The authors entertained at one time the idea that the elastic viscous recovery might only be found to occur with substances which were mixtures; one constituent of which might be elastic and the other viscous. With the object of testing the matter some cylinders of sodium stearate were prepared. It was found that elastic viscous recovery occurred in this substance similar in every respect to that observed with such mixtures as glass or pitch. The viscosity of the substance is given in the general table.

Shoemaker's Wax.

It was desirable that determinations of the viscosity of the same substance should be made by the method here described and also some other. It was found that shoemaker's wax was, on the one hand, just sufficiently viscous to allow cylinders to be made from it for determination with the torsion method, and just sufficiently fluid to admit of its viscosity being determined by allowing a spherical body to drop through it.

The mean value obtained by the torsional method working with two different-sized cylinders was μ=47 x 106. This value is open to considerable doubt, for cylinders of shoemaker's wax sag in the centre rather too quickly to give really reliable results, and would have to be supported by a fluid of the same density in the manner explained earlier in the paper.

The Stokes method adopted for comparison gave very variable results. A steel bicycle-ball answered as the spherical body, the measurements of which showed it to be wonderfully true; nevertheless it did not fall vertically, but irregularly from side to side in its descent. This may have been due to the ball rotating owing to lack of uniformity. The wax itself should have been fairly homogeneous, for it bad been poured when liquid into the containing cardboard cylinder.

The position of the sphere was found from time to time by means of the X-rays. It took a fortnight to travel 1.8 cm. The value for the coefficient of viscosity obtained varied from 6 × 106 to 23 x 106, the mean value being about 10 x 106. This is of the same order of magnitude as that obtained by the torsion method. This latter is probably too small because the sagging of the rod in the torsion experiment was so great that the torque could not be applied long enough to reach the "steady "state.

Observations were made with paraffin-wax and modellingclay to ascertain the character of their behaviour. Paraffin

wax exhibited a behaviour in every way similar to the substances already mentioned, but the modelling-clay acted quite differently. When subjected to a given torque it moved slowly up to a given position and stopped there. On removing the stress, it made an immediate partial recovery to a point where it permanently remained.

The following list contains the results obtained with the several substances experimented with.

XLII. On the Emanation given off by Radium. By J. A. MCCLELLAND, M.A., Professor of Experimental Physics, University College, Dublin*.

THE

HE a rays of radium have been proved to consist of positively charged particles moving with great velocity, the mass of the particle being comparable with that of the hydrogen atom. The B rays have also been shown to consist Brays of charged particles moving with great velocity, the charge in this case being negative, and the mass of the particles very small compared with that of even the hydrogen atom.

Little is known as yet about the y rays, except that they have very great penetrating power.

The emanation produced by radium has been much studied, and many of its properties are known; but it does not appear to have been definitely settled whether or not the emanation particles are charged; and it is important to be certain on this point when framing a conception of the manner in which the radium atom disintegrates. The object in this paper is to test as accurately as possible whether or not the emanation carries an electric charge. Rutherford's work indicates that it is not charged; but I have thought it advisable to make a

From an advance proof of the Transactions of the Royal Dublin Society,' n. s., vol. viii. part vi. pp. 89-94, communicated by the Author.

direct test of the matter, as Rutherford's work is not conclusive on this point.

Rutherford has had emanation for long periods in closed vessels, and under the action of an electric field, in which case we should expect the emanation, if charged, to be driven to one or other of the terminals; and as this does not happen, the indication is that it is not charged. But if the mass travelling with the electric charge were great in comparison with the charge, the motion under electric force would be very slow, and the emanation would not move to the terminals.

Description of Apparatus.

