The inner walls showed horizontal bands of old lava, and intrusions both in lens shape and as dikes. There were red brown puddingstones made up of fragments. A funnel-shaped intrusion looked like the diagram cross section of a volcano, making a perfect T of gray lava, like a mushroom. A large fissure, filling west, rose from bottom to top. A northern rocky horseshoe rim, or somma, at the top made the peak of St. Vincent. The crater lip was a mile wide and the interior a half mile deep; and the green puddle at the bottom was 1,200 feet across. The base of the wall column sputtered fiercely and sent up spurts of black mud and rock fragments. The lake level was 1,100 feet above the ocean, 800 feet lower than before the eruption; and the pool was shallow, with mud flats and islets. We operated cameras, compass, and sketch books; paced off a base line; and noted that the northwest corner of the crater had been blown away to leave a big notch. When we returned to Chateau Belair, the negro peasant women brought out their children to gaze at us, the godlike men who had dared the crater. Mr. MacDonald had to steer us through the crowd, and we felt like the twelve apostles after a miracle. The Soufrière eruption during the first week of May was more voluminous and violent than that of Pelée, for Pelée was concentrated on one target. Soufrière wrought havoc east and west, whereas Pelée was in a sector southwest of the mountain. They were equally devastating, however, and both made downblasts of superheated steam and gravels. Scalding dust killed people, but so did water waves, conflagration, steam, stones, drowning, and burial. Soufrière’s dust fall was reported all the way to Trinidad and Barbados; and from ships east and southeast, directly against the trade winds, from 100 to 900 miles away. The dust column penetrated the antitrades of the upper atmosphere. Sounds were loud 150 miles away, but not heard close to the mountains. In the red hot gravel were innumerable landslides, river waters rushed into the gravel and made false eruptions, and shore cliffs collapsed. No lava, except as fragments, appeared in St. Vincent, whereas it rose as a crateral heap in Pelée. Floods of rivers radial to the volcanoes appeared both before and after the first eruptions, and scientists erroneously attributed them to cloudburst rains. Later, exact descriptions by natives showed that the sources were hot waters gushing out in places where there was no rain. A succession of eruptions at increasing intervals from May to December actuated both volcanoes. In succeeding years, explosions dwindled; but over Pelée’s crater rose a mighty dome and spine of stiff quartz-basalt lava, like ointment from a tube.