It would seem, therefore, that either these shocks or breakages in the earth's crust, or the earth's interior, must be very great indeed, or else that the earth must be composed of rather shaky materials. Well, perhaps both these suppositions are true. We spoke just now of the instruments which seismologists use to record earthquakes. They are known as "seismometers," and a great many of them are used in Japan and on the Californian or Pacific coast of America. Now it is perhaps scarcely necessary to say here (when we recollect how many cyclones and anticyclones England receives from the Atlantic) that a storm or rainy weather is usually heralded or accompanied by a fall in the barometer, or a depression. Now when there is a depression in the barometer that means that the weight of air above the barometer is less than it was before, though it is not so great a difference that human beings could tell it, unless it were accompanied by other signs. But the earth can tell it, and the mere fall of the barometer, owing to changes of the air, will make the earth tremble or quiver slightly, as if it were a jelly. We cannot perceive it; but the delicate seismometers can; and when a storm is coming to Japan or to California from the Pacific, the instruments show that the earth feels the passage of it. The comparison of the earth to a jelly--a very stiff jelly--is on the whole a useful one. If a very tall jelly is allowed to stand for some time, or if the table on which it stands is shaken a good deal, then, as we know, rifts will sometimes appear in the jelly. The reason for these breakdowns in the jelly's composition is that owing to the distribution of its weight it is always in what we call a state of strain; and it is sometimes not strong enough to support this strain, and, almost without apparent cause, will sometimes give way. Much more solid bodies than jelly act in the same way. The great bridge near Quebec which collapsed in 1907 was to all appearance quite sound and strong; but there were strains in the iron girders, and without warning these strains suddenly produced rifts in the iron and steel framework and it broke down. Similarly the towers of churches and cathedrals, which are built on arches, will give away quite suddenly after standing to all appearance quite firm for hundreds of years. There is an architect's maxim which runs, "The arch never sleeps." That means that the arches on which the great weight of a church or cathedral tower rests are always in a state of strain; they are always, as it were, imperceptibly quivering; and they are always liable, if the strain on them should be increased in the slightest degree, to give way, or to resettle the weight on their shoulders in some way.