Radio-activity/Chapter 4

by Ernest Rutherford

 

CHAPTER IV.

NATURE OF THE RADIATIONS.

PART I.

Comparison of the Radiations.

71. The Three Types of Radiation. All the radio-active substances possess in common the power of acting on a photographic plate and of ionizing the gas in their immediate neighbourhood. The intensity of the radiations may be compared by means of their photographic or electrical action; and, in the case of the strongly radio-active substances, by the power they possess of lighting up a phosphorescent screen. Such comparisons, however, do not throw any light on the question whether the radiations are of the same or of different kinds, for it is well known that such different types of radiations as the short waves of ultra-violet light, Röntgen and cathode rays, all possess the property of producing ions throughout the volume of a gas, lighting up a fluorescent screen, and acting on a photographic plate. Neither can the ordinary optical methods be employed to examine the radiations under consideration, as they show no trace of regular reflection, refraction, or polarization.

Two general methods can be used to distinguish the types of the radiations given out by the same body, and also to compare the radiations from the different active substances. These methods are as follows:

(1) By observing whether the rays are appreciably deflected in a magnetic field.

(2) By comparing the relative absorption of the rays by solids and gases.

Examined in these ways, it has been found that there are three different types of radiation emitted from radio-active bodies, which for brevity and convenience have been termed by the writer the α, β, and γ rays.

(i) The α rays are very readily absorbed by thin metal foil and by a few centimetres of air. They have been shown to consist of positively charged bodies projected with a velocity of about 1/10 the velocity of light. They are deflected by intense magnetic and electric fields, but the amount of deviation is minute in comparison with the deviation, under the same conditions, of the cathode rays produced in a vacuum tube.

(ii) The β rays are far more penetrating in character than the α rays, and consist of negatively charged bodies projected with velocities of the same order as the velocity of light. They are far more readily deflected than the α rays, and are in fact identical with the cathode rays produced in a vacuum tube.

(iii) The γ rays are extremely penetrating, and non-deviable by a magnetic field. Their true nature is not definitely settled, but they are analogous in most respects to very penetrating Röntgen rays.

The three best known radio-active substances, uranium, thorium, and radium, all give out these three types of rays, each in an amount approximately proportional to its relative activity measured by the α rays. Polonium stands alone in giving only the α or easily absorbed rays^[1].

72. Deflection of the rays. The rays emitted from the active bodies thus present a very close analogy with the rays which are produced in a highly exhausted vacuum tube when an electric discharge passes through it. The α rays correspond to the canal rays, discovered by Goldstein, which have been shown by Wien to consist of positively charged bodies projected with great velocity (see section 51). The β rays are the same as the cathode rays, while the γ rays resemble the Röntgen rays. In a vacuum tube, a large amount of electric energy is expended in producing the rays, but, in the radio-active bodies, the rays are emitted spontaneously, and at a rate uninfluenced by any chemical or physical agency. The α and β rays from the active bodies are projected with much greater velocity than the corresponding rays in a vacuum tube, while the γ rays are of much greater penetrating power than Röntgen rays.

The effect of a magnetic field on a pencil of rays from a radio-active substance giving out the three kinds of rays is very well illustrated in Fig. 22^[2].

Fig. 22.

Some radium is placed in the bottom of a narrow cylindrical lead vessel R. A narrow pencil of rays consisting of α, β, and γ rays escapes from the opening. If a strong uniform magnetic field is applied at right angles to the plane of the paper, and directed towards the paper, the three types of rays are separated from one another. The γ rays continue in a straight line without any deviation. The β rays are deflected to the right, describing circular orbits the radii of which vary within wide limits. If the photographic plate AC is placed under the radium vessel, the β rays produce a diffuse photographic impression on the right of the vessel R. The α rays are bent in the direction opposite to that of the β rays, and describe a portion of the arc of a circle of large radius, but they are rapidly absorbed after traversing a distance of a few centimetres from the vessel R. The amount of the deviation of the α rays compared with that of the β rays is much exaggerated in the figure.

73. Ionizing and penetrating power of the rays. Of the three kinds of rays, the α rays produce most of the ionization in the gas and the γ rays the least. With a thin layer of unscreened active material spread on the lower of two parallel plates 5 cms. apart, the amount of ionization due to the α, β, and γ rays is of the relative order 10,000, 100, and 1. These numbers are only rough approximations, and the differences become less marked as the thickness of the radio-active layer increases.

The average penetrating power of the rays is shown below. In the first column is given the thickness of the aluminium, which cuts each radiation down to half its value, and in the second the relative power of penetration of the rays.

+———————+—————————-+————————+
| | Thickness of | |
| | Aluminium in cms. | Relative power |
| Radiation |which cuts off half| of penetration |
| | the radiation | |
+———————+—————————-+————————+
| α rays | 0·0005 cms. | 1 |
| β " | 0·05 cms. | 100 |
| γ " | 8 cms. | 10000 |
+———————+—————————-+————————+

The relative power of penetration is thus approximately inversely proportional to the relative ionization. These numbers, however, only indicate the order of relative penetrating power. This power varies considerably for the different active bodies.

The α rays from uranium and polonium are the least penetrating, and those from thorium the most. The β radiations from thorium and radium are very complex, and consist of rays widely different in penetrating power. Some of the β rays from these substances are much less and others much more penetrating than those from uranium, which gives out fairly homogeneous rays.

74. Difficulties of comparative measurements. It is difficult to make quantitative or even qualitative measurements of the relative intensity of the three types of rays from active substances. The three general methods employed depend upon the action of the rays in ionizing the gas, in acting on a photographic plate, and in causing phosphorescent or fluorescent effects in certain substances. In each of these methods the fraction of the rays which is absorbed and transformed into another form of energy is different for each type of ray. Even when one specific kind of ray is under observation, comparative measurements are rendered difficult by the complexity of that type of rays. For example, the β rays from radium consist of negatively charged particles projected with a wide range of velocity, and, in consequence, they are absorbed in different amounts in passing through a definite thickness of matter. In each case, only a fraction of the energy absorbed is transformed into the particular type of energy, whether ionic, chemical, or luminous, which serves as a means of measurement.

The rays which are the most active electrically are the least active photographically. Under ordinary conditions, most of the photographic action of uranium, thorium, and radium, is due to the β or cathodic rays. The α rays from uranium and thorium, on account of their weak action, have not yet been detected photographically. With active substances like radium and polonium, the α rays readily produce a photographic impression. So far the γ rays have been detected photographically from radium only. That no photographic action of these rays has yet been established for uranium and thorium is probably due merely to the fact that the effect sought for is very small, and during exposures for long intervals it is very difficult to avoid fogging of the plates owing to other causes. Considering the similarity of the radiations in other respects, there can be little doubt that the γ rays do produce some photographic action, though it is too small to observe with certainty.

These differences in the photographic and ionizing properties of the radiations must always be taken into account in comparing results obtained by the two methods. The apparent contradiction of results obtained by different observers using these two methods is found to be due to their differences in relative photographic and ionizing action. For example, with the unscreened active material, the ionization observed by the electrical method is due almost entirely to α rays, while the photographic action under the same condition is due almost entirely to the β rays.

It is often convenient to know what thickness of matter is sufficient to absorb a specific type of radiation. A thickness of aluminium or mica of ·01 cms. or a sheet of ordinary writing-paper is sufficient to absorb completely all the α rays. With such a screen over the active material, the effects are due only to the β and γ rays, which pass through with a very slight absorption. Most of the β rays are absorbed in 5 mms. of aluminium or 2 mms. of lead. The radiation passing through such screens consists very largely of the γ rays. As a rough working rule, it may be taken that a thickness of matter required to absorb any type of rays is inversely proportional to the density of the substance, i.e. the absorption is proportional to the density. This rule holds approximately for light substances, but, in heavy substances like mercury and lead, the radiations are about twice as readily absorbed as the density rule would lead us to expect.

PART II.

The β or Cathodic Rays.

75. Discovery of the β rays. A discovery which gave a great impetus to the study of the radiations from active bodies was made in 1899, almost simultaneously in Germany, France, and Austria. It was observed that preparations of radium gave out some rays which were deviable by a magnetic field, and very similar in character to the cathode rays produced in a vacuum tube. The observation of Elster and Geitel that a magnetic field altered the conductivity produced in air by radium rays, led Giesel^[3] to examine the effect of a magnetic field on the radiations. In his experiments, the radio-active preparation was placed in a small vessel between the poles of an electromagnet. The vessel was arranged to give a pencil of rays which was approximately perpendicular to the field. The rays caused a small fluorescent patch on the screen. On exciting the electromagnet, the fluorescent zone was observed to broaden out on one side. On reversing the field, the extension of the zone was in the opposite direction. The deviation of the rays thus indicated was in the same direction and of the same order of magnitude as that for cathode rays.

S. Meyer and Schweidler^[4] also obtained similar results. They showed, in addition, the deviation of the rays by the alteration of the conductivity of the air when a magnetic field was applied. Becquerel^[5], a little later, showed the magnetic deflection of the radium rays by using the photographic method. P. Curie^[6], by the electrical method, showed furthermore that the rays from radium consisted of two kinds, one apparently non-deviable and easily absorbed (now known as the α rays), and the other penetrating and deviable by a magnetic field (now known as the β rays). The ionization effect due to the β rays was only a small fraction of that due to the α rays. At a later date Becquerel, by the photographic method, showed that uranium gave out some deflectable rays. It had been shown previously^[7] that the rays from uranium consisted of α and β rays. The deflected rays in Becquerel's experiment consisted entirely of β rays, as the α rays from uranium produce no appreciable photographic action. Rutherford and Grier^[8], using the electric method, showed that compounds of thorium, like those of uranium, gave out, besides α rays, some penetrating β rays, deviable in a magnetic field. As in the case of radium, the ionization due to the α rays of uranium and thorium is large compared with that due to the β rays.

76. Examination of the magnetic deviation by the photographic method. Becquerel has made a very complete study, by the photographic method, of the β rays from radium, and has shown that they behave in all respects like cathode rays, which are known to be negatively charged particles moving with a high velocity. The motion of a charged ion acted on by a magnetic field has been discussed in section 49. It has been shown that if a particle of mass m and charge e is projected with a velocity u, at an angle α with the direction of a uniform field of strength H, it will describe a helix round the magnetic lines of force. This helix is wound on a cylinder of radius R, with the axis parallel to the field, where R is given by

R = (mu/He) sin α.

When α = π/2, i.e. when the rays are projected normally to the

field, the particles describe circles of radius

R = mu/(He).

The planes of these circles are normal to the field. Thus, for a particular velocity u, the value of R varies inversely as the strength of the field. In a uniform field the rays projected normally to the field describe circles, and their directions of projection are the tangents at the origin.

This conclusion has been verified experimentally by Becquerel for the β rays of radium, by an arrangement similar to that shown in Fig. 23.

Fig. 23.

A photographic plate P, with the film downwards, is enveloped in black paper and placed horizontally in the uniform horizontal magnetic field of an electromagnet. The magnetic field is supposed to be uniform, and, in the figure, is at right angles to the plane of the paper. The plate was covered with a sheet of lead, and on the edge of the plate, in the centre of the magnetic field, is placed a small lead vessel R containing the radio-active matter.

On exciting the magnet, so that the rays are bent to the left of the figure, it is observed that a photographic impression is produced directly below the source of the rays, which have been bent round by the magnetic field. The active matter sends out rays equally in all directions. The rays perpendicular to the field describe circles, which strike the plate immediately under the source. A few of these rays, A_{1}, A_{2}, A_{3}, are shown in the figure. The rays, normal to the plate, strike the plate almost normally, while the rays nearly parallel to the plate strike the plate at grazing incidence. The rays, inclined to the direction of the field, describe spirals and produce effects on an axis parallel to the field passing through the source. In consequence of this, any opaque screen placed in the path of the rays has its shadow thrown near the edge of the photographic plate.

77. Complexity of the rays. The deviable rays from radium are complex, i.e. they are composed of a flight of particles projected with a wide range of velocity. In a magnetic field every ray describes a path, of which the radius of curvature is directly proportional to the velocity of projection. The complexity of the radiation has been shown very clearly by Becquerel^[9] in the following way.

An uncovered photographic plate, with the film upwards, was placed horizontally in the horizontal uniform magnetic field of an electromagnet. A small, open, lead box, containing the radio-active matter, was placed in the centre of the field, on the photographic plate. The light, due to the phosphorescence of the radio-active matter, therefore, could not reach the plate. The whole apparatus was placed in a dark room. The impression on the plate took the form of a large, diffuse, but continuous band, elliptic in shape, produced on one side of the plate.

Such an impression is to be expected if the rays are sent out in all directions, even if their velocities of projection are the same, for it can readily be shown theoretically, that the path of the rays is confined within an ellipse whose minor axis, which is at right angles to the field, is equal to 2R, and whose major axis is equal to πR. If, however, the active matter is placed in the bottom of a deep lead cylinder of small diameter, the rays have practically all the same direction of projection, and in that case each part of the plate is acted on by rays of a definite curvature.

In this case also, a diffuse impression is observed on the plate, giving, so to speak, a continuous spectrum of the rays and showing that the radiation is composed of rays of widely different curvatures. Fig. 24 shows a photograph of this kind obtained by Becquerel,

with strips of paper, aluminium, and platinum placed on the plate.

Fig. 24.

