The accompanying diagram gives a very clear idea of the arrangement of the apparatus. The "boom" is the pendulum. It is customary to think of a pendulum as hanging down like that of a clock, but this is a horizontal pendulum. Professor Milne has built a very solid masonry column, reaching deep into the earth, and so firmly placed that nothing but a tremor of the hard earth itself will disturb it. Upon this is perched a firm metal stand, from the top of which the boom or pendulum, about thirty inches long, is swung by means of a "tie" or stay. The end of the boom rests against a fine, sharp pivot of steel (as shown in the little diagram to the right), so that it will swing back and forth without the least friction. The sensitive end of the pendulum, where all the quakings and quiverings are shown most distinctly, rests exactly over a narrow roll of photographic film, which is constantly turned by clockwork, and above this, on an outside stand, there is a little lamp which is kept burning night and day, year in and year out. The light from this lamp is reflected downward by means of a mirror through a little slit in the metal case which covers the entire apparatus. Of course this light affects the sensitive film, and takes a continuous photograph of the end of the boom. If the boom remains perfectly still, the picture will be merely a straight line, as shown at the extreme right and left ends of the earthquake picture on this page. But if an earthquake wave comes along and sets the boom to quivering, the picture becomes at once blurred and full of little loops and indentations, slight at first, but becoming more violent as the greater waves arrive, and then gradually subsiding. In the picture of the Borneo earthquake of September 20, 1897, taken by Professor Milne in his English laboratory, it will be seen that the quakings were so severe at the height of the disturbance that nothing is left in the photograph but a blur. On the edge of the picture can be seen the markings of the hours, 7.30, 8.30, and 9.30. Usually this time is marked automatically on the film by means of the long hand of a watch which crosses the slit beneath the mirror (as shown in the lower diagram with figure 3). The Borneo earthquake waves lasted in England, as will be seen, two hours fifty-six minutes and fifteen seconds, with about forty minutes of what are known as preliminary tremors. Professor Milne removes the film from his seismograph once a week--a strip about twenty-six feet long--develops it, and studies the photographs for earthquake signs. [Illustration: Seismogram of a Borneo Earthquake that Occurred September 20, 1897.]