We have dwelt at considerable length upon this old clock of the pre-Christian era to show the ingenuity of inventors of that day, and also the careful study that was made of time by ancient mathematicians and astronomers. Of course water clocks were used before the time of Ctesibius. In fact, we read of them in the comedies of Aristophanes, written 400 B. C. Water clocks were used to limit the long speeches of orators at court, and in one place we find Demosthenes accusing a man of “talking in my water,” while at another time, when he was interrupted, he called to the officer to stop the water, showing that he valued every moment of time allotted to him for his speech. In 807 Charlemagne was presented with a clock by the King of Persia. This consisted of an elaborate mechanism in which were all manner of wheels, and the clock would actually strike the hours. The driving power, however, was water. But there were serious disadvantages in the use of water for the measurement of time. No great accuracy was ever obtainable with it, owing to the fact that its volume varied considerably with the temperature, and also with the dryness or moisture in the surrounding atmosphere. The idea of using a weight instead of water is claimed to have originated as far back as 990 A. D. The next important advance in the motive power of clocks was in 1500, when Peter Hele of Nuremberg invented the mainspring. That was long before the pendulum made its appearance. The clock mechanism was slowed down and kept under control by what was known as a balance lever. (See Figure 26.) This was a horizontal lever mounted to oscillate in a horizontal plane. The lever was fitted with sliding weights, so that it could be carefully adjusted. The last wheel of the train of gears was provided with escapement teeth, somewhat similar to those used on our clocks and watches, but which would alternately move the lever this way and that. The inertia of the lever with the heavy weights on it was sufficient to prevent the mechanism from racing, and by this means, the motion was governed and slowed down, so that it measured time with a fair degree of accuracy. DISCOVERY OF THE PENDULUM [Illustration: FIG. 26.--THE BALANCE LEVER, PREDECESSOR OF THE PENDULUM]