If screens of various thickness are placed on the plate, it is observed that the plate is not appreciably affected within a certain distance from the active matter, and that this distance increases with the thickness of the screen. This distance is obviously equal to twice the radius of curvature of the path of the rays, which are just able to produce an impression through the screen.

These experiments show very clearly that the most deviable rays are those most readily absorbed by matter. By observations of this kind Becquerel has determined approximately the inferior limit of the value of HR for rays which are transmitted through different thicknesses of matter.

The results are given in the table below:

+——————-+—————-+————————+
| | Thickness | Inferior limit |
| Substance | in mms. | of HR for |
| | |transmitted rays|
+——————-+—————-+————————+
| Black paper | 0·065 | 650 |
| Aluminium | 0·010 | 350 |
| " | 0·100 | 1000 |
| " | 0·200 | 1480 |
| Mica | 0·025 | 520 |
| Glass | 0·155 | 1130 |
| Platinum | 0·030 | 1310 |
| Copper | 0·085 | 1740 |
| Lead | 0·130 | 2610 |
+——————-+—————-+————————+

If e/m is a constant for all the rays, the value of HR is proportional to the velocity of the rays, and it follows from the table that the velocity of the rays which just produce an effect on the plate through ·13 mms. of lead is about 7 times that of the rays which just produce an impression through ·01 mm. of aluminium. It will be shown, however, in section 82, that e/m is not a constant for all speeds, but decreases with increase of velocity of the rays. The difference in velocity between the rays is in consequence not as great as this calculation would indicate. On examination of the rays from uranium, Becquerel found that the radiation is not as complex as that from radium, but consists wholly of rays for which the value of HR is about 2000.

78. Examination of the β rays by the electric method. The presence of easily deviable rays given off from an active substance can most readily be shown by the photographic method, but it is necessary, in addition, to show that the penetrating rays which produce the ionization in the gas are the same as those which cause the photographic action. This can be conveniently tested in an arrangement similar to that shown in Fig. 25.

Fig. 25.

The radio-active matter A is placed on a lead block B´´ between the two parallel lead plates BB´. The rays pass between the parallel plates and ionize the gas between the plates PP´ of the testing vessel. The magnetic field is applied at right angles to the plane of the paper. The dotted rectangle EEEE represents the position of the pole piece. If a compound of radium or thorium is under investigation, a stream of air is required to prevent the diffusion of the radio-active emanations into the testing vessel. When a layer of uranium, thorium or radium compound is placed at A, the ionization in the testing vessel is due mainly to the action of the α and β rays. The α rays are cut off by adding a layer of aluminium ·01 cm. thick over the active material. When the layer of active matter is not more than a few millimetres thick, the ionization due to the γ rays is small compared with that produced by the β rays, and may be neglected. On the application of a magnetic field at right angles to the mean direction of the rays, the ionization in the testing vessel due to the rays steadily decreases as the strength of the field increases, and in a strong field it is reduced to a very small fraction of its original value. In this case the rays are bent so that none of them enter the testing vessel.

Examined in this way, it has been found that the β rays of uranium, thorium, and radium consist entirely of rays readily deflected by a magnetic field. The rays from polonium consist entirely of α rays, the deviation of which can be detected only in very intense magnetic fields.

When the screen covering the active material is removed, in a strong magnetic field, the ionization in the vessel is mainly due to the α rays. On account of the slight deviation of the α rays under ordinary experimental conditions, a still greater increase of the magnetic field does not appreciably alter the current due to them in the testing vessel.

The action of a magnetic field on a very active substance like radium is easily shown by the electrical method, as the ionization current due to the deviable rays is large. With substances of small activity like uranium and thorium, the ionization current due to the deviable rays is very small, and a sensitive electrometer or an electroscope is required to determine the variation, in a magnetic field, of the very small current involved. This is especially the case for thorium oxide, which gives out only about 1/5 of the amount of deviable rays given out by the same weight of uranium oxide.

79. Experiments with a fluorescent screen. The β rays from a few milligrams of pure radium bromide produce intense fluorescence in barium platino-cyanide and other substances which can be made luminous under the influence of the cathode rays. Using a centigram of radium bromide, the luminosity on a screen, placed upon it, is bright enough to be observed in daylight. With the aid of such a screen in a dark room many of the properties of the β rays may be simply illustrated and their complex nature clearly shown. A small quantity of radium is placed in the bottom of a short, narrow, lead tube open at one end. This is placed between the pole pieces of an electromagnet, and the screen placed below it. With no magnetic field, a faint luminosity of the screen is observed due to the very penetrating γ rays which readily pass through the lead. When the magnetic field is put on, the screen is brightly lighted up on one side over an area elliptical in shape (section 77). The direction of deviation is reversed by reversal of the field. The broad extent of the illumination shows the complex nature of the β rays. On placing a metallic object at various points above the screen, the trajectory of the rays can readily be traced by noticing the position of the shadow cast upon the screen. By observing the density of the shadow, it can be seen that the rays most easily deviated are the least penetrating.

Comparison of the β rays with cathode rays.

80. Means of comparison. In order to prove the identity of the β rays from active bodies with the cathode rays produced in a vacuum tube, it is necessary to show

(1) That the rays carry with them a negative charge;

(2) That they are deviated by an electric as well as by a magnetic field;

(3) That the ratio e/m is the same as for the cathode rays.

Electric charge carried by the β rays. The experiments of Perrin and J. J. Thomson have shown that the cathode rays carry with them a negative charge. In addition, Lenard has shown that the rays still carry a charge after traversing thin layers of matter. When the rays are absorbed, they give up their charge to the body which absorbs them. The total amount of charge carried by the β rays from even a very active preparation of radium is, in general, small compared with that carried by the whole of the cathode rays in a vacuum tube, and can be detected only by delicate methods.

 

Suppose that a layer of very active radium is spread on a metal plate connected to earth, and that the β rays are absorbed by a parallel plate connected with an electrometer. If the rays are negatively charged, the top plate should receive a negative charge increasing with the time. On account, however, of the great ionization produced by the rays between the plates, any charge given to one of them is almost instantly dissipated. In many cases, the plate does become charged to a definite positive or negative potential depending on the metal, but this is due to the contact difference of potential between the plates, and would be produced whether the rays were charged or not. The ionization of the gas is greatly diminished by placing over the active material a metal screen which absorbs the α rays, but allows the β rays to pass through with little absorption.

The rapid loss of any charge communicated to the top plate can be very much reduced, either by diminishing the pressure of the gas surrounding it or by enclosing the plate with suitable insulators. In their experiments to determine the amount of charge carried by the radium rays, M. and Mme Curie^[10] used the second method.

A metal disc MM (Fig. 26) is connected with an electrometer by the wire T. The disc and wire are completely surrounded by insulating matter ii. The whole is surrounded by a metal envelope EEEE connected with earth. On the lower side of the disc, the insulator and the metallic covering are very thin. This side is exposed to the rays of the radium R placed in a depression in a lead plate AA.

Fig. 26.

The rays of the radium pass through the metal cover and insulator with little absorption, but they are completely absorbed by the disc MM. It was observed that the disc received a negative charge which increased uniformly with the time, showing that the rays carry with them a negative charge. The current observed was very small. With an active preparation of radium^[11], forming a layer 2·5 sq. cms. in area and 2 mms. thick, a current of the order of 10^{-11} amperes was observed after the rays had traversed a layer of aluminium ·01 mm. thick and a layer of ebonite ·3 mm. thick. The current was the same with discs of lead, copper, and zinc, and also when the ebonite was replaced by paraffin.

Curie also observed in another experiment of a similar character that the radium itself acquired a positive charge. This necessarily follows if the rays carry with them a negative charge. If the β rays alone carried with them a charge, a pellet of radium, if perfectly insulated, and surrounded by a non-conducting medium, would in the course of time be raised to a high positive potential. Since, however, the α rays carry with them a charge opposite in sign to the β rays, the ratio of the charge carried off by the two types of rays must be determined, before it can be settled whether the radium would acquire a positive or a negative charge. If, however, the radium is placed in an insulated metal vessel of a thickness sufficient to absorb all the α rays, but not too thick to allow most of the β rays to escape, the vessel will acquire a positive charge in a vacuum.

An interesting experimental result bearing upon this point has been described by Dorn^[12]. A small quantity of radium was placed in a sealed glass tube and left for several months. On opening the tube with a file, a bright electric spark was observed at the moment of fracture, showing that there was a large difference of potential between the inside of the tube and the earth.

In this case the α rays were absorbed in the walls of the tube, but a large proportion of the β rays escaped. The inside of the tube thus became charged, in the course of time, to a high positive potential; a steady state would be reached when the rate of escape of negative electricity was balanced by the leakage of positive electricity through the walls of the tube. The external surface of the glass would be always practically at zero potential, on account of the ionization of the air around it.

Strutt^[13] has recently described a simple and striking experiment to illustrate still more clearly that a radium preparation acquires a positive charge, if it is enclosed in an envelope thick enough to absorb all the α particles, but thin enough to allow most of the β particles to escape. The experimental arrangement is clearly seen in Fig. 27. A sealed tube AA containing the radium, was attached at one end to a pair of thin gold leaves in metallic connection with the radium, and was insulated inside a larger tube by means of a quartz rod B. The inner surface of the tube was coated with tinfoil EE connected to earth. The glass surface of AA was made conducting by a thin coating of phosphoric acid. The air in the outer tube was exhausted as completely as possible by means of a mercury pump, in order to reduce the ionization in the gas, and consequently the loss of any charge gained by the gold leaves. After an interval of 20 hours, the gold leaves were observed to diverge to their full extent, indicating that they had acquired a large positive charge. In this experiment Strutt used 1/2 gram of radiferous barium of activity only 100 times that of uranium.

Fig. 27.

If the tube is filled with 30 mgrs. of pure radium bromide, the leaves diverge to their full extent in the course of about a minute. If it is arranged that the gold leaf, at a certain angle of divergence, comes in contact with a piece of metal connected with earth, the apparatus can be made to work automatically. The leaf diverges, touches the metal, and at once collapses, and this periodic movement of the leaf will continue, if not indefinitely, at any rate as long as the radium lasts. This "radium clock" should work at a sensibly uniform rate for many years, but, from evidence considered later (Section 254), there is reason to believe that the number of β particles emitted would decrease exponentially with the time, falling to half value in about 1200 years. The period of movement of the leaf should thus gradually increase with the time, and ultimately the effect would become too small to observe. The action of this radium clock is the nearest approach to an apparent perpetual motion that has so far been observed.

A determination of the amount of the charge carried off by the β rays of radium has been made by Wien^[14]. A small quantity of radium, placed in a sealed platinum vessel, was hung by an insulating thread inside a glass cylinder, which was exhausted to a low pressure. A connection between the platinum vessel and an electrode sealed on to the external glass cylinder could be made, when required, by tilting the tube. Wien found that in a good vacuum the platinum vessel became charged to about 100 volts. The rate of escape of negative electricity from the platinum vessel containing 4 milligrams of radium bromide corresponded to 2·91 × 10^{-12} amperes. If the charge on each particle is taken as 1·1 × 10^{-20} electro-magnetic units, this corresponds to an escape of 2·66 × 10^7 particles per second. From 1 gram of radium bromide the corresponding number would be 6·6 × 10^9 per second. Since some of the β rays are absorbed in their passage through the walls of the containing vessel and through the radium itself, the actual number projected per second from 1 gram of radium bromide must be greater than the above value. This has been found by the writer to be the case. The method employed reduced the absorption of the β rays to a minimum, and the total number emitted per second by 1 gram of radium bromide in radio-active equilibrium was found to be 4·1 × 10^{10}, or about six times the number found by Wien. A detailed account of the method employed cannot be given with advantage at this stage, but will be found later in Section 246.

81. Determination of e/m. We have seen (Section 50) that, in their passage between the plates of a condenser, the cathode rays are deflected towards the positive plate. Shortly after the discovery of the magnetic deviation of the β rays from radium, Dorn^[15] and Becquerel^[16] showed that they also were deflected by an electric field.

By observing separately the amount of the electric and magnetic deviation, Becquerel was able to determine the ratio of e/m and the velocity of the projected particles. Two rectangular copper plates, 3·45 cms. high and 1 cm. apart, were placed in a vertical plane and insulated on paraffin blocks. One plate was charged to a high potential by means of an influence machine, and the other was connected with earth. The active matter was placed in a narrow groove cut in a lead plate parallel to the copper plates and placed midway between them. The photographic plate, enveloped in black paper, was placed horizontally above the plate containing the active substance. The large and diffuse pencil of rays thus obtained was deflected by the electric field, but the deviation amounted to only a few millimetres and was difficult to measure. The method finally adopted was to place vertically above the active matter a thin screen of mica, which cut the field into two equal parts. Thus, in the absence of an electric field, a narrow rectangular shadow was produced on the plate.

When the electric field was applied, the rays were deflected and a part of the pencil of rays was stopped by the mica screen. A shadow was thus cast on the plate which showed the direction of deviation and corresponded to the least deviable rays which gave an impression through the black paper.

If a particle of mass m, charge e, and velocity u, is projected normally to an electric field of strength X, the acceleration α is in the direction of the field, and is given by

α = Xe/m.

Since the particle moves with a constant acceleration parallel to the field, the path of the particle is the same as that of a body projected horizontally from a height with a constant velocity and acted on by gravity. The path of the particle is thus a parabola, whose axis is parallel to the field and whose apex is at the point where the particle enters the electric field. The linear deviation d_{1} of the ray parallel to the field after traversing a distance l is given by

d_{1} = (1/2)(Xe/m)(l^2/u^2).

