and allowed to cool. The mercury in this position closes the upper end of the tube, and as the cooling proceeds it is sucked through the capillary opening and falls, drop by drop, into the bulb below. The process is repeated, if necessary, until the lower bulb is filled with mercury. The thermometer is then heated. The mercury expands, driving out all the air and filling both bulb and tube. The temporary bulb is now removed, and the open end of the tube is closed before the blowpipe. The thermometer is ready for calibration.
The bulb is buried in cracked ice, from which the water is allowed free drainage. When the mercury no longer contracts, a mark is made on the tube at a level with the mercury. This is the freezing-point of water, 0° on the centigrade scale, or 32° on the Fahrenheit. Réaumur's scale, with the freezing-point at 0° and the boiling-point at 80°, although so extensively used in Germany and Russia, is seldom seen in this country. The thermometer is then transferred to a bath of boiling water. The mercury quickly rises, and soon again becomes stationary. The tube is marked for the second time. This is the boiling-point of water, 100° centigrade, or 212° Fahrenheit. As the temperature of the boiling-point varies with the atmospheric pressure, the barometer must be read and a corresponding correction made, or else a standard thermometer must be kept in the bath, and the marking made in harmony with that. These two points determined, the operation of making a thermometer is almost completed. It has now only to be marked.
The tube is dipped into a bath of melted beeswax, and as soon as the thin layer of wax hardens it is taken to the dividing-engine. The space between the freezing and boiling points is here divided off into 100 divisions if the centigrade scale is to be employed, or into 180 divisions if the Fahrenheit be used. Every tenth line is made somewhat longer than the others, and is the only one marked. The marking is done on a machine constructed after the order of a pantograph. The waxed tube is laid on a small sliding platform and secured to its bed by a few drops of melted wax. A sharp stylus is then brought to bear upon the point where the marking is first wanted. The movement of the stylus is controlled by a long lever, whose own movement is, in turn, controlled by the action of a second stylus. This is made to pass over the desired figures cut in brass on a lower platform of the machine. The action of the system of levers is to reduce the motion of the upper stylus, and consequently the size of the figures traced through the wax. In this way accurate marking is secured on a sufficiently small scale. The tube, thus lined and marked, must now be subjected to the action of hydrofluoric acid. A solution of the acid in water, to which some alkaline salt has
