SPECTROSCOPIC WEATHER * FORECASTS.

Mataura Ensign, Volume V, Issue 224, 26 January 1883, Page 6

SPECTROSCOPIC WEATHER FORECASTS.

Mb Piazzi Smith, the Scottish Astronomer Royal, has been using the spec-J troscope to forecast the weather, andf has had such success as would, have surely caused him to be burned m? witchcraft had he made his predictions a couple of centuries ago. Mr Smith, who is chiefly known to science as the author of certain remarkable theories regarding the great pyramid of Egypt, has made extensive observations in spectrum analysis as related to atmospheric conditions, and on the 3rd of September he wrote to the Scotsman a letter, which was published next morning, and in which he took the responsibility of advising farmers regarding their crops. This indicated a presumptuous degree of confidence, both on the part of the Professorsand the newspaper. The essence of the statement was as follows : * There is an absence of the rain-band, and a clearing away of all the vapor lines in the spectrum of skylight, to an extent not equalled during the past two or three months. In a powerful spectroscope the two< solar D lines now stand out clear and clean, in place of being almost lost, as all through last month, in a thicket of terrestrial water-vapor lines. So the farmers may be able to gather in their crops at last, dry and in good condition, though, probably, in rather cold and sharp weather.' On the same morning that this appeared the' meteorological office, which corresponds to our signal service, predicted ' unsettled and rainy weather.' The prediction of Professor Smith was fulfilled to the letter, and the weather service was all wrong. Attention was called to the matter at once, and it is claimed that the spectroscope is destined to become a valuable aid to meteorological re r search. Of course, nobody will be disposed to attach any weight to a single success like the one mentioned, but it is reported that a long series of observations are recorded, showing that certain changes of weather can be foretold with reasonable certainty from twenty-four to forty-eight hours ahead, and sometimes with much precision as to time.