On leaving the electric field, the particle travels in the direction of the tangent to the path at that point. If θ is the angular deviation of the path at that point

tan θ = eXl/(mu^2).

The photographic plate was at a distance h above the extremity of

the field. Thus the particles struck the plate at a distance d_{2} from the original path given by

d_{2} = h tan θ + d_{1}
        = (Xle/(mu^2))(l/2 + h).

In the experimental arrangement the values were

d_{2} = ·4 cms.;
X = 1·02 × 10^{12};
l = 3·45 cms.;
h = 1·2 cms.

If the radius R of curvature of the path of the same rays is observed in a magnetic field of strength H perpendicular to the rays,

e/m = V/(HR).

Combining these two equations we get

u = (X . l(l/2 + h))/(H . R . d_{2}).

A difficulty arose in identifying the part of the complex pencil of rays for which the electric and magnetic deviations were determined. Becquerel estimated that the value of HR for the rays deflected by the electric field was about 1600 C.G.S. units. Thus

u = 1·6 × 10^{10} cms. per second,
and e/m = 10^7.

Thus these rays had a velocity more than half the velocity of light, and an apparent mass about the same as the cathode ray particles, i.e. about 1/1000 of the mass of the hydrogen atom. The β ray is therefore analogous in all respects to the cathode ray, except that it differs in velocity. In a vacuum tube the cathode rays generally have a velocity of about 2 × 10^9 cms. per sec. In special tubes with strong fields the velocity may be increased to about 10^{10} cms. per sec. These β particles, then, behave like isolated units of negative electricity, identical with the electrons set free by an electric discharge in a vacuum tube. The electrons projected from radium have velocities varying from about 0·2V to at least 0·96V, where V is the velocity of light, and thus have an average speed considerably greater than that of the electrons produced in a vacuum tube. These moving electrons are able to pass through much greater thicknesses of matter before they are absorbed than the slower electrons produced in a vacuum tube, but the difference is one merely of degree and not of kind. Since electrons are continuously and spontaneously expelled from radium with enormous velocities, they must acquire their energy of motion from the matter itself. It is difficult to avoid the conclusion, that this velocity has not been suddenly impressed on the electron. Such a sudden gain of velocity would mean an immense and sudden concentration of energy on a small particle, and it is more probable that the electron before its expulsion has been in rapid orbital or oscillatory motion in the atom, and, by some means, suddenly escapes from its orbit. According to this view, the energy of the electron is not suddenly created but is only made obvious by its escape from the system to which it belongs.

82. Variation of e/m with the velocity of the electron. The fact that radium throws off electrons with rates of speed varying from 1/5 to 9/10 the velocity of light has been utilised by Kaufmann^[17] to examine whether the ratio e/m of the electrons varies with the speed. We have seen (Section 48) that, according to the electromagnetic theory, a charge of electricity in motion behaves as if it had apparent mass. For small speeds, this additional electrical mass is equal to (2/3)(e^2/a), where a is the radius of the body, but it increases rapidly as the speed of light is approached. It is very important to settle whether the mass of the electron is due partly to mechanical and partly to electrical mass, or whether it can be explained by virtue of electricity in motion independently of the usual conception of mass.

Slightly different formulae expressing the variation of mass with speed have been developed by J. J. Thomson, Heaviside, and Searle. To interpret his results Kaufmann used a formula

developed by M. Abraham^[18].

Let m_{0} = mass of electron for slow speeds;
    m = apparent mass of electron at any speed;
    u = velocity of electron;
    V = velocity of light.

Let β = u/V; then it can be shown that

m/m_{0} = (3/4)ψ(β) (1),

where ψ(β) = (1/β^2)[(1 + β^2)/(2β) log ((1 + β)/(1 - β)) - 1] (2).

The experimental method employed to determine e/m and u is similar to the method of crossed spectra. Some strongly active radium was placed at the bottom of a brass box. The rays from this passed between two brass plates insulated and about 1·2 mm. apart. These rays fell on a platinum diaphragm, containing a small tube about 0·2 mm. in diameter, which allowed a narrow bundle of rays to pass. The rays then struck a photographic plate enveloped in a thin layer of aluminium.

In the experiments the diaphragm was about 2 cms. from the active material and at the same distance from the photographic plate. When the whole apparatus was placed in a vacuum, a P.D. of from 2000 to 5000 volts could be applied between the plates without a spark. The rays were deflected in their passage through the electric field, and produced what may be termed an electric spectrum on the plate.

Fig. 28.

If a magnetic field is superimposed parallel to the electric field by means of an electromagnet, a magnetic spectrum is obtained perpendicular to the electric spectrum. The combination of the two spectra gives rise to a curved line on the plate. The double trace obtained on the photographic plate with reversal of the magnetic field is shown in Fig. 28. Disregarding some small corrections, it can readily be shown that if y and z are the electric and magnetic deviations respectively,

         β = κ_{1}(z/y) (3),
and e/m = κ(z^2/y) (4).

From these two equations, combined with (1), we obtain

y/(z^2ψ(κ_{1}(z/y))) = κ_{2} (5),

where κ, κ_{1}, κ_{2} are constants.

Equation (5) gives the curve that should be obtained on the plate according to the electromagnetic theory. This is compared by trial with the actual curve obtained on the plate.

In this way Kaufmann^[19] found that the value of e/m decreased with the speed, showing that, assuming the charge constant, the mass of the electron increased with the speed.

The following numbers give some of the preliminary results obtained by this method.

+———————————————+——————-+
| Velocity of electron | e/m |
+———————————————+——————-+
| 2·36 × 10^{10} cms. per sec. | 1·31 × 10^7 |
| 2·48 " " | 1·17 × 10^7 |
| 2·59 " " | 0·97 × 10^7 |
| 2·72 " " | 0·77 × 10^7 |
| 2·85 " " | 0·63 × 10^7 |
+———————————————+——————-+

For the cathode rays S. Simon^[20] obtained a value for e/m of 1·86 × 10^7 for an average speed of about 7 × 10^9 cms. per second.

In a later paper^[21] with some very active radium, more satisfactory photographs were obtained, which allowed of accurate measurement. The given equation of the curve was found to agree satisfactorily with experiment.

The table given below, deduced from the results given by Kaufmann, shows the agreement between the theoretical and experimental values, u being the velocity of the electron and V that of light.

The average percentage error between the observed and calculated value is thus not much more than one per cent. It is remarkable how nearly the velocity of the electron has to approach the velocity of light before the value of m/m_{0} becomes large. This

+————+————————-+——————————-+
|Value of|Observed value of|Percentage difference|
|u/V | m/m_{0} | from theoretical |
| | | values |
+————+————————-+——————————-+
| Small | 1 | |
| ·732 | 1·34 | -1·5 % |
| ·752 | 1·37 | -0·9 " |
| ·777 | 1·42 | -0·6 " |
| ·801 | 1·47 | +0·5 " |
| ·830 | 1·545 | +0·5 " |
| ·860 | 1·65 | 0 " |
| ·883 | 1·73 | +2·8 " |
| ·933 | 2·05 | -7·8 " ? |
| ·949 | 2·145 | -1·2 " |
| ·963 | 2·42 | +0·4 " |
+————+————————-+——————————-+

is shown in the following table which gives the calculated values of m/m_{0} for different velocities of the electron.

Value of u/V small ·1 ·5 ·9 ·99 ·999 ·9999 ·999999
Calculated value m/m_{0} 1·00 1·015 1·12 1·81 3·28 4·96 6·68 10·1

Thus for velocities varying from 0 to 1/10 the velocity of light, the mass of the electron is practically constant. The increase of mass becomes appreciable at about half the velocity of light, and increases steadily as the velocity of light is approached. Theoretically the mass becomes infinite at the velocity of light, but even when the velocity of the electron only differs from that of light by one part in a million, its mass is only 10 times the value for slow speeds.

The above results are therefore in agreement with the view that the mass of the electron is altogether electrical in origin and can be explained purely by electricity in motion. The value of e/m_{0}, for slow speeds, deduced from the results was 1·84 × 10^7, which is in very close agreement with the value obtained by Simon for the cathode rays, viz. 1·86 × 10^7. If the electricity carried by the electron is supposed to be distributed uniformly over a sphere of radius a, for speeds slow compared with the velocity of light, the apparent mass m_{0} = (2/3)(e^2/a).

Therefore a = (2/3)(e/m_{0}) . e.

Taking the value of e as 1·13 × 10^{-20}, a is 1·4 × 10^{-13} cms.

Thus the diameter of an electron is minute compared with the diameter of an atom.

83. Distribution of velocity amongst the β particles. Some interesting experiments have been recently made by Paschen^[22] to determine the relative number of β particles which are expelled from radium at the different speeds. The experimental arrangement is shown in Fig. 29.

Fig. 29.

A small thin silvered glass tube b, containing 15 mgrs. of radium bromide, was placed in the axis of a number of lead vanes arranged round a cylinder of diameter 2 cms. and length 2·2 cms. When no magnetic field was acting, the β particles from the radium passed through the openings and were absorbed in an outer concentric cylinder aa of lead of inner diameter 3·7 cms. and of thickness 5·5 mms. This outer cylinder was rigidly connected to the inner cylinder cc by quartz rods ii, which also served to insulate it. The cylinder c and the radium were connected with earth. A gold-leaf electroscope E was attached to a, and the whole apparatus was enclosed in a glass vessel which was exhausted to a low vacuum by means of a mercury pump. The glass vessel was placed in the uniform field of a large electromagnet, so that the axis of the lead cylinder was parallel to the lines of force.

The outer cylinder gains a negative charge on account of the particles which are absorbed in it. This negative charge, which is indicated by the movement of the gold-leaf, tends to be dissipated by the small ionization produced in the residual gas by the passage of the β rays. This action of the gas can be eliminated by observing the rate of movement of the gold leaf when charged alternately to an initial positive and negative potential. The mean of the two rates is proportional to the number of β particles which give up their charge to the lead cylinder. This is evidently the case, since, when the charge is positive, the ionization of the gas assists the rate of movement of the gold-leaf, and, when negative, diminishes it to an equal extent.

When a magnetic field is applied, each of the particles describes a curved path, whose radius of curvature depends on the velocity of the particle. For weak fields, only the particles of smallest velocity will be deflected sufficiently not to strike the outer cylinder, but, as the field is raised, the number will increase until finally all the β particles fail to reach the outer cylinder. The decrease of the charge communicated to the outer cylinder with the increase of the strength of the magnetic field is shown graphically in Fig. 30, Curve I.

The ordinates represent in arbitrary units the charge communicated to the lead cylinder per second, and thus serve as a measure of the number of β particles which reach the cylinder. Knowing the dimensions of the apparatus, and assuming the value e/m found by Kaufmann, the velocity of the particles which just fail to reach the lead cylinder can be deduced from any strength of the magnetic field. Curve II, Fig. 30 is the first differential of Curve I, and the ordinates represent the relative number of β particles which are projected at each velocity.

Fig. 30.

From the data given by Kaufmann (see section 82) Paschen deduced that the group of rays examined by the former, which had velocities lying between 2·12 × 10^{10} and 2·90 × 10^{10} cms. per second, corresponded to the group of rays between the points A and B, that is, to the group of rays which were completely deflected from the lead cylinder between the magnetic fields of strengths of 1875 and 4931 C.G.S. units. Since radium gives off β particles which require a field of strength over 7000 units to deflect them, Paschen concluded that β particles are expelled from radium with still greater velocities than the highest recorded by Kaufmann.

Paschen considered that the small charge observed in still higher fields was mainly due to the γ rays. The effect is small and is probably not due to an actual charge carried by the γ rays but to a secondary effect produced by them. This question will be discussed in more detail in section 112.

There is a group of low velocity β particles emitted by radium (see Fig. 30) which have about the same speed as the electrons set free in a vacuum tube. In consequence of their small velocity, these probably produce a large proportion of the ionization due to the β rays at short distances from the radium, for it will be shown (section 103) that the ionization produced by an electron per unit length of path steadily decreases with increase of its velocity above a small limiting value. This observation is confirmed by experiments on the absorption of the β rays in passing through matter.

In Paschen's experiments, the glass tube containing the radium was ·5 mms. thick, so that a considerable proportion of the low velocity β particles must have been stopped by it. This is borne out by some later experiments of Seitz which will be described in section 85.

84. Absorption of the β rays by matter. The β particles produce ions in their passage through the gas and their energy of motion is consequently diminished. A similar action takes place also when the β rays pass through solid and liquid media, and the mechanism of absorption is probably similar in all cases. Some of the particles in their passage through matter are completely stopped, while others have their velocity reduced. In addition, there is a considerable scattering or diffuse reflection of the rays in traversing matter. The amount of this scattering depends upon the density of the substance and also upon the angle of incidence of the rays. This scattering of the rays will be discussed later in section 111.

There are two general methods of determining the absorption of the β rays. In the first method, the variation of the ionization current is observed in a testing vessel when the active matter is covered by screens differing in material and thickness. This ionization in the vessel depends upon two quantities, viz. the number of β particles which pass through the matter and also upon the number of ions produced by them per unit path. In the absence of any definite information in regard to the variation of ionization by the electron with its velocity, no very definite conclusions can be drawn from such experiments.

The advent of pure radium-bromide has made it possible to determine the actual number of electrons which are absorbed in their passage through a definite thickness of matter, by measuring the negative charge carried by the issuing rays. Experiments of this character have been made by Seitz and will be considered later.

