Just when or by whom the instrument was thus scientifically perfected is not known. The calculations necessary to the projection of the hour lines upon a flat surface could hardly have been performed before Greek times. The Greeks ascribed the invention of the sundial to Anaximander, in the sixth century B. C., but sundials of various types had been known in various parts of the world long before then. On the other hand, the Hemicycle remained the common form of the instrument all through the classic period and even afterwards. The Babylonians were quite capable of understanding the principle of the sloping gnomon. And once this was discovered, it would have been entirely practical to set up the new dial beside a Hemicycle or Clepsydra, and find the angles of the hour lines by experiment. These, once laid out correctly, would be determined once for all. Even at its best the sundial had certain very marked limitations. Scientifically constructed, it would keep accurate time according to the visible sun. But it could not be read accurately unless made inconveniently large. It was inaccurate when removed from its original latitude, or displaced from a true North and South position; so that in any portable form it became a very rough measure indeed. Moreover, it was of course entirely useless at night or in bad weather or in shadow. And finally, it was never absolutely exact under the most ideal conditions, because of what is known as the Equation of Time. The Earth does not, in fact, move around the sun at an absolutely regular rate of speed; it moves a trifle faster during certain parts of the year and slower at others. The sun therefore varies correspondingly his apparent speed along the Ecliptic, so that even from noon to noon the sun is not always precisely on time. He may be as much as fifteen minutes late or early, according to the season. And our modern days are measured according to the sun's average rate, so as to allow for this variation and keep every day exactly twenty-four hours long. This of course no sun-dial can possibly be made to do, since it must follow the actual sun. The sun-dial has remained in use to the present day. It seems strange to think of a sun-dial being used as a standard for setting clocks and actually to regulate the running of trains. But these things were done in civilized Europe within the last half century. It was only when the railroad and the telegraph had made standard time at once necessary and easy to obtain that the sun-dial altogether lost its position of authority.