THE CONSTANT VOLUME AIR THERMOMETER.

BY A. O. WILKINSON, DETROIT, MICH.

I claim nothing new in what I have to present. It is merely a method that I have made use of in the laboratory to show the constant relation between the temperature of a gas and its pressure. The apparatus (Fig. 5) consists of a glass flask containing dry air, made dry by a little sulphuric acid `placed in the flask. This is inclosed in a tin can filled so as to cover the flask. Leading out of the flask is a glass tube bent twice at right angles,

Fig. 5.

which is connected to another glass tube having a rubber tube and a pinch cock at the bottom. These tubes can be placed in front of a mirror scale. Mercury is placed in the longer arm; when the mercury is at the same height in both arms, the air, of course, is under the pressure indicated by the barometer; its temperature is that of the water.

To obtain other readings heat is applied to the can and enough mercury is poured into the longer tube to bring the mercury in the short arm to its original position, when the volume of the inclosed air is the same as the original volume. The difference in the heights of the mercury in the arms measures the increase of pressure. These results are tabulated as pressure and absolute temperature. The temperature is then divided by the pressure to show the constant relation, i. e., the temperature measured on the absolute scale varies directly as the pressure.

COEFFICIENT OF EXPANSION OF A GAS AT CONSTANT

PRESSURE.

BY N. H. WILLIAMS, DETROIT, MICH.

Among the physical constants that have been made the subjects of laboratory experiments, none is more important than the coefficient of expansion of a gas. Its relation to the absolute zero and its immediate application in the volumetric work of the laboratory make it of peculiar interest and importance to both the physicist and the chemist. Elaborate experiments have made known to us the value of this constant with considerable accuracy, but a simple apparatus that will give reasonably accurate results in the hands of students has not till the present time been devised. Experiments for finding the pressure coefficient at constant volume, however, have been successful.

Nearly a dozen different pieces of apparatus have been used by the author of this article, but with little success. Mercury as a valve to enclose the air is unsatisfactory. The gas, if in contact with water, expands very irregularly, and even if corrections are made for aqueous tension, no concordant results are obtained. Methods involving only measurements by weighing have been tried; in these cases glycerine being used in contact with the gas, but still the presence of water vapor introduced a large error. Some active drying agent as a valve to enclose the air is indispensible, and sulphuric acid seems to be the only liquid available. Its vapor tension is insignificant. Its density is much less than that of mercury, and hence the pressure can be more accurately adjusted. One method in which sulphuric acid was used is of interest because it shows how important it is to exclude every trace of moisture. A long U tube an eighth of an inch in diameter was employed. A little sulphuric acid was put into it to separate the air in the two sides. Into one arm there was fitted a piston of paraffined wood. A little mercury over the piston made it perfectly air tight. As the gas expanded or contracted with changes of temperature, the piston was moved so as to keep the surfaces of the acid in the two arms at the same level. The measurements were made by putting the tube into a deep jar of hot water and observing the temperature of the water and the position of the piston after the adjustment for pressure had been made. The apparatus was then put into cold water and the piston pushed down as the gas contracted. The temperature was again observed and the length of the air column measured. These data are sufficient for calculating the coefficient of expansion. The results obtained in this way were good, but if the apparatus is put into cold water first, it will be cooled below the dew point, and traces of moisture will be deposited in

side the tube above the piston. As the piston is afterward raised in the second part of the operation, sufficient moisture will slip by it to cause an error of ten per cent. in the result.