Five milligrammes of radium bromide were dissolved in water contained in a small vessel R covered with a slip of thin paper through which the emanation readily passes. The vessel R is placed under a large air-tight bell-jar A. A second large air-tight bell-jar B is joined up as shown in the figure. C is a vessel filled with glass-wool; and D is a metal

cylindrical vessel resting on blocks of paraffin, and fitted with a paraffin stopper, in which is fixed the metal rod T. The glass tube F dips into mercury, and acts as a gauge to show the pressure when D is partially exhausted, the exhaustion being produced by applying a pump at G. The letters t1, t2, and to denote taps by which the tubes can be closed at the points indicated. E is a quadrant electrometer. one pair of quadrants being permanently to earth, and the other pair joined to a mercury cup b in a block of paraffin. The cup b is kept connected to an earthed cup a, except when an observation is to be taken, and then the connecting piece

is removed by a string from a distance, so as not to disturb the electrometer by induction effects. The mercury cups b and e in the same block of paraffin are joined to D and T; while a fourth cup d is joined to one pole of a battery of small storage-cells, the other pole of which is to earth. The vessel D is screened from outside electrical disturbances by a surrounding earthed conductor not shown in the diagram.

Method of Working.

The radium emanation passes readily through the slip of thin paper covering the vessel R; and thus, after A bas remained closed for a short time, it contains a large quantity of emanation. The tap to being closed, the vessels B, C, and D are partially exhausted to any desired pressure; 3 is then closed, and t1 and to opened; and thus the vessel B is filled with air containing radium emanation, t1 and to being then closed again. The mercury cups b and c being joined together, the piece connecting a to b is removed, and the emanation allowed to rush into D by opening the tap t3. If this emanation carries a charge, it will be shown by a deflexion of the electrometer spot of light. The glass-wool in the vessel C stops dust particles which might get electrified by friction and produce a deflexion. The glass-wool also stops the ions which have been produced by the radiation from the radium emanation. As the ionized gas has been for some time in B and the tube leading to C, there would be a tendency for more negative than positive ions to be lost by diffusion to the walls; and the excess of positive would produce a deflexion when admitted into the vessel D.

To test whether or not the air thus admitted into D has carried emanation with it, and how much, the ionization current between T and D is measured immediately after the gas is admitted to D. To do this e is disconnected from b and joined to d. The terminal T is thus kept at a high potential, and the air in D being kept ionized by radiation from the emanation, the vessel D will gradually be charged; and the rate of charging is measured by the rate of movement of the spot of light when the connexion between a and b is broken.

Before the emanation is admitted there is only a very small current to D, when T is connected to the storage-battery, this small current being due to the weak ionization which is always present in atmospheric air.

We thus, by one experiment, measure the charge (if any) carried by the emanation, and by a second experiment we measure the ionizing power of this emanation.

The Observations.

We shall now give the numbers obtained in an experiment similar to a great number of others carried out.

The capacity of the electrometer and the necessary connexions, including the vessel D, was 131 electrostatic units, or 000145 microfarad, and the electrometer gave a deflexion of 60 scale-divisions for 1 volt difference of potential between its quadrants.

The admission of the radiam emanation produced a deflexion of only 4 scale-divisions.

The ionization current to D was then measured as described above immediately after the admission of the emanation. To sufficiently reduce the rate of movement of the spot of light a capacity of 1 microfarad was joined to the electrometer, and the deflexion was then 100 scale-divisions in 47 seconds.

The admission of the air containing emanation into the vessel D produced, as stated, a small deflexion of 4 scaledivisions. Preliminary observations had been made to see if any deflexion would be produced when an equal quantity of air free from emanation was admitted in the same way. It was found that a small deflexion was produced probably by some friction effect, the deflexion varying in different experiments between 0 and 5 scale-divisions, and being always in the same direction. The direction of the deflexion of 4 divisions obtained when the air contained emanation was the same as that obtained without emanation.

Judging not only from this particular experiment, but from several others, we are safe in saying that the emanation did not carry a charge sufficient to produce a deflexion of more than 1 scale-division.

The experiment did not, therefore, detect any charge on the emanation; but it is important to calculate whether or not the emanation might be charged, and the charge be less than what could have been detected in the above experiment.

Let us suppose that each emanation particle has a charge equal to that carried by the gaseous ion; we have no case of a charge less than this, so that if the emanation is charged its charge is probably at least equal to that of the gaseous ion, and may be greater. Denote this charge by e in electromagnetic units.

The capacity of the electrometer and connexions was 000145 microfarad, and 1 scale-division corresponds to a potential-difference between the quadrants of of a volt. A deflexion of 1 scale-division would therefore be produced