These two methods of determining the absorption of β rays are quite distinct in principle, and it is not to be expected that the values of the coefficients of absorption obtained in the two cases should be the same. The whole question of the absorption of electrons by matter is very complicated, and the difficulty is still further increased by the complexity of the β rays emitted by the radio-active substances. Many of the results obtained by different methods, while pointing to the same general conclusion, are quantitatively in wide disagreement. Before any definite advance can be made to a better understanding of the mechanism of absorption, it will be necessary to determine the variation of the ionization with the speed of the electron over a very wide range. Some work has already been done in this direction but not between sufficiently wide limits.

Ionization method.

We shall first consider the results obtained on the absorption of β rays by measuring the variation of the ionization current, when screens of different thickness are placed over the active substance. When the active matter is covered with aluminium foil of thickness ·1 mm., the current in a testing vessel such as is shown in Fig. 17, is due almost entirely to the β rays. If a uranium compound is used, it is found that the saturation current decreases with the thickness of matter traversed nearly according to an exponential law. Taking the saturation current as a measure of the intensity of the rays, the intensity I after passing through a thickness d of matter is given by

I/I_{0} = e^{-λd},

where λ is the constant of absorption of the rays and I_{0} is the initial intensity. For uranium rays, the current is reduced to half its value after passing through about ·5 mm. of aluminium.

If a compound of thorium or radium is examined in the same way, it is found that the current does not decrease regularly according to the above equation. Results of this kind for radium rays have been given by Meyer and Schweidler^[23]. The amount of absorption of the rays by a certain thickness of matter decreases with the thickness traversed. This is exactly opposite to what is observed for the α rays. This variation in the absorption is due to the fact that the β rays are made up of rays which vary greatly in penetrating power. The rays from uranium are fairly homogeneous in character, i.e. they consist of rays projected with about the same velocity. The rays from radium and thorium are complex, i.e. they consist of rays projected with a wide range of velocity and consequently with a wide range of penetrating power. The electrical examination of the deviable rays thus leads to the same results as their examination by the photographic method.

Results on the absorption of cathode rays have been given by Lenard^[24], who has shown that the absorption of cathode rays is nearly proportional to the density of the absorbing matter, and is independent of its chemical state. If the deviable rays from active bodies are similar to cathode rays, a similar law of absorption is to be expected. Strutt^[25], working with radium rays, has determined the law of absorption, and has found it roughly proportional to the density of matter over a range of densities varying from 0·041 for sulphur dioxide to 21·5 for platinum. In the case of mica and cardboard, the values of λ divided by the density were 3·94 and 3·84 respectively, while the value for platinum was 7·34. In order to deduce the absorption coefficient, he assumed that the radiation fell off according to an exponential law with the distance traversed. As the rays from radium are complex, we have seen that this is only approximately the case.

Since the β rays from uranium are fairly homogeneous, and are at the same time penetrating in character, they are more suitable for such a determination than the complex rays of radium. I have in consequence made some experiments with uranium rays to determine the dependence of absorption on the density. The results obtained are given in the following table, where λ is the coefficient of absorption. It will be observed that the value of the absorption constant divided by the density is very nearly the same for such different

+—————-+—————+————-+—————+
| Substance | [Greek: lambda] | Density | [Greek: lambda] |
| | | | Density |/Density]
+—————-+—————+————-+—————+
| Glass | 14·0 | 2·45 | 5·7 |
| Mica | 14·2 | 2·78 | 5·1 |
| Ebonite | 6·5 | 1·14 | 5·7 |
| Wood | 2·16 | ·40 | 5·4 |
| Cardboard | 3·7 | ·70 | 5·3 |
| Iron | 44 | 7·8 | 5·6 |
| Aluminium | 14·0 | 2·60 | 5·4 |
| Copper | 60 | 8·6 | 7·0 |
| Silver | 75 | 10·5 | 7·1 |
| Lead | 122 | 11·5 | 10·8 |
| Tin | 96 | 7·3 | 13·2 |
+—————-+—————+————-+—————+

substances as glass, mica, ebonite, wood, iron and aluminium. The divergences from the law are great, however, for the other metals examined, viz. copper, silver, lead and tin. In tin the value of [Greek: lambda] divided by the density is 2·5 times its value for iron and aluminium. These differences show that a law for the absorption of the [Greek: beta] rays depending only on the density does not hold for all substances. With an exception in the case of tin, the value of [Greek: lambda] divided by the density for the metals increases in the same order as their atomic weights.

The absorption of the [Greek: beta] rays by matter decreases very rapidly with increase of speed. For example, the absorption of cathode rays in Lenard's experiment (loc. cit.) is about 500 times as great as for the uranium [Greek: beta] rays. The velocity of the [Greek: beta] rays of uranium was found by Becquerel to be about 1·6 × 10^{10} cms. per sec. The velocity of the cathode rays used in Lenard's experiment was certainly not less than 1/10 of this, so that, for a decrease of speed of less than 10 times, the absorption has increased over 500 times.

85. Number of electrons stopped by matter. An account will now be given of the experiments made by Seitz^[26], to determine the relative number of electrons which are stopped in their passage through different thicknesses of matter. The experimental arrangement is shown in Fig. 31.

The radium was placed outside a glass vessel containing an insulated brass plate P, the connection of which with a wire leading to the electrometer could be made or broken by a simple electromagnetic device. The β rays from the radium R, after passing through openings in a brass plate A, covered with thin aluminium foil, were absorbed in the plate P. The glass vessel was exhausted, and the charge communicated to P by the β rays was measured by an electrometer.

In a good vacuum, the magnitude of the current observed is a measure of the number of β particles absorbed by the upper plate^[27]. The following table shows the results obtained when different thicknesses of tin foil were placed over the radium. The second table gives the ratio I/I_{0} where I_{0} is the rate of discharge observed before the absorbing screen is introduced. The mean value of the absorption constant λ was deduced from the equation I/I_{0} = e^{-λd} where d is the thickness of matter traversed.

Fig. 31.

The values included in the brackets have not the same accuracy as the others. There is thus a wide difference in penetrating power of the β particles emitted from radium, and some of them are very readily absorbed. When a lead screen 3 mms. thick was placed over the radium—a thickness sufficient to absorb all the readily deflectable β rays—a small negative charge was still given to the plate, corresponding to ·29 per cent. of the maximum. This is a very much smaller value than was observed by Paschen (see Fig. 30). This

| Thickness of | | |
| Tin in mms. |I/I_{0}|λ |
+———————+—————-+ |
| 0·00834 | ·869 | 175 |
| 0·0166 | ·802 | 132·5 |
| 0·0421 | ·653 | 101·5 |
| 0·0818 | ·466 | 93·5 |
| 0·124 | ·359 | 82·5 |
| 0·166 | ·289 | 74·9 |
| 0·205 | ·230 | 71·5 |
| 0·270 | ·170 | 65·4 |
| 0·518 | ·065 } | 53} |
| 0·789 | ·031 } | 44} |
| 1·585 | ·0059} | 32} |
| 2·16 | ·0043} | 25} |
+———————+—————-+————-+

difference may, in part, be due to the fact that, in Paschen's experiments, a large proportion of the slow velocity electrons were absorbed in the glass tube of ·5 mm. thickness containing the radium.

Seitz also determined the relative thickness, compared with tin, of different substances which reduced the negative charge communicated to P by a definite amount. A few of the numbers are given below, and expressed in terms of tin as unity.

+—————-+—————-+
| | Thickness |
| Substance | Tin = 1 |
+—————-+—————-+
| Lead | ·745 |
| Gold | ·83 |
| Platinum | ·84 |
| Silver | 1 |
| Steel | 1·29 |
| Aluminium | 1·56 |
| Water | 1·66 |
| Paraffin | 1·69 |
+—————-+—————-+

The thickness required to stop a given proportion of the β rays thus decreases with the density, but not nearly so fast as the density increases. These results are difficult to reconcile with the density-law of absorption found by Lenard from the cathode rays, or with the results of the ionization method already considered. A further experimental examination of the whole question is very much to be desired.

86. Variation of the amount of radiation with the thickness of the layer of radiating material. The radiations are sent out equally from all portions of the active mass, but the ionization of the gas which is measured is due only to the radiations which escape into the air. The depth from which the radiations can reach the surface depends on the absorption of the radiation by the active matter itself.

Let λ be the absorption constant of the homogeneous radiation by the active material. It can readily be shown that the intensity I of the rays issuing from a layer of active matter, of thickness d, is given by

I/I_{0} = 1 - e^{-λd},

where I_{0} is the intensity at the surface due to a very thick layer.

This equation has been confirmed experimentally by observing the current due to the β rays for different thicknesses of uranium oxide. In this case I = (1/2)I_{0} for a thickness of oxide corresponding to ·11 gr. per sq. cm. This gives a value of λ divided by density of 6·3. This is a value slightly greater than that observed for the absorption of the same rays in aluminium. Such a result shows clearly that the substance which gives rise to the β rays does not absorb them to a much greater extent than does ordinary matter of the same density.

The value of λ will vary, not only for the different active substances, but also for the different compounds of the same substance. PART III.

The [Greek: alpha] Rays.

87. The [Greek: alpha] rays. The magnetic deviation of the [Greek: beta] rays was discovered towards the end of 1899, at a comparatively early stage in the history of radio-activity, but three years elapsed before the true character of the [Greek: alpha] rays was disclosed. It was natural that great prominence should have been given in the early stages of the subject to the [Greek: beta] rays, on account of their great penetrating power and marked action in causing phosphorescence in many substances. The [Greek: alpha] rays were, in comparison, very little studied, and their importance was not generally recognized. It will, however, be shown that the [Greek: alpha] rays play a far more important part in radio-active processes than the [Greek: beta] rays, and that the greater portion of the energy emitted in the form of ionizing radiations is due to them.

88. The nature of the [Greek: alpha] rays. The nature of the [Greek: alpha] rays was difficult to determine, for a magnetic field sufficient to cause considerable deviation of the [Greek: beta] rays produced no appreciable effect on the [Greek: alpha] rays. It was suggested by several observers that they were, in reality, secondary rays set up by the [Greek: beta] or cathode rays in the active matter from which they were produced. Such a view, however, failed to explain the radio-activity of polonium, which gave out [Greek: alpha] rays only. Later work also showed that the matter, which gave rise to the [Greek: beta] rays from uranium, could be chemically separated from the uranium, while the intensity of the [Greek: alpha] rays was unaffected. These and other results show that the [Greek: alpha] and [Greek: beta] rays are produced quite independently of one another. The view that they are an easily absorbed type of Röntgen rays fails to explain a characteristic property of the [Greek: alpha] rays, viz. that the absorption of the rays in a given thickness of matter, determined by the electrical method, increases with the thickness of matter previously traversed. It does not seem probable that such an effect could be produced by a radiation like X rays, but the result is to be expected if the rays consist of projected bodies, which fail to ionize the gas when their velocity is reduced below a certain value. From observations of the relative ionization produced in gases by the [Greek: alpha] and [Greek: beta] rays, Strutt^[28] suggested in 1901 that the [Greek: alpha] rays might consist of positively charged bodies projected with great velocity. Sir William Crookes^[29], in 1902, advanced the same hypothesis. From a study of the [Greek: alpha] rays of polonium Mme. Curie^[30] in 1900 suggested the probability that these rays consisted of bodies, projected with great velocity, which lost their energy by passing through matter.

The writer was led independently to the same view by a mass of indirect evidence which received an explanation only on the hypothesis that the rays consisted of matter projected with great velocity. Preliminary experiments with radium of activity 1000 showed that it was very difficult to determine the magnetic deviation of the [Greek: alpha] rays. When the rays were passed through slits sufficiently narrow to enable a minute deviation of the rays to be detected, the ionizing effect of the issuing rays was too small to be measured with certainty. It was not until radium of activity 19,000 was obtained that it was possible to detect the deviation of these rays in an intense magnetic field. How small the magnetic deviation is may be judged from the fact that the [Greek: alpha] rays, projected at right angles to a magnetic field of 10,000 C.G.S. units, describe the arc of a circle of about 39 cms. radius, while under the same conditions the cathode rays produced in a vacuum tube would describe a circle of about ·01 cm. radius. It is therefore not surprising that the [Greek: alpha] rays were for some time thought to be non-deviable in a magnetic field.

89. Magnetic deviation of the [Greek: alpha] rays. The general method employed^[31] to detect the magnetic deviation of the [Greek: alpha] rays was to allow the rays to pass through narrow slits and to observe whether the rate of discharge of an electroscope, due to the issuing rays, was altered by the application of a strong magnetic field. Fig. 32 shows the general arrangement of the experiment. The rays from a thin layer of radium of activity 19,000 passed upwards through a number of narrow slits G, in parallel, and then through a thin layer of aluminium foil, ·00034 cm. thick, into the testing vessel V. The ionization produced by the rays in the testing vessel was measured by the rate of movement of the leaves of a gold-leaf electroscope B. The gold-leaf system was insulated inside the vessel by a sulphur bead C, and could be charged by means of a movable wire D, which was afterwards earthed. The rate of movement of the gold-leaf was observed through small mica windows in the testing vessel by means of a microscope provided with a micrometer eye-piece.

Fig. 32.

In order to increase the ionization in the testing vessel, the rays passed through 20 to 25 slits of equal width, placed side by side. This was arranged by cutting grooves at regular intervals in side-plates into which brass plates were slipped. The width of the slit varied in different experiments between ·042 cm. and ·1 cm. The magnetic field was applied perpendicular to the plane of the paper, and parallel to the plane of the slits. The rays are thus deflected in a direction perpendicular to the plane of the slits and a very small amount of deviation is sufficient to cause the rays to impinge on the sides of the plate where they are absorbed.

