon may occur, the finer will be the weather during the ensuing week, and consequently the nearer to midday it may occur, the more humid will be the weather. The barometer which shows the atmospheric pressure at a mean altitude will consequently be affected in terms of the above statements, and as it ranges from 28° to 30° over these islands, we may also indicate the state of the barometer from the Moon’s phases. For it will be seen that, so long as there is but moderate wind, the barometer will answer to the Moon’s phases in harmony with the weather. Very strong winds without rainfall will affect the barometer and register a fall. I have endeavoured to embody the whole of this theory of the Moon’s action on the weather in a single diagram. [Illustration: Figure 26.] In attempting to account for a very wet summer in the year 1912, and a record downfall of rain in the month of August, I shall have to refer to the times of the phases of the Moon since the solstice. Here are the dates and times, taken from the Ephemeris of Greenwich. June 21. First Quarter 8.39 p.m. Fair. ” 29. Full Moon 1.34 p.m. Wet. Stormy. July 7. Last Quarter 4.47 p.m. Unsettled. ” 14. New Moon 1.13 p.m. Wet. Stormy. ” 21. First Quarter 5.18 a.m. Change. ” 29. Full Moon 4.28 a.m. Change. Aug. 6. Last Quarter 4.18 a.m. Change. ” 12. New Moon 7.58 p.m. Change to Fair. ” 19. First Quarter 4.57 p.m. Unsettled. ” 29. Full Moon 7.59 p.m. Change to Fair. Here there are indications of plenty of variable and changeable weather, with splashes of very wet and stormy weather, and just a suspicion of fair in the whole period.