A chance clearing of the whole dome came two months after the obliteration of St. Pierre. This we photographed, when brown dust was rising, and steam jets appeared southeast on the dome and in the gulch. On top was an extraordinary spine, shaped like a shark fin, with steep escarpment to the east, curved and smooth and scraped to the west, pushed up and out of a central rupture of the dome. It was like paste from a tube, a hard central pencil of lava that had been shoved up by the expansive force within. Jagged surfaces of breaking showed on the vertical east cliff and long, smooth, arched striations of scrape appeared on the rounded west profile of the protuberance. Other hornlike projections showed on the dome. The summit spine was 200 feet above the surface of the heap. On July 6, 1902, came the first report of the famous Pelée spine. It crumbled in August, and a year later a new spine, facing in the opposite direction, reached a height of 1,000 feet. It was a central tongue of the semisolid lava of the dome, sufficiently plastic to be urged out by forces within. Otherwise the dome was a nearly solid extrusion covered with fallen bombs. This was the magma, or lava, of the Pelée-Soufrière eruptions. Dike ribs extended radially from the spine athwart the dome. I published an erroneous explanation that the dome of boulders consisted of old fragments melted by a superblast and was not true lava. I was so far right, however, as to anticipate the gas-heat theory and melting of all volcanism. The direct crisis of these Carib islands in 1902 was introduced by Soufrière Volcano on St. Vincent, 100 miles south of Martinique, at 1 p.m. on May 7, nineteen hours before the St. Pierre disaster. Soufrière exploded, as the common saying is, through a crater lake pit southwest of its 4,000-foot summit, the crater edge being 3,500 feet high. It is notable how many volcanoes are 4,000 feet high, and how many have crater pits, not at the top, but along a rift below the peak. Just this was the case of Pelée, just this characterizes the calderas of Kilauea and Mauna Loa. A dozen other volcanoes could be named where the vents are through the flank of the heap. Hovey, Curtis, and I were taken by the _Dixie_ to St. Vincent, where the hospitable English colonists provided us with houses at the base of Soufrière, and with servants and horses; and the Government supply steamer took us around the island. We made the ascent of Soufrière to the edge of the great crater and looked down at boiling waters far below, green and muddy, and sending up a column of steam on one wall. We three Americans guided by T. M. MacDonald, a Scottish planter, made the first ascent after the fearful eruptions of May 7 and 18. Leaving our quarters at Chateau Belair, we climbed on foot from the southwest base, with six stalwart negroes carrying instruments, water, and food. In the ruins of Wallibu sugar mill we encountered a wild-eyed East Indian coolie and his helpers looting sugar.