The testing vessel and system of plates were waxed to a lead plate P so that the rays entered the vessel V only through the aluminium foil. It is necessary in these experiments to have a steady stream of gas passing downwards between the plates in order to prevent the diffusion of the emanation from the radium upwards into the testing vessel. The presence in the testing vessel of a small amount of this emanation, which is always given out by radium, would produce great ionization and completely mask the effect to be observed. For this purpose, a steady current of dry electrolytic hydrogen of about 2 c.c. per second was passed into the testing vessel; it then streamed through the porous aluminium foil, and passed between the plates carrying the emanation with it away from the apparatus. The use of a stream of hydrogen instead of air greatly simplifies the experiment, for it increases the ionization current due to the [Greek: alpha] rays in the testing vessel, and at the same time greatly diminishes that due to the [Greek: beta] and [Greek: gamma] rays. This is caused by the fact that the [Greek: alpha] rays are much more readily absorbed in air than in hydrogen, while the rate of production of ions due to the [Greek: beta] and [Greek: gamma] rays is much less in hydrogen than in air. The intensity of the [Greek: alpha] rays after passing between the plates is consequently greater when hydrogen is used; and since the rays pass through a sufficient distance of hydrogen in the testing vessel to be largely absorbed, the total amount of ionization produced by them is greater with hydrogen than with air.

The following is an example of an observation on the magnetic deviation:—

Pole-pieces 1·90 × 2·50 cms.

Strength of field between pole-pieces 8370 units.

Apparatus of 25 parallel plates of length 3·70 cms., width ·70 cm., with an average air-space between plates of ·042 cm.

Distance of radium below plates 1·4 cm.

                                            Rate of discharge
                                           of electroscope in
                                            volts per minute
(1) Without magnetic field 8·33
(2) With magnetic field 1·72
(3) Radium covered with thin layer of mica to
      absorb all [Greek: alpha] rays 0·93
(4) Radium covered with mica and magnetic field
      applied 0·92

The mica plate, ·01 cm. thick, was of sufficient thickness to

absorb completely all the α rays, while it allowed the β rays and γ rays to pass through without appreciable absorption. The difference between (1) and (3), 7·40 volts per minute, gives the rate of discharge due to the [Greek: alpha] rays alone; the difference between (2) and (3), 0·79 volts per minute, that due to the α rays not deviated by the magnetic field employed.

The amount of [Greek: alpha] rays not deviated by the field is thus about 11% of the total. The small difference between (3) and (4) measures the small ionization due to the [Greek: beta] rays, for they would be completely deviated by the magnetic field; (4) comprises the effect of the [Greek: gamma] rays together with the natural leak of the electroscope in hydrogen.

In this experiment there was a good deal of stray magnetic field acting on the rays before they reached the pole-pieces. The diminution of the rate of discharge due to the [Greek: alpha] rays was found to be proportional to the strength of field between the pole-pieces. With a more powerful magnetic field, the whole of the α rays were deviated, showing that they consisted entirely of projected charged particles.

In order to determine the direction of deviation of the rays, the rays were passed through slits one mm. in width, each of which was half covered with a brass strip. The diminution of the rate of discharge in the testing vessel for a given magnetic field in such a case depends upon the direction of the field. In this way it was found that the rays were deviated in the opposite sense to the cathode rays. Since the latter consist of negatively charged particles, the [Greek: alpha] rays must consist of positively charged particles.

These results were soon after confirmed by Becquerel^[32], by the photographic method, which is very well adapted to determine the character of the path of the rays acted on by a magnetic field. The radium was placed in a linear groove cut in a small block of lead. Above this source, at a distance of about 1 centimetre, was placed a metallic screen, formed of two plates, leaving between them a narrow opening parallel to the groove. Above this was placed the photographic plate. The whole apparatus was placed in a strong magnetic field parallel to the groove. The strength of the magnetic field was sufficient to deflect the β rays completely away from the plate. When the plate was parallel to the opening, there was produced on it an impression, due to the α rays alone, which became more and more diffuse as the distance from the opening increased. This distance should not exceed 1 or 2 centimetres on account of the absorption of the rays in air. If, during the exposure, the magnetic field is reversed for equal lengths of time, on developing the plate two images of the α rays are observed which are deflected in opposite directions. This deviation, even in a strong field, is small though quite appreciable and is opposite in sense to the deviation observed for the β or cathodic rays from the same material.

M. Becquerel^[33], by the same method, found that the α rays from polonium were deviated in the same direction as the α rays from radium; and thus that they also consist of projected positive bodies. In both cases, the photographic impressions were sharply marked and did not show the same diffusion which always appears in photographs of the β rays.

90. Electrostatic deviation of the α rays. If the rays are charged bodies, they should be deflected in passing through a strong electric field. This was found by the writer to be the case, but the electric deviation is still more difficult to detect than the magnetic deviation, as the intensity of the electric field must of necessity be less than that required to produce a spark in the presence of radium. The apparatus was similar to that employed for the magnetic deviation (Fig. 32) with this exception, that the brass sides which held the plates in position, were replaced by ebonite. Alternate plates were connected together and charged to a high potential by means of a battery of small accumulators. The discharge in the electroscope, due to the α rays, was found to be diminished by application of the electric field. With plates ·055 cm. apart and 4·5 cms. high, the diminution was only 7% with a P.D. of 600 volts between the slits. With a special arrangement of plates, with slits only ·01 cm. apart, the discharge was diminished about 45% with an electric field corresponding to 10,000 volts per cm.

91. Determination of the constants of the rays. If the deviation of the rays in both an electric and magnetic field is known, the values of the velocity of the rays, and the ratio e/m of the charge of the particle to its mass can be determined by the method, first used by J. J. Thomson for the cathode rays, which is described in section 50. From the equations of a moving charged body, the radius of curvature [Greek: rho] of the path of the rays in a magnetic field of strength H perpendicular to the path of the rays is given by

H[Greek: rho] = (m/e)V.

If the particle, after passing through a uniform magnetic field for a distance l_{1}, is deviated through a small distance d_{1} from its original direction,

        2[Greek: rho]d_{1} = l_{1}^2
or d_{1} = (l_{1}^2/2) (e/m) (H/V) (1).

If the rays pass through a uniform electric field of strength X and length l_{2} with a deviation d_{2},

d_{2} = (1/2) (Xel_{2}^2})/(mV^2) (2),

since Xe/m is the acceleration of the particle, at right angles to its direction, and l_{2}/V is the time required to travel through the electric field.

From equations (1) and (2)

         V = (d_{1}/d_{2}) (l_{2}^2/l_{1}^2) (X/H),
and e/m = (2d_{1}/l_{1}^2) (V/H).

The values of V and e/m are thus completely determined from the combined results of the electric and magnetic deviation. It was found that

V = 2·5 × 10^9 cms. per sec.
e/m = 6 × 10^3.

On account of the difficulty of obtaining a large electrostatic deviation, these values are only approximate in character. The results on the magnetic and electric deviation of the [Greek: alpha] rays of radium have been confirmed by Des Coudres^[34], by the photographic method. Some pure radium bromide was used as a source of radiation. The whole apparatus was enclosed in a vessel which was exhausted to a low vacuum. In this way, not only was he able to determine the photographic action of the rays at a much greater distance from the source, but he was also able to apply a stronger electric field without the passage of a spark. He found values of the constants given by

V = 1·65 × 10^9 cms. per sec.
e/m = 6·4 × 10^3.

These values are in very good agreement with the numbers found by the electric method. The [Greek: alpha] rays from radium are complex, and probably consist of a stream of positively charged bodies projected at velocities lying between certain limits. The amount of deviation of the particles in a magnetic field will thus differ according to the velocity of the particle. The photographic results of Becquerel seem to indicate that the velocity of the rays of radium can vary only within fairly narrow limits, since the trajectory of the rays in a magnetic field is sharply marked and not nearly as diffuse as in similar experiments with the [Greek: beta] rays. The evidence, however, discussed in the following section, shows that the velocities of the [Greek: alpha] particles from a thick layer of radium vary over a considerable range.

92. Becquerel^[35] has examined the amount of magnetic deviation of the [Greek: alpha] rays at different distances from the source of the rays in a very simple way. A narrow vertical pencil of the rays, after its passage through a narrow slit, fell on a photographic plate, which was inclined at a small angle to the vertical and had its lower edge perpendicular to the slit. The trajectory of the rays is shown by a fine line traced on the plate. If a strong magnetic field is applied parallel to the slit, the trajectory of the rays is displaced to the right or left according to the direction of the field. If equal times of exposure are given for the magnetic field in the two directions, on developing the plate two fine diverging lines are found traced on the plate. The distance between these lines at any point is a measure of twice the average deviation at that point, corresponding to the value of the magnetic field. By measuring the distance between the trajectories at various points, Becquerel found that the radius of curvature of the path of the rays increased with the distance from the slit. The product H[Greek: rho] of the strength of the field and the radius of curvature of the path of the rays is shown in the following table.

Distance in mms.
 from the slit H[Greek: rho]
       1 2·91 × 10^5
       3 2·99 "
       5 3·06 "
       7 3·15 "
       8 3·27 "
       9 3·41 "

The writer (loc. cit.) showed that the maximum value of H[Greek: rho] for complete deviation of the [Greek: alpha] rays was 390,000. The results are thus in good agreement. Since H[Greek: rho] = (m/e)V these results show that the values either of V or of e/m for the projected particles vary at different distances from the source. Becquerel considered that the rays were homogeneous, and, in order to explain the results, has suggested that the charge on the projected particles may gradually decrease with the distance traversed, so that the radius of curvature of the path steadily increases with the distance from the source. It, however, seems more probable that the rays consist of particles projected with different velocities, and that the slower particles are more quickly absorbed in the gas. In consequence of this, only the swifter particles are present some distance from the source.

This conclusion is borne out by some recent experiments of Bragg and Kleeman^[36] on the nature of the absorption of [Greek: alpha] particles by matter, which are discussed in more detail in sections 103 and 104. They found that the [Greek: alpha] particles from a thick layer of radium are complex, and have a wide range of penetrating power and presumably of velocity. This is due to the fact that the [Greek: alpha] particles emitted from the radium come from different depths. Since their velocity is reduced in their transit through matter, a pencil of [Greek: alpha] rays will consist of particles which differ considerably in speed. Those which are just able to emerge from the radium will be absorbed in a very short depth of air, while those that come from the surface will be able to pass through several centimetres of air before they lose their power of ionizing the gas. Since the [Greek: alpha] particles have different velocities, they will be unequally deflected by the magnetic field, the slower moving particles describing a more curved path than the swifter ones. Consequently, the outer edge of the trace of the pencil of rays on the photographic plate, as obtained by Becquerel, will be the locus of the points where the photographic action of the [Greek: alpha] particles end. It was found that the [Greek: alpha] particles are most efficient as ionizers of the gas just before their power of ionizing ends. The loss of ionizing power of the [Greek: alpha] particles seems to be fairly abrupt, and, for particles of the same velocity, to occur always after traversing a definite distance in air. On the assumption that the photographic as well as the ionizing action is most intense just before the particles are stopped, and ceases fairly abruptly, Bragg has been able to account numerically for the measurements (see above table) recorded by Becquerel. Quite apart from the special assumptions required for such a quantitative comparison of theory with experiment, there can be little doubt that the increase of value of H[Greek: rho] with distance can be satisfactorily explained as a consequence of the complex character of the pencil of rays^[37].

Becquerel states that the amount of deviation, in a given magnetic field, was the same for the [Greek: alpha] rays of polonium and of radium. This shows that the value of (m/e)V is the same for the [Greek: alpha] rays from the two substances. Since the [Greek: alpha] rays from polonium are far more readily absorbed than the [Greek: alpha] rays from radium, this result would indicate that the value of m/e is greater for the [Greek: alpha] particles of polonium than of radium. Further experimental evidence is required on this important point. 93. Charge carried by the [Greek: alpha] rays. We have seen that the negative charge carried by the [Greek: beta] particles has been readily measured. Since there is reason to believe (section 228) that four [Greek: alpha] particles are expelled from radium for each [Greek: beta] particle, it is to be expected that the positive charge carried by the [Greek: alpha] particles should be determined still more readily. All the initial experiments, however, made to detect this charge, gave negative results; and, before successful results were obtained, it was found necessary to eliminate some secondary actions, which at first completely masked the effects to be looked for.

In consequence of the importance of this question, a brief account will be given of the methods of measurement adopted and the special experimental difficulties which have arisen.

In the first place, it must be remembered that only a small fraction of the [Greek: alpha] rays, emitted from a layer of powdered radium bromide, escape into the surrounding gas. On account of the ease with which the [Greek: alpha] rays are stopped in their passage through matter, only those escape which are expelled from a superficial layer, and the rest are absorbed by the radium itself. On the other hand, a much larger proportion of the [Greek: beta] rays escape, on account of their greater power of penetration. In the second place, the [Greek: alpha] particle is a far more efficient ionizer of the gas than the [Greek: beta] particle, and, in consequence, if the charge carried by the [Greek: alpha] rays is to be determined by methods similar to those employed for the [Greek: beta] rays (see section 80), the pressure of the gas surrounding the conductor to be charged must be very small in order to eliminate, as far as possible, the loss of charge resulting from the ionization of the residual gas by the [Greek: alpha] rays^[38].

