THE
ERUPTION
OF
PELÉE
51
with May 27, 1902, and increasing to Jan., 1903. On Feb. 21-22,1903, the diminution amounted to one-half the normal value of the radiation. I t was assumed t h a t at about this time some of the upper dust fell into the lower region of calm. I n August, 1903, the diminution still amounted to one-tenth of the normal radiation. Among the associated phenomena noted were: decreased intensity of the sky blue; green color of the moon; pink coloring in high clouds; and a bright illumination on lofty mountain summits (Annuaire de la Soc. Météor. de France, Feb., 1905, pp. 40-45; Monthly Weather Review, March, 1905, pp. 101-103). C. G. Abbott, of the Astrophysical Observatory of the Smithsonian Institution, has also placed on record the recognition of " an actual decrease in the transparency of the air, beginning somewhere between Nov. 15, 1902 and Feb. 19, 1903" (Smithsonian Misc. Collections, Dec. 9, 1903). I t is b y no means unlikely t h a t m a n y of the dark days noted toward the latter p a r t of the year 1906 m a y have been associated with the great outpourings from Vesuvius in April of t h a t year. The General Optical Phenomena (Twilight Glows, Bishop's Ring, Anthelia, etc.), as observed at the Observatory of the Pic du Midi, Pyrenees, and at Bagnères. The record of observations made at these points b y E. Marchand and his associates is the most complete t h a t has been made in any locality; indeed, the observations are the only ones touching all the phases of the optical phenomena t h a t have been made b y any set of observers. They are here presented in consecutive form as a translation from the Annuaire de la Société Météorologique de France in the Monthly Weather Review for March, 1905 (pp. 101-103). " I will summarize here, very briefly, the observations made on these phenomena a t Pic du Midi or at Bagnères, b y m y co-workers, Messrs. Ginet, Latreille, Dort, and myself, since the end of the year 1902; observations which have been communicated a t different times to the Société R a m o n d , principally in the meetings of February 3, and April 7 and 12, 1904. " Three principal phenomena have been distinguished, which can be attributed, nevertheless, to the same general cause, the presence of dust, or perhaps of extremely fine particles of ice, in the upper regions of the atmosphere, the dust coming probably from the great eruptions of the volcanoes of the Antilles in May, 1902. These three principal phenomena are: (a) twilight glows; (b) solar or lunar coronas; (c) diminution of solar radiation. And we shall have occasion to mention several others of less importance. (a)
TWILIGHT
GLOWS.
" Two phases of this phenomenon m a y be distinguished : (1) The appearance, fifteen or sixteen minutes after sunset,* of a first twilight segment, pink, purplish, or copper-red, which lasts from twenty to twenty-five minutes and then * " W e h a v e reference here to t h e astronomical sunset, which m a y differ s o m e w h a t from t h e real sunset. At Pic du Midi, t h e real sunset t a k e s place from eight to t h i r t e e n m i n u t e s after t h e astronomical sunset, on account of t h e large depression, 1° 42', of t h e sensible horizon."