came aware that it was not correct for the meridian of Rome, although the latitude of Catania differs from that of the capital by more than four degrees. From that time forward sun-dials came into general use; and since many have been recovered their construction is well known. The most common form is that of a concave hemisphere cut in two. Within one of these quarters the hours were marked. A stylus or hand fastened in the top indicated the time of day, when the sun shone. Pliny says the first water-clock was set up in Rome 159 B.C. These water-clocks appear to have differed from the clepsydræ that had long been in use in the countries farther east. They consisted of an earthen vessel tapering downward to a point, in the bottom of which there was a small hole through which the water flowed in a given time. It was comparatively easy to ascertain when the sun was on the meridian; but not so easy to determine the exact period of midnight. This was moreover, an affair of small practical importance. In the larger cities, the periods or hours were announced by the sound of a trumpet; in the country few persons cared how the hours of the night passed. The custom of proclaiming the hours of the night prevailed in some countries of Europe, especially in Germany, long after clocks had come into almost universal use. It is not known when the Romans began to divide the day into twenty-four hours. At any rate there were two kinds of days in vogue: the astronomical day, the hours of which were all of the same length, and the civil or ordinary day which corresponded with the former at the equinoxes only. The popular day was a matter of latitude. In Rome the longest contains somewhat more than fifteen hours according to mathematical calculation, but owing to the Appennines which lie east of it the fact does not quite correspond with the figures. The hour in Rome was therefore at one time of the year about seventy-five minutes in length, while the hours of the night were correspondingly shorter, and vice versa.
Every schoolboy is taught that twelve inches make a foot, but not one in a million thinks to ask what is the basis of this measurement. It must at once occur to the occasional inquirer that the average human foot is not twelve inches long. When, however, a unit of measurement has been once fixed, the rest is easy. The metric system was the first attempt to establish an invariable standard to which recourse could always be had in cases of doubt. A table before me gives twenty-six different lengths for the foot in the German empire, twenty-five for the rest of Europe, eight for America and four for Asia. Of these the longest is that of Lombardy, which contains a little more than 435 millimeters, the shortest the foot of Siam, which is only 245.6 mm. Even in Germany the foot varies from 429.5 to 250 mm. There is of course the same divergence between the square and the cubic foot. The Eng-
