Practical Physics is a free textbook on basic laboratory physics. See the editorial for more information....

Construction of a Water Thermometer

The method of filling a thermometer is given in full in Garnett's 'Heat', 10-18, also in Deschanel's(1) or Ganot's 'Natural Philosophy,' and Maxwell's ' Heat'.

In this case water is to be used instead of mercury.

One or two points may be noticed:

(1) The tube and bulb have not always a cup at the top as in Garnett (fig. 1). When this is the case, a piece of wide glass tubing must be drawn out to serve as a funnel, and joined by means of clean india-rubber to the tube of the thermometer.

(2) It would be difficult to seal the glass tube when full of water, unless it has been previously prepared for closing. After the bulb has been filled, but before it is again heated to the high temperature, the upper part of the tube is softened in a blow-pipe flame, and drawn out so as to leave a fine neck in the tube. Then the bulb is heated until the liquid rises above this neck, and when this is the case the tube is sealed by applying a small blow-pipe flame at the thinnest part.

At the moment of sealing the source of heat must be removed from the bulb, otherwise the liquid will continue to expand, owing to the rise of temperature, and will burst the bulb. The safest way of heating the bulb is to put it in a bath of liquid - melted paraffin, for example, or water if the thermometer be not required for use near the boiling point - and apply heat to the bath until the liquid in the thermometer reaches beyond the neck. Remove the source of heat from the bath and seal off the tube as the level of the water sinks past the narrow neck.

(3) The water used for filling the thermometer should be distilled water from which the dissolved air has been driven by long-continued boiling. This precaution is essential, as otherwise bubbles of air separate from the water in the bulb and stem after sealing, and this often renders the thermometer useless until it has been unsealed and the air removed and the tube re-sealed.

We proceed to show how to use the thermometer to determine the coefficient of expansion of the water.

We. require, for this purpose, to know the volume of any given length of the tube and the whole volume of water contained in the thermometer.

To find the Volume of any Length of the Tube.

Before filling the thermometer, introduce into the tube a small pellet of mercury and measure its length, which should be from 10 to 20 cm. Then warm the bulb and force the mercury out into a beaker, of which the weight is known. Weigh the beaker and mercury, and get by subtraction the weight of the mercury. Now, we may take the density of mercury to be 13.6. If, then, we divide the mass in grammes by this number, we get the volume in cubic centimetres.

We thus find the volume of a known length - that of the pellet of mercury - of the tube, and from this can determine the volume of any required length. For greater accuracy it is necessary to measure the length of the same pellet of mercury at different parts of the tube, thus calibrating the tube (see 8).

To find the Volume of the Water which is contained in the Thermometer.

Weigh the bulb and tube when empty, then weigh it again when filled, before sealing off. The difference in the weights gives the number of grammes of water in the bulb and tube, and hence the number of cubic centimetres of water in the two can be calculated.

It may be more convenient to seal off before weighing, but in this case great care must be taken not to lose any of the glass in the act of sealing, and to put the piece of glass which is drawn off on the balance with the tube.

If the thermometer be filled with some other liquid than water, we obtain the volume from the mass by dividing by the density of the liquid.

Let us suppose the volume of 1 cm. length of the tube is 0.01 c.c., and that the volume of the water contained is 4.487 c.c.

After sealing the tube as already described, immerse it in a bath of water at the temperature of the room, noting that temperature by means of a thermometer; suppose it to be 15C.

Make a mark on the tube at a known distance above the level of the water in it; let us say at 10 cm.

Now raise the temperature of the bath until the level of the water in the tube rises to this mark, and then note the temperature as indicated by the other thermometer. We shall find that with the numbers given it will be about 70C.

The water has risen 10 cm., and the volume of 1 cm. is 0.01 c.c. Thus the volume of water has been increased relatively to the glass by 0.1 c.c.

The original volume was 4.487 c.c. The new volume is 4.587 c.c. The rise of temperature is 70-15, or 55C.

Thus the coefficient of expansion of water relatively to the glass between these temperatures is 0.1/(4.487x55) per degree centigrade.

This, on reduction, comes to 0.000405.

The coefficient of expansion of water varies considerably with the temperature, so that the result will be the mean coefficient between the limits of temperature 15 and 70.

Experiment. - Determine by means of a water thermometer the coefficient of thermal expansion of water.

Enter results thus:

Length of pellet of mercury: 15.3 cm.
Weight of do.: 2.082 gm.
Vol of 1 cm. of tube: 0.01 c.c.
Vol. water initially: 4.487 c.c.  Temp. 15
Vol. finally: 4.587 c.c. Temp. 70
Coeff. of expansion = 0.000405 per 1.

(1) Deschanel, Natural Philosophy, p. 245, etc.

Last Update: 2011-03-27