The experimental arrangement used by the writer is shown in Fig. 33.

A thin film of radium was obtained on a plate A by evaporation of a radium solution containing a known weight of radium bromide. Some hours after evaporation, the activity of the radium, measured by the [Greek: alpha] rays, is about 25 per cent. of its maximum value, and the [Greek: beta] rays are almost completely absent. The activity measured by the [Greek: alpha] and [Greek: beta] rays is then slowly regained, and recovers its original value after about a month's interval (see chapter XI.). The experiments were made on the active plate when its activity was a minimum, in order to avoid complications due to the presence of [Greek: beta] rays. The film of radium was so thin that only a very small fraction of the [Greek: alpha] rays was absorbed.

Fig. 33.

The active plate A was insulated in a metal vessel D, and was connected to one pole of the battery, the other pole being earthed. The upper electrode, which was insulated and connected with a Dolezalek electrometer, consisted of a rectangular copper vessel BC, the lower part of which was covered with a thin sheet of aluminium foil. The [Greek: alpha] rays passed through the foil, but were stopped by the copper sides of the vessel. This arrangement was found to reduce the secondary ionization produced at the surface of the upper plate. The outside vessel D could be connected with either A or B or with earth. By means of a mercury pump, the vessel was exhausted to a very low pressure. If the rays carry a positive charge, the current between the two plates measured by the electrometer should be greater when A is charged positively. No certain difference, however, between the currents in the two directions was observed, even when a very good vacuum was obtained. In some arrangements, it was found that the current was even greater when the lower plate was negative than when it was positive. An unexpected experimental result was also noticed. The current between the parallel plates at first diminished with the pressure, but soon reached a limiting value which was not altered however good a vacuum was produced. For example, in one experiment, the current between the two parallel plates, placed about 3 mms. apart, was initially 6·5 × 10^{-9} amperes and fell off directly as the pressure. The current reached a limiting value of about 6 × 10^{-12} amperes, or about 1/1000 of the value at atmospheric pressure. The magnitude of this limiting current was not much altered if the air was replaced by hydrogen.

Experiments of a similar character have been made by Strutt^[39] and J. J. Thomson^[40]; using an active bismuth plate coated with radio-tellurium (polonium) after Marckwald's method. This substance emits only [Greek: alpha] rays, and is thus especially suitable for experiments of this kind. Strutt employed the method used by him to show the charge carried by the [Greek: beta] rays (Fig. 27). He found, however, that, even in the lowest possible vacuum, the electroscope rapidly lost its charge and at the same rate whether it was charged positively or negatively. This is in agreement with the results found by the writer with radium.

In the experiments of J. J. Thomson, the electroscope was attached to a metal disc placed 3 cms. from the plate of radio-*tellurium. A very low vacuum was produced by Dewar's method by absorbing the residual gas in cocoanut charcoal immersed in liquid air. When the electroscope was charged negatively, an extremely slow rate of leak was observed, but when charged positively the leak was about 100 times greater. This showed that the polonium gave out large quantities of negative electricity, but not enough positive to be detected. By placing the apparatus in a strong magnetic field, the negative particles were prevented from reaching the electroscope and the positive leak was stopped.

These results indicate that these negative particles are not projected with sufficient velocity to move against the repulsion exerted by the electrified body, and are bent by a magnetic field. There thus seems little doubt that a stream of negative particles (electrons) is projected from the active surface at a very slow speed. Such low velocity electrons are also projected from uranium and radium. It is probable that these electrons are a type of secondary radiation, set up at the surfaces on which the [Greek: alpha] rays fall. The particles would be extremely readily absorbed in the gas, and their presence would be difficult to detect except in low vacua. J. J. Thomson at first obtained no evidence that the [Greek: alpha] particles of polonium were charged; but in later experiments, where the plates were closer together, the electro-*scope indicated that the [Greek: alpha] rays did carry a positive charge.

In order to see whether the positive charge due to the [Greek: alpha] rays from radium could be detected when the slow moving ions were prevented from escaping by a magnetic field, I placed the apparatus of Fig. 33 between the pole-pieces of a large electro-*magnet, so that the magnetic field was parallel to the plane of the plates^[41]. A very marked alteration was observed both on the magnitude of the positive and negative currents. In a good vacuum, the upper plate received a positive charge, independently of whether the lower plate was charged positively or negatively or was connected with earth. After the magnetic field had reached a certain value, a great increase in its strength had no appreciable effect on the magnitude of the current.

The following table illustrates the results obtained when the two plates were 3 mms. apart, and were both coated with thin aluminium foil.

                   Current in arbitrary units
                /———————/\——————-\
Potential of Without magnetic With magnetic
lower plate field field
  0 — +·36
 +2 volts 2·0 +·46}
                                            } ·39
 -2 " 2·5 +·33}

 +4 " 2·8 +·47}
                                            } ·41
 -4 " 3·5 +·35}

 +8 " 3·1 +·56}
                                            } ·43
 -8 " 4·0 +·31}

 +84 " 3·5 +·77}
                                            } ·50
 -84 " 5·2 +·24}

Let n be the number of [Greek: alpha] particles, carrying a charge e, which are absorbed in the upper plate. Let [Greek: iota]_{0}, see next page] be the current due to the slight ionization of the residual gas.

If only a small potential is applied to the lower plate, this current should be equal in magnitude but opposite in sign when the potential is reversed. Let [Greek: iota]_{1} be the charge per sec. communicated to the upper electrode when the lower plate is charged positively and ι_{2} the value when charged negatively. Then

                [Greek: iota]_{1} = [Greek: iota]_{0} + ne,
                [Greek: iota]_{2} = [Greek: iota]_{0} + ne;

adding we get ne = ([Greek: iota]_{1} + [Greek: iota]_{2})/2._{0} in one of the above two equations]

Now in the third column of the above table it is seen that ([Greek: iota]_{1} + [Greek: iota]_{2})/2 has the values ·39, ·41, ·43 for 2, 4, and 8 volts respectively. The numbers are thus in fairly good agreement. Similar results were obtained when a brass plate was substituted for the upper electrode shown in the figure. Taking into consideration that the magnitude of ne is independent of the strength of the magnetic field above a certain small value, and the good agreement of the numbers obtained with variation of voltage, I think that there can be no doubt that the positive charge communicated to the upper electrode was carried by the [Greek: alpha] particles. This positive charge was not small, for using a weight of ·48 mgrs. radium bromide spread in a thin foil over an area of about 20 sq. cms., the charge communicated by the particles corresponded to a current 8·8 × 10^{-13} amperes, and, with the Dolezalek electrometer employed, it was necessary to add a capacity of ·0024 microfarads to the electrometer system.

In these experiments, the film of radium bromide was so thin, that only a very small percentage of the [Greek: alpha] particles was stopped by the radium itself. Assuming that each [Greek: alpha] particle carries the same charge as an ion, viz. 1·1 × 10^{-19} coulombs, and remembering that half of the [Greek: alpha] particles are absorbed in the lower plate, the total number N of [Greek: alpha] particles expelled per second from one gram of radium bromide (at its minimum activity) can be deduced. In two separate experiments where the amount of radium used was ·194 and ·484 mgrs. respectively, the values of N were in close agreement and equal to 3·6 × 10^{10}. Now it will be shown later that in radium there are three other products in radio-active equilibrium, each of which probably gives out the same number of [Greek: alpha] particles as radium itself. If this is the case, the total number of [Greek: alpha] particles expelled per second from 1 gram of radium bromide in radio-active equilibrium is 4N or 1·44 × 10^{11}. Assuming the composition of radium bromide as RaBr_{2}, the number per second per gram of radium is 2·5 × 10^{10}. This number will be found to be in very good agreement with that deduced from indirect data (chapter XIII.). The value of N is of great importance in determining the magnitude of various quantities in radio-active calculations.

94. Mass and energy of the α particle. It has been pointed out that the α rays from radium and polonium are analogous to the Canal rays of Goldstein, for both carry a positive charge and are difficult to deflect by a magnetic field. The experiments of Wien have shown that the velocity of projection of the canal rays varies with the gas in the tube and the intensity of the electric field applied, but it is generally about 1/10 of the velocity of the α particle from radium. The value of e/m is also variable, depending upon the gas in the tube.

It has been shown that for the α rays of radium

V = 2·5 × 10^9 and e/m = 6 × 10^3.

Now the value of e/m for the hydrogen atom, liberated in the electrolysis of water, is 10^4. Assuming the charge carried by the α particle to be the same as that carried by the hydrogen atom, the mass of the α particle is about twice that of the hydrogen atom. Taking into consideration the uncertainty attaching to the experimental value of e/m for the α particle, if the α particle consists of any known kind of matter, this result indicates that it consists either of projected helium or hydrogen. Further evidence on this important question is given in section 261.

The α rays from all the radio-active substances and their products, such as the radio-active emanations and the matter causing excited activity, possess the same general properties and do not vary very much in penetrating power. It is thus probable that in all cases the α rays from the different radio-active substances consist of positively charged bodies projected with great velocity. Since the rays from radium are made up in part of α rays from the emanation stored in the radium, and from the excited activity which it produces, the α rays from each of these products must consist of positively charged bodies; for it has been shown that all the α rays from radium are deviated in a strong magnetic field. The kinetic energy of each projected particle is enormous, compared with its mass. The kinetic energy of each [Greek: alpha] particle is

(1/2)mV^2 = (1/2)(m/e)V^2e = 5·9 × 10^{-6} ergs.

Taking the velocity of a rifle bullet as 10^5 cms. per second, it is seen that, mass for mass, the energy of motion of the [Greek: alpha] rays is 6 × 10^8 times as great as that of the rifle bullet. In this projection of bodies atomic in size with great velocity probably lies the principal cause of the heating effects produced by radium (chapter XII).

95. Atomic disintegration. The radio-activity of the radio-*elements is an atomic and not a molecular property. The rate of emission of the radiations depends only on the amount of the element present and is independent of its combination with inactive substances. In addition, it will be shown later that the rate of emission is not affected by wide variations of temperature, or by the application of any known chemical or physical forces. Since the power of radiating is a property of the radio-atoms, and the radiations consist for the most part of positively and negatively charged masses projected with great velocity, it is necessary to suppose that the atoms of the radio-elements are undergoing disintegration, in the course of which parts of the atom escape from the atomic system. It seems very improbable that the [Greek: alpha] and β particles can suddenly acquire their enormous velocity of projection by the action of forces existing inside or outside the atom. For example, the [Greek: alpha] particle would have to travel from rest between two points differing in potential by 5·2 million volts in order to acquire the kinetic energy with which it escapes. Thus it seems probable that these particles are not set suddenly in motion, but that they escape from an atomic system in which they were already in rapid oscillatory or orbital motion. On this view, the energy is not communicated to the projected particles, but exists beforehand in the atoms from which they escape. The idea that the atom is a complicated structure consisting of charged parts in rapid oscillatory or orbital motion has been developed by J. J. Thomson, Larmor and Lorentz. Since the [Greek: alpha] particle is atomic in size, it is natural to suppose that the atoms of the radio-active elements consist not only of the electrons in motion, but also of positively charged particles whose mass is about the same as that of the hydrogen or helium atom.

It will be shown later that only a minute fraction of the atoms of the radio-element need break up per second in order to account for the radiations even of an enormously active element like radium. The question of the possible causes which lead to this atomic disintegration and the consequences which follow from it will be discussed later in chapter XIII.

96. Experiments with a zinc sulphide screen. A screen of Sidot's hexagonal blend (phosphorescent crystalline zinc sulphide) lights up brightly under the action of the [Greek: alpha] rays of radium and polonium. If the surface of the screen is examined with a magnifying glass, the light from the screen is found not to be uniformly distributed but to consist of a number of scintillating points of light. No two flashes succeed one another at the same point, but they are scattered over the surface, coming and going rapidly without any movement of translation. This remarkable action of the radium and polonium rays on a zinc sulphide screen was discovered by Sir William Crookes^[42], and independently by Elster and Geitel^[43], who observed it with the rays given out from a wire which has been charged negatively either in the open air or in a vessel containing the emanation of thorium.

In order to show the scintillations of radium on the screen, Sir William Crookes has devised a simple apparatus which he has called the "Spinthariscope." A small piece of metal, which has been dipped in a radium solution, is placed several millimetres away from a small zinc sulphide screen. This screen is fixed at one end of a short brass tube and is looked at through a lens fixed at the other end of the tube. Viewed in this way, the surface of the screen is seen as a dark background, dotted with brilliant points of light which come and go with great rapidity. The number of points of light per unit area to be seen at one time falls off rapidly as the distance from the radium increases, and, at several centi-*

• metres distance, only an occasional one is seen. The experiment

is extremely beautiful, and brings vividly before the observer the idea that the radium is shooting out a stream of projectiles, the impact of each of which on the screen is marked by a flash of light.

The scintillating points of light on the screen are the result of the impact of the [Greek: alpha] particles on its surface. If the radium is covered with a layer of foil of sufficient thickness to absorb all the [Greek: alpha] rays the scintillations cease. There is still a phosphorescence to be observed on the screen due to the [Greek: beta] and [Greek: gamma] rays, but this luminosity is not marked by scintillations to any appreciable extent. Sir William Crookes showed that the number of scintillations was about the same in vacuo as in air at atmospheric pressure. If the screen was kept at a constant temperature, but the radium cooled down to the temperature of liquid air, no appreciable difference in the number of scintillations was observed. If, however, the screen was gradually cooled to the temperature of liquid air, the scintillations diminished in number and finally ceased altogether. This is due to the fact that the screen loses to a large extent its power of phosphorescence at such a low temperature.

