but in the centre a small cone of about two metres had formed, pointed in such a manner that it possessed but a very narrow opening at the apex, from which smoke issued with a hissing sound, and from which were spurted a few very small incandescent scoriæ. This little cone increased in size as well as activity until it filled the crater, and rose four or five metres above the brim.[A] New and more abundant lavas appeared near the base of this cone, and, pouring continually into the Atria del Cavallo, rushed into the Fossa della Vetrana in the direction of the Observatory and towards the Crocella, where they accumulated to such an extent as to cover the hill-side for a distance of about 300 metres; then turning below the Canteroni, they formed a hillock there without spreading much farther. These very leucitic lavas are capable of great extension, the pieces which are ejected forming for the most part very fine filiform masses, which may be collected on the mountain in great quantities, and specimens of which I presented to the Academy under the name of _filiform lapilli_. These threads were often of a clear yellowish colour, and, when observed under the microscope, were found to consist of very minute crystals of leucite embedded in a homogeneous paste. The crystals were still smaller as the diameter of the threads was less, and never formed knots or swellings even in the most hair-like threads. These observations led me to reject the opinion of those who hold that crystals of leucite are pre-existent in the lava. The viscous nature of these lavas prevented their being covered with fragmentary scoriæ, but caused the formation at first of a skin, which, thickening, became at last a more or less pliable shell, that, when more solidified, allowed the still fluid part to run as in a tube formed of this solid shell. For many months the lava descended thus from the cone and traversed the Atria del Cavallo, always covered, appearing below the Canteroni of a lively fluidity, until it could no longer be enveloped in its skin, which was stretched by the addition of new lava, and finally rent asunder to give room to the current until, owing to diminished liquidity, it was constrained to stop. When the lava, having traversed the covered channel it had made for itself from the top of the mountain to below the Canteroni, made its appearance still running, it frequently formed large bubbles on the surface, which mostly burst to give vent to smoke, and then disappeared.