Now volcanic cones with craters in their centres occur in great numbers in many parts of the earth's surface. In some districts, like the Auvergne, the Catacecaumene in Asia Minor, and certain parts of New Zealand, these volcanic cones occur by hundreds and thousands. In some instances, these volcanic cones have been formed in historic times, but in the great majority of cases we can only infer their mode of origin from their similarity to others of which the formation has been witnessed. Most of the smaller volcanic hills, with their craters, have been thrown up during a single eruption from a volcanic fissure; but, as Hamilton conclusively proved, the grandest volcanic mountains must have been produced by frequent repetitions of similar operations upon the same site. For not only are these great volcanic piles found to be entirely composed of materials which have evidently been ejected from volcanic vents, but, when carefully watched, such mountains are found undergoing continual changes in form, by the addition of materials thrown out from the vent, and falling upon their sides. This fact will be well illustrated by a comparison of the series of drawings of the summit of Vesuvius which were made by Sir William Hamilton in 1767, and which we have copied in fig. 12. During the earlier months of that year the summit of the mountain was seen to be of truncated form, a great crater having been originated by the violent outbursts of the preceding year. This condition of the mountain-top is represented in the first figure of the series. The drawing made by Hamilton, on July 8, shows that not only was the outer rim of the great crater being modified in form by the fall of materials upon it, but that in the centre of the crater a small cone was being gradually built up by the quiet ejections which were taking place. [Illustration: Fig. 12.--Outlines of the Summit of Vesuvius during the Eruption of 1767.] [Sidenote: CHANGES IN FORM OF VESUVIUS.] If we compare the drawings made at successive dates, we shall find that the constant showers of falling materials were not only raising the edge of the great crater but were at the same time increasing the size of the small cone inside the crater. By the end of October the small cone had grown to such an extent that its sides were confluent with those of the principal cone, which had thus entirely lost its truncated form and been raised to a much greater height. The comparison of these drawings will be facilitated by the dotted lines, which represent the outline of the top of the mountain at the preceding observation; so that the space between the dotted and the continuous line in each drawing shows the extent to which the bulk of the cone had increased in the interval between two observations.