escape. Having, probably, found one or more points of discharge far below the summit of the cone, it flowed out in such vast quantities that it left a cavity large enough to engulf the whole of the unsupported mountain mass. The sinking was doubtless aided, and lessened in violence, by the partial fusion of the overlying rocks as they became more and more depressed, and probably the action took place around a common center. When the mountain summit had completely disappeared, there was left around about a regular curve of unbroken walls bearing witness to the comparative gentleness with which the action had been carried out. It is possible to consider the central Aso range as part of the old mountain that did not sink or become totally engulfed, but it seems more likely that it is a later growth. The completed work probably left the whole of the sunken mountain melted in a level lake within the great caldron. The radiating lava flows described in a later paragraph may help to account for the material removed.
After nearly two weeks spent in and about Aso we left it, setting out eastward to continue our march across Kiushiu to the Pacific, on the opposite side of the island from our starting point. The less precipitous portions of the crater wall are well-watered and clothed with beautiful groves of pines and cryptomerias, bamboos, oaks and chestnut trees, among which one finds little meadows and mossy places and banks overgrown with rich grass, where thrive an abundance of wild violets of various colors and sweet-smelling daphnes. Through these woods our road wound up out of the pit at a comparatively low and gently-rising portion of the wall, and finally over the crest of the rim to the far-sloping outer reaches. Within a few days more we