Any shaking of the earth's crust produces what may be called an earth-shake or earthquake. The mere falling of a raindrop on the earth produces a slight shaking. The falling of a heavy stone produces a stronger shaking, and sets up a series of minute waves, generally called vibrations, that spread around the place in all directions from where the stone struck. These movements, however, while they spread in all directions, just as they do in a surface of a lake, when a stone is thrown into it, are of course much more quickly stopped by the solid earth than similar movements are by the more readily movable water. But, while any shaking of the earth's crust constitutes an earthquake, yet, strictly speaking, an earthquake is produced only by some force that acts suddenly on the earth, _at a point below its surface_, and, therefore, out of sight. This, of course, would rule out all such shakings as are caused by bodies striking the outer surface of the earth. Earthquakes may occur in any part of the world, and at any time of the day or year. They do occur, however, most frequently in certain parts of the world, at certain seasons of the year and at certain hours of the day. Earthquakes are far from being unusual occurrences. In some parts of the world, such as the island of Java, they are very common, and in Japan, under certain circumstances, scarcely a day passes without one or more shocks in some part of that little empire. Professor Mallet, who has made a very extensive study of earthquakes, published in 1850 to 1858, in the Philosophical Transactions, brief abstracts or descriptions of all the more important earthquakes he could find records of during the past 3,456 years. The number of earthquakes thus recorded during this period reached 6,830. Of this great number nearly one-half occurred during the last fifty years. It should not be inferred from the above figures that the number of earthquakes has really increased so greatly in the past half-century. The explanation of the apparent increase is that greater care has been taken recently in recording earthquakes, and that an apparatus called a _seismometer_, or _earthquake-recorder_, has been invented which automatically produces a record of the smallest shocks; so that a great many have been recorded that would otherwise have passed undetected. It is the opinion of Le Conte that if the records of all the earthquakes of 3,456 years had been thus made there would have been found during the entire time of Mallet's researches to have occurred no less than 200,000, while during the last four years of Mallet's records, the number would have probably reached two earthquakes per week.