Page images
PDF
EPUB

α. A thermometer inserted near the bottom of a tall cylindrical vessel filled with water and cooled in the middle part by some surrounding ice will after some time indicate a stationary temperature, which will be approximately that of the maximum density. (Hope's method.)

b. Joule and Playfair's experiment for determining the point of maximum density of water is also based on this second general method. They devised an apparatus and a mode of working which greatly improved on Hope's and gave means of attaining the utmost accuracy. The velocity of the convection currents set up in two tall cylinders communicating with each other by two channels, one at the top and one at the bottom, when filled with water and kept at different temperatures was measured. A glass float is put into the channel at the top, and the velocity of the current is measured by the motion of this float relative to a scale fixed at the back. A set of values for the velocity of the current was thus obtained, the mean temperature of the two vessels being different, and chosen so that in at least one case the current went in the opposite direction from that in the others. With these values a curve was constructed, from which the mean temperature at which the velocity of the current would be zero can be deduced. This is the temperature of maximum density. A description of the apparatus used, the mode of working and the results obtained, will be found in the paper by Joule and Playfair, Phil. Mag., series 3, vol. XXX. page 41.

The apparatus used in the Cavendish Laboratory for determining the point of maximum density of water by this method is constructed on the same principle as that originally used by Joule and Playfair, and has also the same dimensions. To guard against rapid changes of temperature from outside it is wrapped round with hemp. A small tube runs down along the side of each vessel, with which it communicates through some holes. Thermometers graduated to degree Centigrade and reading with accuracy to are suspended in these tubes by means of strings. The freezing points of these thermometers were re-determined, and a suitable correction for the zero reading applied to

Joule's original method of working is slightly modified, as instead of measuring the velocity of the current produced, the temperatures are so adjusted that no current is produced on making communication at the top and bottom between the two vessels. The water in the two vessels must then clearly have the same density, and the mean temperature of the two will give the point of maximum density. The assumption made here is, that the mean coefficient of expansion in the range of temperature taken above and below the point of maximum density is the same. To ascertain the presence and direction, or absence of a convection current the following plan of working is adopted. The water is well stirred, the slide at the top opened to allow the two levels to equalise themselves, then shut again and the water stirred again. An exact reading of the temperature in each vessel is then taken, the stopcock in the lower channel opened, the slide at the top carefully removed, the float put into the middle of the top channel and its motion watched.

The experiment was carried on at the ordinary temperature of the laboratory (about 16°C.) and the water cooled by putting ice into it till the temperature of the one was somewhat above, that of the other below 4°C. Care had to be taken not to make any observations before all the ice was melted, as otherwise on opening the slide at the top local currents would have been set up there, which might altogether have hidden the effect of those looked for. For the same reason the water must always be kept well stirred. A movement of the glass float at the top on opening the stopcock and slide indicated by its direction in which of the two vessels the water is denser, the direction of the convection current being from denser to lighter in the lower channel, from lighter to denser in the upper channel. By either increasing or decreasing the density of the water in one side, two temperatures can be obtained at which the density is the same. The change in density is effected by either using some ice or some warm water. According as to whether the water is above or below its temperature of maximum density the change in density produced by cooling or warming is reversed.

The best way of working in carrying out the experiment is to lower the temperature of the water in one vessel to near the freezing point, and then to adjust that in the other to a temperature such that no current is produced. The temperature of the colder water is then allowed to rise slowly by taking in heat from the surrounding air, and that of the other is cooled by ice. Another value for two temperatures, at which the density is the same, is thus obtained, these two last being nearer together; the process is repeated and a set of values nearer and nearer to each other is got. The mean temperature of the two vessels is taken in each case as that of the maximum density.

Some of the results obtained by this method are:

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

Temperature of maximum density (mean of 3) = 4.30.

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

Temperature of maximum density (mean of 3) = 4.25.

This value is somewhat too high, and the error can at least in part be accounted for by the following two circumstances:

(1) The water used was not pure. There was a great deal of iron-rust in suspension, and it probably contained also some impurities in solution.

(2) The sensitiveness was not sufficiently great, as was shewn by a separate experiment undertaken for the purpose of determining the degree of sensitiveness. This consisted in ascertaining the least difference in temperature which would give a current noticeable by the motion of the bead. The result obtained was that at 10.8° a difference of temperature of 0.2° between the two vessels produced a decided current. This corresponds to a difference of density which at 4° is only attained by a difference of temperature of about 04°C. The limit of error is therefore 0·4 ̊C., a value much larger than what it ought to be.

IDA FREUND.

No. 8. ON THE MANUFACTURE OF AN AIR THERMOMETER FOR RAPIDLY INDICATING COMPARATIVELY HIGH TEMPERATURES1.

A LONG, uniform glass tube has a bulb blown near to one extremity, both ends being left open, and is bent round at right angles to itself at a distance of about two or three inches from the commencement of the bulb.

It is first carefully cleaned, by connecting the long end with the aspirating pump, and drawing consecutively acids and potash

[blocks in formation]

into the bulb, and then allowing them to run down the tube, afterwards well washing with spring water and distilled water.

When this operation is complete, the free end is connected with a series of calcium chloride tubes, and air is passed through,

1 Compare Berthelot's instrument.

« PreviousContinue »