The following account of earthquake sounds is given by Daubeny, in his book on volcanoes. It appears that during March, 1822, the people living on the island of Melida, opposite Ragusa, in Dalmatia, were greatly alarmed by sounds that at first they believed due to cannonading either at sea or on the neighboring coast. They afterwards found that these sounds were due to something that was taking place under the ground. The noises continued at intervals until August 23d, 1823, when a great earthquake occurred, during which one of the highest mountains on the island was cleft or split in one place. The underground noises continued from time to time and so frightened the people that they were about to leave the island permanently and emigrate to the mainland of Dalmatia. They were dissuaded from doing so by the government, and while the noises continued at intervals it so happened that no damage came to them. It is said, however, that twenty years after an active volcano broke out on the island. There are various causes that produce earthquake sounds. A very slight rubbing or grinding together of rock surfaces may produce fairly loud noises, the volume of the sound being increased by transmission through the rock masses that lie in the path of the waves. An example of such an increase in the loudness of sounds is seen in the case of several of the large blocks of stone used for some of the piers of Kingston Harbor, in Ireland. When these rocks are moved together by blows of the waves they produce loud and appalling sounds, as if the whole island were being washed away. The same rocks, however, when left high and dry on the falling of the tide, can be caused to rub together, when moved by the hand. Under these circumstances, they produce but feeble sounds that can only be heard in their immediate neighborhood. No doubt, some find it difficult to understand how it is possible for comparatively feeble sound-waves to be strengthened by their passage through large masses of solids. This is important and should be made clear. As everyone well knows, the ticking of a watch can only be heard at a short distance when the watch is held in the hand, because the sound-waves cannot readily pass through the body of the person holding the watch to the earth, the materials of the body not being sufficiently elastic. If, however, the watch be placed on the bare surface of a large wooden table from which the tablecloth has been removed, so that the watch can come directly in contact with the wood, and nothing else is placed on the table but the watch, the sound-waves are transmitted to the mass of the table and its entire surface sends them out into the air. The ticking of the watch can then be heard distinctly in almost any part of a large room. Mallet states that in nearly all great earthquakes sounds are heard before the principal shock, and in his description of the Calabrian earthquake Hamilton says: