In most instances the molten rock stopped short on the flanks of its parent mountain, but it occasionally descended into the plain, as in the familiar example on the Via Appia, where the stream flowed from the side of the Alban volcano for some six miles to within a short distance of the site of the future capital of the world. The lava is here a firm, compact, durable stone admirably adapted for pavements, a use to which it has been extensively put for more than two thousand years, both within the walls and on the great high roads that radiate from them. Here, again, we see how bountifully Rome was favoured in regard to the materials needed for the construction of a great city. The heaping of so much volcanic rock over the surface of the country must have greatly modified its topographical features. The drainage would especially be affected. Streams descending from the Apennines would find their direct passage to the sea blocked by the newly formed ridges, hills and mountains, and they might have to make long circuits before finding an exit. Lakes would gather in the hollows of the irregularly deposited tuff and others would fill up the cavities blown out by explosions, so as to become crater-lakes. The case of the Tiber may be cited in illustration of the deflection of drainage. In earlier times, as I have already remarked, this river probably flowed southwestwards across the site of the volcanic district of Bolsena and Viterbo; but in consequence of the subsequent eruptions, the lower part of its course was buried and the stream, diverted at a right angle, was made to run southeastwards, skirting the volcanic heights until, near Monte Soracte, it reached the plain between the base of the Bracciano and that of the Alban volcano, where it was able at length to find a seaward passage across the site of the future Rome.