Not only are scintillations produced by radium, actinium, and polonium, but also by the emanations and other radio-active products which emit [Greek: alpha] rays. In addition, F. H. Glew^[44] has found that they can be observed from the metal uranium, thorium compounds and various varieties of pitchblende. In order to show the scintillations produced by pitchblende, a flat surface was ground, and a transparent screen, whose lower surface was coated with zinc sulphide, placed upon it. Glew has designed a modified and very simple form of spinthariscope. A transparent screen, coated on one side with a thin layer of zinc sulphide, is placed in contact with the active material, and the scintillations observed by a lens in the usual way.

Since there is no absorption in the air, the luminosity is a maximum. The relative transparency of different substances placed between the active material and the screen may, in this way, be directly studied.

The production of scintillations appears to be a general property of the [Greek: alpha] rays from all radio-active substances. The scintillations are best shown with a zinc sulphide screen; but are also observed with willemite (zinc silicate), powdered diamond, and potassium platinocyanide (Glew, loc. cit.). If a screen of barium platinocyanide is exposed to the [Greek: alpha] rays from radium, the scintillations are difficult to observe, and the luminosity is far more persistent than for a zinc sulphide screen exposed under the same conditions. The duration of the phosphorescence in this case probably accounts for the absence of visible scintillations.

There can be no doubt that the scintillations result from the continuous bombardment of the sensitive screen by the [Greek: alpha] particles. Each of these particles moves with enormous velocity, and has a considerable energy of motion. On account of the ease with which these particles are stopped, most of this energy is given up at the surface of the screen, and a portion of the energy is in some way transformed into light. Zinc sulphide is very sensitive to mechanical shocks. Luminosity is observed if a penknife is drawn across the screen, or if a current of air is directed on to the screen. The disturbance effected by the impact of the [Greek: alpha] particle extends over a distance very large compared with the size of the impinging particle, so that the spots of light produced have an appreciable area. Recently Becquerel^[45] has made an examination of the scintillations produced by different substances, and has concluded that the scintillations are due to irregular cleavages in the crystals composing the screen, produced by the action of the [Greek: alpha] rays. Scintillations can be mechanically produced by crushing a crystal. Tommasina^[46] found that a zinc sulphide screen removed from the action of the radium rays for several days, showed the scintillations again when an electrified rod was brought near it.

The number of scintillations produced in zinc sulphide depends upon the presence of a slight amount of impurity and on its crystalline state. It can be shown that even with the most sensitive zinc sulphide screens, the number of scintillations is probably only a small fraction of the total number of [Greek: alpha] particles which fall upon it. It would appear that the crystals are in some way altered by the bombardment of the [Greek: alpha] particles, and that some of the crystals occasionally break up with emission of light^[47]. Although the scintillations from a particle of pure radium bromide are very numerous, they are not too numerous to be counted. Close to the radium, the luminosity is very bright, but by using a high power microscope the luminosity can still be shown to consist of scintillations. Since the number of scintillations probably bears no close relation to the number of [Greek: alpha] particles emitted, a determination of the number of scintillations would have no special physical significance. The relation between the number of [Greek: alpha] particles and the number of scintillations would probably be variable, depending greatly on the exact chemical composition of the sensitive substance and also upon its crystalline state.

97. Absorption of the [Greek: alpha] rays by matter. The [Greek: alpha] rays from the different radio-active substances can be distinguished from one another by the relative amounts of their absorption by gases or by thin screens of solid substances. When examined under the same conditions, the [Greek: alpha] rays from the active substances can be arranged in a definite order with reference to the amount of absorption in a given thickness of matter.

In order to test the amount of absorption of the [Greek: alpha] rays for different thicknesses of matter, an apparatus similar to that shown in Fig. 17, p. 98, was employed^[48]. A thin layer of the active material was spread uniformly over an area of about 30 sq. cms., and the saturation current observed between two plates 3·5 cms. apart. With a thin layer^[49] of active material, the ionization between the plates is due almost entirely to the [Greek: alpha] rays. The ionization due to the [Greek: beta] and [Greek: gamma] rays is generally less than 1% of the total.

The following table shows the variation of the saturation current between the plates due to the [Greek: alpha] rays from radium and polonium, with successive layers of aluminium foil interposed, each ·00034 cm. in thickness. In order to get rid of the ionization due to the [Greek: beta] rays from radium, the radium chloride employed was dissolved in water and evaporated. This renders the active compound, for the time, nearly free from [Greek: beta] rays. The initial current with 1 layer of aluminium over the active material is taken as 100. It will be observed that the current due

          Polonium. Radium.
+————-+———-+——————+ +————-+———-+——————+
|Layers of|Current| Ratio of | |Layers of|Current| Ratio of |
|aluminium| |decrease for| |aluminium| |decrease for|
| | | each layer | | | | each layer |
+————-+———-+——————+ +————-+———-+——————+
| 0 |100 | | | 0 | 100 | |
| | | ·41 | | | | ·48 |
| 1 | 41 | | | 1 | 48 | |
| | | ·31 | | | | ·48 |
| 2 | 12·6 | | | 2 | 23 | |
| | | ·17 | | | | ·60 |
| 3 | 2·1 | | | 3 | 13·6 | |
| | | ·067 | | | | ·47 |
| 4 | ·14 | | | 4 | 6·4 | |
| | | | | | | ·39 |
| 5 | 0 | | | 5 | 2·5 | |
| | | | | | | ·36 |
| | | | | 6 | ·9 | |
| | | | | | | |
| | | | | 7 | 0 | |
+————-+———-+——————+ +————-+———-+——————+

to the radium rays decreases very nearly by half its value for each additional thickness until the current is reduced to about 6% of the maximum. It then decays more rapidly to zero. Thus, for radium, over a wide range, the current decreases approximately according to an exponential law with the thickness of the screen,

or i/i_{0} = e^{-[Greek: lambda]d},

where i is the current for a thickness d, and i_{0} the initial current. In the case of polonium, the decrease is far more rapid than would be indicated by the exponential law. By the first layer, the current is reduced to the ratio ·41. The addition of the third layer cuts the current down to a ratio of ·17. For most of the active bodies, the current diminishes slightly faster than the exponential law would lead one to expect, especially when the radiation is nearly all absorbed.

98. The increase of absorption of the [Greek: alpha] rays of polonium with the thickness of matter traversed has been very clearly shown in some experiments made by Mme Curie. The apparatus employed is shown in Fig. 34. The saturation current was measured between two parallel plates PP´ 3 cms. apart. The polonium A was placed in the metal box CC, and the rays from it, after passing through an opening in the lower plate P´, covered with a layer of thin foil T, ionized the gas between the plates. For a certain distance AT, of 4 cms. or more, no appreciable current was observed between P and P´. As the distance AT was diminished, the current increased in a very sudden manner, so that for a small variation of the distance AT there was a large increase of current. With still further decrease of distance the current increases in a more regular manner. The results are shown in the following table, where the screen T consisted of one and two layers of aluminium foil respectively. The current due to the rays, without the aluminium screen, is in each case taken as 100.

Fig. 34.

+———————————————————+—-+—-+—-+——+—-+
| Distance AT in cms. |3·5|2·5|1·9|1·45|0·5|
+———————————————————+—-+—-+—-+——+—-+
|For 100 rays transmitted by one layer | 0 | 0 | 5 | 10 |25 |
|For 100 rays transmitted by two layers| 0 | 0 | 0 | 0 |0·7|
+———————————————————+—-+—-+—-+——+—-+

The metallic screen thus cuts off a greater proportion of the rays the greater the distance of air which the radiations traverse. The effects are still more marked if the plates PP´ are close together. Results similar but not so marked are found if radium is substituted for the polonium.

It follows from these experiments that the ionization per unit volume, due to a large plate uniformly covered with the radio-active matter, falls off rapidly with the distance from the plate. At a distance of 10 cms. the [Greek: alpha] rays from uranium, thorium, or radium have been completely absorbed in the gas, and the small ionization then observed in the gas is due to the more penetrating [Greek: beta] and [Greek: gamma] rays. The relative amount of the ionization observed at a distance from the source will increase with the thickness of the layer of active matter, but will reach a maximum for a layer of a certain thickness. The greater proportion of the ionization, due to unscreened active matter, is thus entirely confined to a shell of air surrounding it not more than 10 cms. in depth.

Fig. 35.

99. The α rays from different compounds of the same active element, although differing in amount, have about the same average penetrating power. Experiments on this point have been made by the writer^[50] and by Owens^[51]. Thus in comparing the relative power of penetration of the α rays from the different radio-elements, it is only necessary to determine the penetrating power for one compound of each of the radio-elements. Rutherford and Miss Brooks^[52] have determined the amount of absorption of the α rays from the different active substances in their passage through successive layers of aluminium foil ·00034 cm. thick. The curves of absorption are given in Fig. 35. For the purpose of comparison in each case, the initial current with the bare active compound was taken as 100. A very thin layer of the active substance was used, and, in the case of thorium and radium, the emanations given off were removed by a slow current of air through the testing vessel. A potential difference of 300 volts was applied between the plates, which was sufficient to give the maximum current in each case.

Curves for the minerals organite and thorite were very nearly the same as for thoria.

For comparison, the absorption curves of the excited radiations of thorium and radium are given, as well as the curve for the radio-elements uranium, thorium, radium, and polonium. The [Greek: alpha] radiations may be arranged in the following order, as regards their power of penetration, beginning with the most penetrating.

Thorium}
Radium } excited radiation.
Thorium.
Radium.
Polonium.
Uranium.

The same order is observed for all the absorbing substances examined, viz., aluminium, Dutch metal, tinfoil, paper, and air and other gases. The differences in the absorption of the [Greek: alpha] rays from the active bodies are thus considerable, and must be ascribed either to a difference of mass or of velocity of the [Greek: alpha] particles or to a variation in both these quantities.

Since the [Greek: alpha] rays differ either in mass or velocity, it follows that they cannot be ascribed to any single radio-active impurity common to all radio-active bodies.

100. Absorption of the [Greek: alpha] rays by gases. The [Greek: alpha] rays from the different radio-active substances are quickly absorbed in their passage through a few centimetres of air at atmospheric pressure and temperature. In consequence of this, the ionization of the air, due to the [Greek: alpha] rays, is greatest near the surface of the radiating body and falls off very rapidly with the distance (see section 98). A simple method of determining the absorption in gases is shown in Fig. 36. The maximum current is measured between two parallel plates A and B kept at a fixed distance of 2 cms. apart, and then moved by means of a screw to different distances from the radio-*active surface. The radiation from this active surface passed through a circular opening in the plate A, covered with thin aluminium foil, and was stopped by the upper plate. For observations on other gases besides air, and for examining the effect at different pressures, the apparatus is enclosed in an air-*tight cylinder.

Fig. 36.

If the radius of the active surface is large compared with the distance of the plate A from it, the intensity of the radiation is approximately uniform over the opening in the plate A, and falls off with the distance x traversed according to an exponential law. Thus

I/I_{0} = e^{-[Greek: lambda]x},

where [Greek: lambda] is the "absorption constant" of the radiation for the gas under consideration^[53]. Let

x = distance of lower plate from active material,
l = distance between the two fixed plates.

The energy of the radiation at the lower plate is then I_{0}e^{-[Greek: lambda]x}, and at the upper plate I_{0}e^{-[Greek: lambda](l + x)}. The total number of ions produced between the parallel plates A and B is therefore proportional to

e^{-[Greek: lambda]x} - e^{-[Greek: lambda](l + x)} = e^{-[Greek: lambda]x}(1 - e^{-[Greek: lambda]l}).

Since the factor 1 - e^{-[Greek: lambda]l} is a constant, the saturation currentparticles coming from all points of the large radio-active layer, [Greek: lambda] is not the same as the coefficient of absorption of the rays from a point source. It will however be proportional to it. For this reason [Greek: lambda] is called the "absorption constant."] between A and B varies as e^{-[Greek: lambda]x}, i.e. it decreases according to an exponential law with the distance traversed.

Fig. 37.

The variation of the current between A and B with the distance from a thin layer of uranium oxide is shown in Fig. 37 for different gases. The initial measurements were taken at a distance of about 3·5 mms. from the active surface. The actual values of this initial current were different for the different gases, but, for the purposes of comparison, the value is in each case taken as unity.

It will be seen that the current falls off with the distance approximately in a geometrical progression, a result which is in agreement with the simple theory given above. The distance through which the rays pass before they are absorbed is given below for different gases.

                   Distance in mms. to
     Gas absorb half of radiation
Carbonic acid 3
Air 4·3
Coal-gas 7·5
Hydrogen 16

The results for hydrogen are only approximate, as the absorption is small over the distance examined. The absorption is least in hydrogen and greatest in carbonic acid, and follows the same order as the densities of the gases. In the case of air and carbonic acid, the absorption is proportional to the density, but this rule is widely departed from in the case of hydrogen. Results for the relative absorption by air of the [Greek: alpha] rays from the different active bodies are shown in Fig. 38.

Fig. 38.

The initial observation was made about 2 mms. from the active surface, and the initial current is in each case taken as 100. The current, as in the case of uranium, falls off at first approximately in geometrical progression with the distance. The thickness of air, through which the radiation passes before the intensity is reduced to half value, is given below.

