In this same epoch apparatus of precision developed in other fields. Weight clocks had been in use as time-measurers since the thirteenth century, but they were, as we have seen, difficult to control and otherwise unreliable. Even in the seventeenth century scientists in their experiments preferred some form of water-clock. In 1636 Galileo, in a letter, mentioned the feasibility of constructing a pendulum clock, and in 1641 he dictated a description of the projected apparatus to his son Vincenzo and to his disciple Viviani. He himself was then blind, and he died the following year. His instructions were never carried into effect. However, in 1657 Christian Huygens applied the pendulum to weight clocks of the old stamp. In 1674 he gave directions for the manufacture of a watch, the movement of which was driven by a spring. Galileo, to whom the advance in exact science is so largely indebted, must also be credited with the first apparatus for the measurement of temperatures. This was invented before 1603 and consisted of a glass bulb with a long stem of the thickness of a straw. The bulb was first heated and the stem placed in water. The point at which the water, which rose in the tube, might stand was an indication of the temperature. In 1631 Jean Rey just inverted this contrivance, filling the bulb with water. Of course these thermoscopes would register the effect of varying pressures as well as temperatures, and they soon made way for the thermometer and the barometer. Before 1641 a true thermometer was constructed by sealing the top of the tube after driving out the air by heat. Spirits of wine were used in place of water. Mercury was not employed till 1670. Descartes and Galileo had brought under criticism the ancient idea that nature abhors a vacuum. They knew that the _horror vacui_ was not sufficient to raise water in a pump more than about thirty-three feet. They had also known that air has weight, a fact which soon served to explain the so-called force of suction. Galileo's associate Torricelli reasoned that if the pressure of the air was sufficient to support a column of water thirty-three feet in height, it would support a column of mercury of equal weight. Accordingly in 1643 he made the experiment of filling with mercury a glass tube four feet long closed at the upper end, and then opening the lower end in a basin of mercury. The mercury in the tube sank until its level was about thirty inches above that of the mercury in the basin, leaving a vacuum in the upper part of the tube. As the specific gravity of mercury is 13, Torricelli knew that his supposition had been correct and that the column of mercury in the tube and the column of water in the pump were owing to the pressure or weight of the air.