In all probability, this is what really happened. There were two sorts of sieges. One was that described above, which only consisted in heaping up rubbish against the wall. Or else, with huge toil they undermined its foundations, and shoring it up with a framework of strong beams, set fire to the timber, and burned it, so that the masonry might come down with a crash. A positive mention of battering rams does not occur before the Peloponnesian war; and among the Romans even somewhat later. If Veii was really taken by means of a _cuniculus_, it is to be explained by the second mode. The draining of the Alban lake must surely belong to this period. We cannot gainsay it, nor is there any reason for putting in here a work of earlier date. It is likely that owing to some stoppage in the channels by which it was drained, there was danger of the lake’s overflowing the whole of Latium; and possibly they may have taken advantage of the credulity of the people to stir them to this immense undertaking, though I believe that, if the senate decided upon this necessary work, it readily found obedience. It is to be supposed that the Alban lake had a subterraneous outlet through clefts, like the Fucinus and all lakes which have been formed in the craters of volcanoes: these chasms may have been filled up by an earthquake. Livy, somewhat further on, speaks of a severe winter when the Tiber was covered with ice, and of a sickly summer which followed it. The newly discovered _excerpta_ of Dionysius place the construction of the tunnel in the year after that severe winter. Livy says that during that winter the snow lay seven feet deep, and that the trees were killed by the frost; a statement quite in the style of the annals, which, although the old annals were lost at the Gallic invasion, is yet very credible, as that winter must have survived in the memory of all. Just as severe was the winter of 483, when the snow lay for forty days on the Forum. The earlier Roman history shows traces of the average height of the thermometer having been at that time much less than it is now.[115] In Roman and Greek history, the periods of extraordinary appearances in the weather are almost always the precursors of frightful earthquakes: thus an eruption of Ætna happens at this time (354). Vesuvius was then quiet; yet the earthquakes were awful. By one of these, the outlets of the Alban lake may have been stopped: generally speaking, all lakes which have no _emissarius_, exhibit wonderful periods of rise and fall. The lake Copais even had artificial drains, which, however, were afterwards choked up, and Bœotia during the Macedonian era was not able to pay the cost of clearing them; the consequence of this was, that the lake began to swell, and overflowed the country all about. On the whole, as Aristotle has already remarked, Greece may have lost in the supply of water. The lake Copais is at present merely a marsh,