There are many forms of this escapement and there is no need to explain it in detail. But the main idea is this: At the end of each vibration or swing of the pendulum, the escape-teeth, instead of being made to recoil by the downward motion of the pallets, simply remains stationary or at rest until the commencement of the return swing of the pendulum. This was brought about by applying certain curves to the acting faces of the pallets. But the acting faces of both tooth and pallet are beveled, so that the tooth in slipping by gives the pallet a "kick" or impulse outward and keeps it in motion. Nowadays, even a common alarm-clock has an escapement working in this way. Then came another remarkably interesting contribution. Have you ever wondered why the pendulums of fine clocks were weighted with a _gridiron_ of alternate rods of brass and steel? For purpose of ornament? Not at all—it constitutes a scientific solution of an embarrasing problem, due to the inevitable variations in temperature. Metals expand with heat and contract with cold. Notched iron bars can be made to "crawl" along a flat surface by alternately heating and cooling them. Bridge-builders sometimes arrange sliding points, or rocking points to adjust the differences in the length of the steel. Contraction and expansion are important factors in all their calculations. But a pendulum would change its rate of motion if it changed its length and this would interfere with its accuracy as a measurer of time. Graham worked upon this problem, too, and attached a jar of mercury to the rod of his pendulum for a weight. When the heat lengthened the rod, it also caused the mercury to rise, just as in a thermometer, and this left the "working-length" the same. Such mercury-weighted pendulums are not uncommon to this day, but the more familiar _gridiron_ came from the brain of John Harrison, who, in 1726, fixed the alternate rods in such a way that the expanding brass rods raised the weight as much as the expanding steel rods lowered it. Thus they neutralized each other. The clock as we know it was now virtually complete. There were structural refinements, but no more radical improvements to be made. In tracing its development from the fourteenth to the eighteenth century, we note one curious likeness to the ancient history of recorded time. In this case, as before in Babylon, the people first concerned with the science were the priests, and after them the astronomers, but we note a still more important difference.