of the room on starting the experiment, the correction for loss by radiation, &c., may be made as follows:

Let the rate of falling of temperature at the initial and final temperatures be observed. Take the mean of these rates, and multiply this by the time the experiment has lasted This product must be used instead of above.

Two observers are required, one to turn the handwheel, and the other to note the revolutions of the cogwheel and the temperature.

Make a note of the time of the beginning of the experiment, and also the time at which each successive 100 turns of the spindle are completed. This will be a great check on accuracy of counting.

Stir the water all the time, by moving the stirrer gently up and down. Do not splash. Place two or three (not six or seven) drops of oil on the inside of the outer vessel, and place the inner vessel in it before the oil has run down to the bottom of the vessel.

Hang a sensitive thermometer from a clip, so as to pass through the hole in the centre of the wooden wheel, and so as to have its bulb not quite touching the bottom of the vessel.

Place weights symmetrically on the wooden wheel so as to produce enough friction to raise the weight P when the wheel is worked at a convenient speed.

On starting, the cones slip with much greater difficulty than when once started.

The following plan is convenient :-Fasten a string to P, and attach the other end to a weight Q, which rests on the floor. On starting, P will not be sufficient to keep the inner cup from revolving, and Q will come into play; as soon as the statical friction has been overcome Q will fall to the ground again, and the driving-wheel must then be so manipulated that the string P Q is always slack. Great care must be taken that the string supporting P is always a tangent to the wheel.

The mass of P should be about 200 grammes.

CHAPTER XI.

PRESSURE OF VAPOUR AND HYGROMETRY.

41. Dalton's Experiment on the Pressure of Mixed Gases. To shew that the Maximum Pressure produced by a Vapour in a given Space depends on the Temperature and not on the Presence of Air or other Vapours in that Space.

FIG. 21.

-10

E

B

The apparatus and experiment are de- . scribed in Garnett's 'Heat.'

A, B, G, fig. 21, are three barometer tubes. A and B are to be filled with mercury and inverted over the cistern of mercury D E. G contains some air above the mercury.

We require, first, to explain how to fill the tubes with mercury.

They must first be cleaned by washing out with dilute acid, and then dried by being repeatedly exhausted with the air-pump and filled with air that has passed through chloride of calcium tubes. This can be done by means of a three-way cock, as already described (§ 16). Having cleaned and dried a tube, we may proceed to fill it.

For this purpose it is connected with a double-necked receiver which contains enough mercury to fill the tube, the other neck of the receiver being connected with the air-pump, and the tube and receiver are exhausted by working the air-pump. Then by raising the end of the tube to which the receiver is attached and tilting the receiver the mercury is allowed to flow into the empty tube from the receiver. We are thus able to fill the tube with mercury free from air without its being necessary to boil the mercury.

The three tubes should be filled in this way and inverted

over the mercury cistern. A convenient arrangement for the latter is a hemispherical iron basin screwed on to the end of a piece of iron tubing, the lower end of the tubing being closed.

Connect the open end of G by means of a bent piece of small-sized glass tubing with the drying tubes, and allow a small quantity of dry air to flow in. The amount of air introduced should be such as to cause the mercury in G to rise to about half the height that it reaches in A and B. The quantity can be regulated by pinching the india-rubber tube which connects G with the drying tubes.

Adjust in a vertical position behind the three tubes a scale of millimetres, and hang up close to them a thermometer. Place a telescope at some distance off, so as to read on the millimetre scale the height at which the mercury columns stand and also the thermometer. The tube G should be so placed that it can be depressed into the iron tubing below the cistern.

Mark the height at which the mercury stands in G by means of a piece of gummed paper fastened to the tube.

Read on the millimetre scale the heights of A, B, and G, above the level of the mercury in the cistern.

Introduce, by the aid of a pipette with a bent nozzle, a little ether into B and G, putting into each tube just so much that a small quantity of the liquid rests above the mercury.

The mercury in B will fall. The amount of fall will depend on the temperature. Let us suppose that the new reading in B is 354 mm., then the mercury has fallen through 765-354 mm. ; thus the ether exerts a pressure equivalent to that of 411 mm. of mercury.

The mercury in G will fall also, but not by so much as that in B, for the pressure in G is the pressure of the ether

vapour together with that of the contained air; and as the mercury falls, the volume of the contained air increases and its pressure consequently decreases.1

Now lower the tube G in the cistern until the level of the mercury in G just comes back again to the paper mark. The volume of the contained air is now the same as before, therefore so also is its pressure. The depression of the mercury column in G below its original height is due therefore to the pressure of the ether vapour. Now read the height of G on the scale; it will be found to be about 113 mm. The column in G, therefore, has been depressed through 524-113 mm., or 411 mm. Thus B and G are depressed through equal amounts provided that the volume of air in G is allowed to remain the same.

The assumption has been made that the temperature remains constant during the experiment. This will not be far from the truth in the laboratory, provided that the readings are taken from a distance so as to avoid the heating effects of the body; if necessary, a correction must be applied for a change in temperature.

Having made these measurements, depress B into the iron tube; it will be found that the consequence is simply to increase the amount of condensed liquid above the sur. face of B without altering the height of that surface.

The difference between the heights of the colurans in a and B gives in millimetres of mercury the maximum pressure which can be exerted by ether vapour at the temperature of the laboratory.

Experiment.-Determine the maximum pressure exerted by the vapour of ether at the temperature of the laboratory, and shew that it is independent of the presence of air.

Enter results thus :

Height of mercury in A

765 mm.

The presence of the air in G retards the evaporation of the ether; a considerable time must therefore be allowed for the mercury to arrive at its final level.

The volumenometer described in § 26 will afford us another means of testing Dalton's law. Introduce a small quantity of water or other liquid into the bulb E (fig. 16), and screw it on. As the water evaporates the pressure will increase and the level of the mercury change. When it has become steady read the level in both tubes, and note the height of the barometer. Alter the position of the tube A and take another reading, and thus obtain a series of corresponding values of volume and pressure. Let us suppose the volume of the flask is known, so that v, the actual volume occupied by the air, can be found. Allowance must be made for the volume occupied by the water, which of course changes slightly; this is easily done by weighing the flask empty, then with the water, at the beginning and end of the experiment. These last two will differ, but very slightly, owing to the evaporation. From the mean of the two weights and the weight of the empty flask we can obtain the average volume of the water, which will be sufficient for our present purpose.

Write down the reciprocals of the observed values of v, and then plot a curve with these reciprocals as abscissæ and the observed pressures as ordinates. If Dalton's law is true, or has the same actual error at all pressures, the curve will be found to be a straight line, as A B in fig. xxiv, cutting the axis of y in B. Let P M be any ordinate, and through o draw o Q parallel to A B, Cutting PM in Q. Let O B=Po, then