A DESCRIPTION of the scene of devastation in Martinique was published by the writer in this magazine in August, 1902. Some of the readers of The Popular Science Monthly may be interested in the details of a great eruption, and scientific deductions from observation of the same; the present article aims to present the results of such observation. The writer returned to Martinique from Barbados in June, 1902, and had the good fortune to see an eruption of first magnitude on Mount Pelée. On the evening of the ninth of July, he was in Fort de France, and the commission sent out by the Royal Society, Drs. Tempest Anderson and John S. Flett, were living on board the sloop Minerva near Carbet; thus the scientific record kept of this eruption was unusually complete.
At 8:15 in the evening, just at twilight, there was seen from Fort de France a very high column of black dust with the characteristic cauliflower surface, boiling straight up from the direction of the volcano to an enormous height (compare Fig. 1). The blackness of this dust cloud increased, the bulbous profile gave place to a smooth balloon-shaped outline slightly flattened above, and from the moment that the interior of the cloud first became obscure, small spicular lightnings were seen dancing like a myriad of bright white sparks all over the cloud surface. At first this whole extraordinary display was perfectly silent; the writer was called out from the public library in Fort de France by an attendant at about 8:20 p.m., and there was a singularly oppressive stillness noticeable. When the phenomenon was first observed in the streets, some slight disturbance akin to panic had been audible, but the people as a whole had grown so callous to these happenings, that after a few moments very few individuals seemed to take any note of the extraordinary phenomena in the northwestern sky.
At 8:30 the cloud had the form of a balloon, swelling rapidly to the zenith, and filling with its diameter perhaps 70° of the sky north-northwest. The wonderful coruscations grew more brilliant, leaping rapidly from place to place over the surface of the cloud, but at first they had the form of points of light, rather than lines. There was no thunder accompanying these early flashes, and their frequency was incessant, the whole cloud surface fairly scintillating with dancing lights.