1.

LI. Low-velocity X-ray Electrons. By LEWIS SIMONS, D.Sc. Reader in Physics in the University of London *.

WE

HILST studying the absorption of X-ray electrons in various gases certain peculiarities in the absorption curves led me to conclude that groups of electrons, as predicted by O. W. Richardson †, possessing speeds less than that given by the equation

where is the maximum frequency of the incident radiation, played an important role in electronic emission ‡. Further work on this subject by the writer § and by de Broglie and Whiddington ¶ showed conclusively that the maximum speed of each sub-group could be represented by the equation

where vo represents the frequency of the radiation associated with the level within the atom from which the electron is ejected, the value of which is determined from a knowledge of the X-ray spectrum of the element emitting the electrons. The validity of this equation has been confirmed in the region of y-ray frequencies by Ellis ** It marks the position of the head of each B-ray spectral line. It would be very difficult to determine the width of these lines or bands, whilst equation (ii) indicates that there should be no lower limit to the velocity of photo-electrons.

2. The experiments described were directed to the determination of the distribution with velocity of those electrons emerging with minimum energy from a screen of high atomic weight when irradiated with X-rays. They have shown that, in point of number, of the electrons described in paragraph 1 practically all the emission is confined to those of velocity less than 2 volts. I have therefore employed a retarding electrostatic field method, such as is described by Richardson and Brown†† in their study of the distribution of the normal

* Communicated by the Author.

+ Proc. Roy. Soc. A. xciv. p. 272 (1918).

Trans. Roy. Soc. S. Africa, viii. (1) p. 73 (1919).

Phil. Mag. xli. p. 121 (1921).

Comptes Rendus, clxxii. pp. 274, 527, 764, 806 (1921); Journal de Physique, vi. T. ii. p. 265 (1921).

Phil. Mag. xliii. p. 1116 (1922).

** Proc. Roy. Soc. A. xcix. p. 261 (1921), ci. p. 1 (1922).

†† Phil. Mag. xvi. p. 353 (1908).

component of the velocity of thermions emitted from a comparatively small surface of hot platinum towards an extensive plane opposite. In the corresponding X-ray experiment the electronic currents measured were necessarily much smaller, and, moreover, no variation of the opposing electrostatic field, caused, for example, by the charging of an electrometer, could be allowed during a run. A "null" method of employing a tilted electroscope was therefore devised and other precautions employed that are given below.

3. In fig. 1, A and B represent two sheets of filter-paper, each about 16 sq. cm. in area, insulated from each other and about 1 mm. apart. The side of A remote from B was made conducting by rubbing with pure graphite. The side towards B was coated with one sheet of ordinary gold-leaf. Fig. 1.

Both faces of B were rubbed with graphite; whilst this makes the surface a good conductor, I have never been able to detect any electronic emission from it. The gold-leaf of the screen A was connected directly to a Wilson's tilted electroscope by a short lead sheathed for 10 cm. of its length with an insulated metal tube 2 cm. in diameter, the potential of which could be varied continuously from 0 to -10 volts by turning a handle on the continuous rheostat D. The potential of B could be maintained steady at any desired value. It was not varied beyond the range + 2 volts.

4. The method of experimenting was as follows:-- After the gold-leaf system with A attached had been earthed and insulated, C earthed, and B at, say, 1 volt, the X-ray beam was started. The system A would normally begin to

acquire a positive potential due to loss of electrons from the gold. This change of potential was prevented by continuously diminishing the potential of C from zero downwards, so that the potential of the insulated gold-leaf system remained at zero during the run, as denoted by the stationary position of the needle of the electroscope on the eyepiece scale. After the one-minute run was over C was earthed, A at once acquired its normal positive potential due to loss of negative charge, the deflexion of the electroscope needle giving a measure of this charge after suitable calibration. Thus the arrangement might be called a "null" method of using an electroscope. It was extremely useful, for it quite got over the difficulty involved in measuring the rate of charging of A without the potentials of A or B

Arrangement of Lead Stops and Radiator A, within evacuated chamber.

varying. Its advantages are obvious in the accurate measurement of extremely low electronic speeds. Incidentally, it might also be noted that the final position of the jockey on D maintaining A at zero potential at the end of the run is also a measure of the charge acquired by A.

5. Great precaution was taken to avoid the spurious charging of A by electronic emission from any part of the apparatus other than the gold-leaf upon it. The X-ray beam was rendered parallel by means of three lead stops and limited to a cross-section area of 1 sq. cm. The last lead stops, the sheets A and B, and the general arrangements are shown in fig. 2, the whole system being enclosed in an exhausted brass cylinder lined inside with filter-paper.

All lead parts were given a thin film of paraffin-wax and earthed. F, F are sheets of graphited filter-paper, about 1 mm. from A and B respectively, placed there for the double purpose of limiting the volume in the neighbourhood of A and of screening it from stray electrostatic fields. That the whole arrangement was quite successful for the purpose for which it was intended was determined by replacing A by an exactly similar sheet, but without the gold-leaf upon it. On now making a run in a manner described above, even under the most favourable conditions with B charged positively, no readable deflexion could be obtained. The readings could therefore be plotted without any corrections or adjustments. To approach ideal conditions, the area radiating elettrons must be small in comparison with the whole area of the parallel plates used for testing their speeds. The effect of increasing the area of these plates is to diminish the currents measured. To obtain measurable currents the full radiation from a tungsten target "radiator " pattern Coolidge tube was employed, the tube being excited by means of a transformer.

6. Two characteristic curves are shown in fig. 3, the ordinates representing the charge acquired in 1 minute by the system A, maintained at zero potential, the abscissæ give the potential of B in volts. Curve I is for a current of 2 milliamperes through the tube, curve II for 3.5 milliamperes. The points on the two curves were taken alternately and after equal intervals of time, the heating current of the cathode of the Coolidge tube being maintained during the whole experiment and varied slightly for alternate points to give the above outputs. The smoothness of the two curves indicates the accuracy of the observational points.

7. Both curves show that 85 per cent. of the electrons emitted have velocities less than that given by a fall through 2 volts, and the argument against their being true dependent photo-electrons described in paragraph 1 is as follows:The results are complicated by the incident radiation being heterogeneous, but M. and L. de Broglie have shown from considerations of the absorption law of Bragg and Peirce that selective absorption of frequency v will occur when v=vo, where vo is a natural frequency existing in an absorbing element. Taking this to be the La, frequency for gold (w.l.1.27 x 10-8 cm.), the major velocity of the dependent electrons emitted in these experiments should be

*Comptes Rendus, Sept. 1921, p. 527.

given by

Ve=h(v-vo)=hv。/3 .

(iii)

..

V=3.35 x 103 volts.

The curves show that this cannot be the case. Considering now that the energy-distribution curve for the radiation from the Coolidge tube is continuous, it follows that residues. such as (iii) will also form a continuous series, but, as the

major number of electrons emitted have comparatively zero speed, if these were the true photo-electrons, then the emission spectrum from the Coolidge tube should correspond exactly to that from gold. Again, interesting results on these low-speed electrons have recently been obtained by Shearer, though the design and method of using of his

*Phil. Mag. xliv. p. 806 (1922).