while the bulb and the tube are warmed with a spirit lamp to facilitate the drying.

As soon as it is completely dry, the drying tubes are removed, and a small glass tap is connected with the open end of the bulb by a short piece of india-rubber tubing, which has been previously well washed and dried.

The lower end of the tap is introduced into a small beaker of pure, dry mercury (at a known temperature), and the pump being again put in action, the small tap is slightly turned, and the mercury allowed to ascend slowly into the bulb. As soon as it reaches the neck of the bulb the tap is shut off, and the thermometer disconnected from the pump by slipping off the connecting tube.

The beaker is now removed, and a second beaker, whose weight has been previously determined, is placed under the tap, which is turned on, and the mercury allowed to sink through the whole length of the bulb, being stopped at exactly the point at which it is intended to seal off the bulb. The quantity of mercury that has fallen through is weighed, and thus we determine the weight of mercury at the temperature of the experiment, which is required to fill the bulb, and hence deduce the volume of the bulb.

To find the volume of any length of the tube, the bulb end of the instrument is connected with the air-pump, and the tap fastened on to the other end. When a sufficient length of mercury has run in the tap is turned off. After the length

of the column introduced has been measured off, by placing the stem of the thermometer over a scale (the reading of which may be afterwards corrected with the beam-compass), it is dropped into a beaker and weighed as before.

The cleaning and drying processes have now to be repeated -with the utmost care1. The instrument should first be well washed with nitric acid, to remove any small particles of mercury which may have adhered to the inside.

In drying, the air is caused to pass first through calcium chloride, and then through phosphoric pentoxide tubes.

This

1 The facility of motion of mercury in a narrow glass tube depends very greatly upon the glass being perfectly clean and dry.

operation must be allowed to go on for a considerable time, and meanwhile the bulb and tube be thoroughly heated.

Two little stoppers, each consisting of a short piece of indiarubber tubing, having one end closed with a small glass rod, are got ready, and as soon as the thermometer is perfectly dry, it is disconnected from the pump and drying tubes, and a stopper is put on to each end.

The end of the bulb is now sealed up. The instrument is mounted vertically on a stand, with a scale alongside the stem

T

U

T

Fig. 5.

(for convenience in determining the boiling point, and the melting point of ice, the arm holding the bulb is bent down to an angle of about 45° with the stem). A small beaker of pure mercury is put underneath, and the stopper withdrawn from the end of the tube, beneath the surface of the mercury.

The bulb is heated in a large Bunsen flame till it begins to get red-on allowing to cool the mercury rises in the stem. The instrument is thus filled with air which is under a pressure nearly equal to that of the atmosphere at the highest temperature for which it can be used, but at ordinary temperatures the pressure is of course smaller. Before graduating the instrument, a diamond scratch is made upon the stem as a reference mark.

The temperatures 0° and 100° C. are obtained by putting the bulb in a funnel of melting ice and in a steam jacket.

The height of the barometer and the temperature of the room are observed at the time of the determination.

Fig. 6.

A tube is used to allow the steam to escape-it should issue very gently.

Observations for graduation:

Wt. of mercury which fills the bulb at 15° = 113·460,

Vol. of 113:46 grms. Hg at 0° = 8·345 c. c.,

.. at 15° C. = 8.368 c. c.

Length of column introduced into tube 28.15,

Weight of this 5.8055,

=

=

Temperatures up to the softening point of glass (about 500° C.) may be conveniently determined with very considerable accuracy by the method employed by Rudberg and Regnault to determine the coefficient of expansion of air. The method in the form it naturally takes for a class experiment consists in exposing to the source of heat a glass bulb with a neck drawn out to a fine point. The extremity of the neck must project from the hot enclosure, the rest of the instrument can be brought to the temperature it is required to measure. When this is done the extremity of the neck is sealed by a blowpipe flame and the bulb is withdrawn, allowed to cool, and weighed, let the weight be w。. The end of the neck is then broken off under water or mercury, and the liquid allowed to enter until the pressure in the interior is equal to the atmospheric pressure. The nozzle is closed with wax while the bulb is being withdrawn from the liquid, and the bulb with the mercury or water which it contains is weighed again, taking care that the glass broken off is collected and added, let this weight be w1. The bulb is then completely filled with the liquid and re-weighed— let this third weight be w2.

Then, supposing that the atmospheric pressure at the time of sealing off is the same as when the nozzle is broken off under the liquid, it is evident that a volume of air corresponding to a weight w, w, of the liquid has by expansion at constant pressure, namely, the atmospheric pressure, filled the bulb at the high temperature.

Its volume (not allowing for the expansion of the glass) would correspond to the weight w2- w, of liquid, we therefore have, if be the unknown high temperature, and t the temperature of the air and liquid at the time of the experiment, y the coefficient of expansion of glass

when a is the coefficient of expansion of air at constant pressure. This plan of measuring temperature works very well in practice. It is of course necessary that the bulb should be very thoroughly dried before being used. The precautions described by Regnault (Mém. de l'Inst. T. XXI.) are in general unnecessary unless extreme accuracy is required. Some results obtained by various observers with this method are appended.

1 The glass softened in the flame, and the bulb collapsed somewhat after sealing. 2 July, 1886.