This difficulty is eliminated by the apparatus described below. A glass tube a little over a meter long is bent into the form shown in figure 6. The shorter arm is closed by a short glass rod fused into the end of the tube. Some sulphuric acid is put in and another tube of smaller size is put into the open end and pushed down into the acid. The displacement of the acid by this tube as it is pushed downward raises the surface. This method permits an adjustment of the level through about eight inches, thus it is always possible to bring the surfaces of the liquid in the two arms to the same level and produce atmospheric pressure upon the enclosed gas. A metal band CC is arranged to slide upon the tubes. It is fastened to a wire, R, so that it may be adjusted to mark the liquid surface. The apparatus is put into cold water in a deep cylinder of glass and the level of the liquid in the long arm is adjusted to that in the short arm by moving the inner, tube T up or down. The metal band is then brought to this level and the temperature of the water noted. Next the apparatus is placed before the mirrored scale of a Jolly balance and the length of the air column accurately measured. The operation is then repeated with water at a temperature thirty or forty degrees higher.

2

2

If V, and V, represent the two volumes and t, and to the corresponding temperatures, we shall have the two equations V,V,(1+at,) and V1 =V (1+at), in which V, is the volume of the gas at zero degrees C, and "a" the coefficient of expansion. Eliminating V, between these two equations and solving for "a," we have

Fig. 6.

The following results will give an idea of what may be expected of the apparatus. None of them show more than one per cent. of error. .00002 is added to each result as found by the formula to correct for the expansion of the glass: .00366, .00365, .00363, .00368, .00366, .00363, .00364, .00366, .00368, .00366, .00366.

The following results are obtained by one of the laboratory sections. The error in two cases is greater than one per cent. .00365, .00366, .00371, .00370, .00366, .00373, .00353, .00368.

A MODIFICATION OF HARE'S METHOD FOR DENSITIES OF

LIQUIDS.

BY DE FORREST ROSS, YPSILANTI, MICH.

Fig. 7 represents an apparatus for determining the specific gravity of liquids by the Hare's method. It consists of a piece of board H about 100 cm. long by 15 cm. wide, rigidly fastened to a base K about 6 cm. wide, which, by being clamped to a table, supports the piece in a vertical position. 1, 2, 3, 4, 5 are test tubes 2 cm. in diameter by 10 cm. long, which contain the liquids to be tested; 3 contains water with which the other liquids are to be compared. R is a meter-stick for measuring the heights of the liquids in the various tubes by means of a carpenter's try-square.

A is a bottle with the bottom cut off and fitted with a rubber stopper perforated to receive the tubes a, b, c, d, e. The top is closed with the rubber stopper S, perforated to receive the tube s, to which is attached the rubber bulb B, by means of which a part of the air in A is removed, thus causing the liquids to rise in the tubes. The pinch-cock O prevents the return of the air.

The advantages of this apparatus over that of two tubes connected by a "Y" tube are several. The thin glass test tubes permit of more accurate reading of the surfaces of the liquids. The same liquids are always in the same tubes. Several liquids may be compared at the same time with great accuracy, as the tongue of the try-square will come in contact with the tubes, thus avoiding the error of parallax and always being at right angles to the

tubes. The hand-pressure on the bulb B is a much more agreeable way, to say the least, than the mouth for removing the air from the tubes. The piece is cheap, easily made and always ready.

A MODIFIED AMPÈRE APPARATUS.

BY H. N. CHUTE, ANN ARBOR, MICH.

Supported on a stout steel wire, within a vertical rectangular helix of wire, is a series of three electrically independent rectangular helices rigidly fastened so as to turn together (Fig. 8) with their naked ends dipping into a circular trough of mercury divided into two opposite sectors of about 60° arc electrically connected to the binding posts on the base. By means of a

Fig. 8.

sliding commutator the direction of the current in these movable helices can be reversed, and by means of a switch the stationary rectangular helix can be placed in or out of circuit at pleasure. The revolving helices consist, each, of ten turns of No. 24 copper wire, and are carefully balanced on the supporting point. When properly arranged only two of them can be in circuit at any one time. A current of four or five ampéres will produce continuous rotation of the hexagonal helix in the earth's field with the stationary helix out of circuit. A rapid rotation is secured by holding the pole of a bar magnet near it. With the stationary helix in circuit the attraction or repulsion of parallel currents, according to the direction given the current through the movable helices, causes it to rotate very rapidly.