                                           Distance in mms.
Uranium 4·3
Radium 7·5
Thorium 10
Excited radiation from Thorium and Radium 16·5

The order of absorption by air of the radiations from the active substances is the same as the order of absorption by the metals and solid substances examined. 101. Connection between absorption and density. Since in all cases the radiations first diminish approximately according to an exponential law with the distance traversed, the intensity I after passing through a thickness x is given by I = I_{0}e^{-λx} where λ is the absorption constant and I_{0} the initial intensity.

The following table shows the value of λ with different radiations for air and aluminium.

    Radiation λ for aluminium λ for air
Excited radiation 830 ·42
Thorium 1250 ·69
Radium 1600 ·90
Uranium 2750 1·6

Taking the density of air at 20° C. and 760 mms. as 0·00120 compared with water as unity, the following table shows the value of λ divided by density for the different radiations.

    Radiation Aluminium Air
Excited radiation 320 350
Thorium 480 550
Radium 620 740
Uranium 1060 1300

Comparing aluminium and air, the absorption is thus roughly proportional to the density for all the radiations. The divergence, however, between the absorption-density numbers is large when two metals like tin and aluminium are compared. The value of λ for tin is not much greater than for aluminium, although the density is nearly three times as great.

If the absorption is proportional to the density, the absorption in a gas should vary directly as the pressure, and this is found to be the case. Some results on this subject have been given by the writer (loc. cit.) for uranium rays between pressures of 1/4 and 1 atmosphere. Owens (loc. cit.) examined the absorption of the α radiation in air from thoria between the pressures of 0·5 to 3 atmospheres and found that the absorption varied directly as the pressure.

The variation of absorption with density for the projected positive particles is thus very similar to the law for the projected negative particles and for cathode rays. The absorption, in both cases, depends mainly on the density, but is not in all cases directly proportional to it. Since the absorption of the α rays in gases is probably mainly due to the exhaustion of the energy of the rays by the production of ions in the gas, it seems probable that the absorption in metals is due to a similar cause.

102. Relation between ionization and absorption in gases. It has been shown (section 45) that if the α rays are completely absorbed in a gas, the total ionization produced is about the same for all the gases examined. Since the rays are unequally absorbed in different gases, there should be a direct connection between the relative ionization and the relative absorption. This is seen to be the case if the results of Strutt (section 45) are compared with the relative absorption constants (section 100).

                Relative Relative
    Gas absorption ionization
Air 1 1
Hydrogen ·27 ·226
Carbon dioxide 1·43 1·53

Considering the difficulty of obtaining accurate determinations of the absorption, the relative ionization in a gas is seen to be directly proportional to the relative absorption within the limits of experimental error. This result shows that the energy absorbed in producing an ion is about the same in air, hydrogen, and carbon dioxide.

103. Mechanism of the absorption of α rays by matter. The experiments, already described, show that the ionization of the gas, due to the α rays from a large plane surface of radio-active matter, falls off in most cases approximately according to an exponential law, until most of the rays are absorbed, whereupon the ionization decreases at a much faster rate. In the case of polonium, the ionization falls off more rapidly than is to be expected on the simple exponential law.

The ionization produced in the gas is due to the collision of the rapidly moving α particles with the molecules of the gas in their path. On account of its large mass, the α particle is a far more efficient ionizer than the β particle moving at the same speed. It can be deduced from the results of experiment that each projected α particle is able to produce about 100,000 ions in passing through a few centimetres of the gas before its velocity is reduced to the limiting value, below which it no longer ionizes the gas in its path.

Energy is required to ionize the gas, and this energy can only be obtained at the expense of the kinetic energy of the projected α particle. Thus it is to be expected that the α particle should gradually lose its velocity and energy of motion in its passage through the gas.

Since the rate of absorption of the α rays in gases is deduced from measurements of the ionization of the gas at different distances from the source of radiation, a knowledge of the law of variation of the ionizing power of the projected α particle with its speed is required in order to interpret the results. The experimental data on this question are, however, too incomplete to be applied directly to a solution of this question. Townsend^[54] has shown that a moving electron produces ions in the gas after a certain limiting velocity is reached. The number of ions produced per centimetre of its path through the gas then rises to a maximum, and for still higher speeds continuously decreases. For example, Townsend found that the number of ions produced by an electron moving in an electric field was small at first for weak fields, but increased with the strength of the electric field to a maximum corresponding to the production of 20 ions per cm. of path in air at a pressure of 1 mm. of mercury. Durack^[55] found that the electrons, generated in a vacuum tube, moving with a velocity of about 5 × 10^9 cms. per second produced a pair of ions every 5 cms. of path at 1 mm. pressure. In a later paper, Durack showed that for the electrons from radium, which are projected with a velocity greater than half the velocity of light, a pair of ions was produced every 10 cms. of path. The high speed electron from radium is thus a very inefficient ionizer and produces only about 1/100 of the ionization per unit path observed by Townsend for the slow moving electron.

104. In the case of the α particle, no direct measurements have been made upon the variation of the ionization with the velocity of the particle, so that the law of absorption of the rays cannot be deduced directly. An indirect attack upon the question has, however, been made recently by Bragg and Kleeman^[56] who have formulated a simple theory to account for the experimental results which they have obtained upon the absorption of the α rays. The α particles from each simple type of radio-active matter are supposed to be projected with the same velocity, and to pass through a definite distance a in air at atmospheric pressure and temperature before they are all absorbed. As a first approximation the ionization per unit path is supposed to be the same over the whole length traversed before absorption, and to cease fairly suddenly at a definite distance from the source of radiation. This is in agreement with the observed fact that the ionization between parallel plates increases very rapidly when it approaches nearer than a certain distance to the radiant source. The range a depends upon the initial energy of motion of the α particle and will thus be different for different kinds of radio-active matter. If a thick layer of radio-active matter is employed, only the α particles from the surface have a range a. Those which reach the surface from a depth d have their range diminished by an amount ρd, where ρ is the density of the radio-active matter compared with air. This is merely an expression of the fact that the absorption of the α rays is proportional to the thickness and density of matter traversed. The rays from a thick layer of active matter will thus be complex, and will consist of particles of different velocity whose ranges have all values between 0 and a.

Suppose that a narrow pencil of α rays is emitted from a thick layer of radio-active material, and confined by metal stops as in Fig. 39.

Fig. 39.

The pencil of rays passes into an ionization vessel AB through a fine wire gauze A. The amount of ionization is to be determined between A and B for different distances h from the source of the rays R to the plate A. All the particles coming from a depth x of the material given by h = a - [Greek: rho]x will enter the ionization vessel. The number of ions produced in a depth dh of the ionization vessel is equal to nxdh, i.e. to n((a - h)/[Greek: rho])dh, where n is a constant.

If the depth of the ionization vessel be b, the total number of ions produced in the vessel is

[integral]_{h}^{h + b} (n((a - h)/[Greek: rho])dh) = (nb/[Greek: rho])(a - h - b/2).

This supposes that the stream of particles passes completely across the vessel. If not, the expression becomes

[integral]_{h}^a (n((a - h)/[Greek: rho])dh) = n(a - h)^2/(2[Greek: rho]).

If the ionization in the vessel AB is measured, and a curve plotted showing its relation to h, the curve in the former case should be a straight line whose slope is nb/[Greek: rho] and in the latter a parabola.

Thus if a thin layer of radio-active material is employed and a shallow ionization vessel, the ionization would be represented by a curve such as APM (Fig. 40), where the ordinates represent distances from the source of radiation, and the abscissae the ionization current between the plates AB.

Fig. 40.

In this case, PM is the range of the [Greek: alpha] particles from the lowest layer of the radio-active matter. The current should be constant for all distances less than PM.

For a thick layer of radio-active matter, the curve should be a straight line such as APB.

Curves of the above character should only be obtained when definite cones of rays are employed, and where the ionization vessel is shallow and includes the whole cone of rays. In such a case the inverse square law need not be taken into account. In the experiments previously recorded (sections 99 and 100), the ionization was measured between parallel plates several centimetres apart for a large area of radio-active material. Such an arrangement was necessary at the time at which the experiments were made, as only weak radio-active material was available. Measurable electrical effects could not then be obtained with narrow cones of rays and shallow ionization vessels, but this disadvantage is removed by the advent of pure radium bromide as a source of radiation.

The interesting experiments described by Bragg and Kleeman show that the theoretical curves are approximately realized in practice. The chief difficulty experienced in the analysis of the experimental results was due to the fact that radium is a complex radio-active substance and contains four radio-active products each of which gives rise to [Greek: alpha] rays which have different ranges. The general character of the results obtained from radium are shown graphically in Fig. 41, curves A, B, C, D.

Fig. 41.

The ordinates represent the distance between the radium and the gauze of the testing vessel; the abscissae the current in the ionization vessel in arbitrary units. Five milligrams of radium bromide were used, and the depth of the ionization vessel was about 5 mms. Curve A is for a cone of rays of angle 20°. The initial current at a distance of 7 cms. is due to the β and γ rays and natural leak. This curve is initially parabolic, and then is made up of two straight lines. Curve B is for a smaller cone, and shows the straight line character of the curve to within a short distance of the radium. Curve C was obtained under the same condition as curve A, but with a layer of gold beater's skin placed over the radium. The effect of this is to reduce all the ordinates of curve A by the same quantity. This is to be expected on the simple theory already considered. Curve D was obtained when the radium was heated so as to get rid of the emanation and its products. The [Greek: alpha] particles of greatest range are quite absent and the curve is simpler in character.

Fig. 42.

The complex character of the radium curves are more clearly brought out by a careful examination of a portion of the curve at distances between 2 and 5 cms. from the radium, using an ionization vessel of depth only 2 mms. The results are shown in Fig. 42, where the curve is seen to consist approximately of four straight lines of different slopes represented by PQ, QR, RS, ST.

Such a result is to be expected, for it will be shown later that four distinct [Greek: alpha] ray products exist in radium when in radio-active equilibrium. Each of these products of radium emits an equal number of [Greek: alpha] particles per second, but the range of each is different. If a_{1} is the range of one stream, a_{2} of another, the ionization in the vessel AB, when two streams enter the vessel, should be

(nb/ρ)(a_{1} - h - b/2) + (nb/ρ)(a_{2} - h - b/2), i.e. (nb/ρ)(a_{1} + a_{2} - 2h - b).

Thus the slope of the curve should in this case be 2nb/ρ, while if only one stream enters, it should be nb/ρ. When three reach it, the slope should be 3nb/ρ and for four 4nb/ρ. These results are realized fairly closely in practice. The curve (Fig. 42) consists of four parts, whose slopes are in the proportion 16, 34, 45, 65, i.e. very nearly in the ratio 1, 2, 3, 4.

Experiments were also made with very thin layers of radium bromide, when, as we have seen (Fig. 40) a very different shape of curve is to be expected. An example of the results is shown Fig. 43, curves I., II. and III. Curve I. is obtained from radium bromide which has been heated to drive off the emanation, and curves II. and III. from the same substance several days later, when the emanation was again accumulating. The portion PQ, which is absent in the first curve, is probably due to the "excited" activity produced by the emanation. By careful examination of the successive changes in the curves after the radium has been heated to drive off the emanation, it is possible to tell the range of the [Greek: alpha] rays from each of the different products, and this has been done to some extent by Bragg and Kleeman.

It will be seen later that the results here obtained support in a novel way the theory of radio-active changes which has been advanced from data of quite a different character.

The inward slope of the curve in Fig. 43 due to the radium indicates that the [Greek: alpha] particles become more efficient ionizers as their velocity decreases. This is in agreement with observations on the β rays. In some cases Bragg also observed that the [Greek: alpha] particles are the most efficient ionizers just before they lose their power of ionizing the gas.

Fig. 43.

Thus we may conclude from these experiments that the [Greek: alpha] particles from a simple radio-active substance traverse a definite distance in air, at a definite pressure and temperature, and that the ionization ends fairly abruptly. If the rays traverse a sheet of metal, the effective range of ionization is diminished by a distance corresponding to ρd, where ρ is the density of the material compared with air and d its thickness. The [Greek: alpha] rays from a thick layer of a simple radio-active substance consist of [Greek: alpha] particles of different velocities, which have ranges in air lying between 0 and the maximum range. The ionization of the particles per unit path is greatest near the end of its range, and decreases somewhat as we approach the radiant source. A complex source of rays like radium gives out four types of rays, each of which has a different but distinct range.

From this theory it is possible to calculate approximately the decrease of current to be observed when sheets of metal foil are placed over a large area of radio-active substance. This is the method that has been employed to obtain the curves of Figs. 35 and 38.

Suppose a very thin layer of simple radio-active matter is employed (for example a bismuth plate covered with radio-tellurium or a metal plate made active by exposure to the presence of the thorium or radium emanations) and that the ionization vessel is of sufficient depth to absorb the [Greek: alpha] rays completely.

Let d be the thickness of the metal plate, ρ its density compared with air. Consider a point P close to the upper side of the plate. The range of the particles moving from a point, when the path makes an angle θ with the normal at P, is a - ρd sec θ, where a is the range in air. The rays coming from points such that the paths make an angle with the normal greater than cos^{-1} (ρd/a) will thus be absorbed in the plate. By integrating over the circular area under the point P, it is easy to show that the total ionization in the vessel is proportional to

[integral]_{0}^{cos^{-1} (ρd/a)} 2π sin θ cos θ(a - ρd sec θ) dθ = π(a - ρd)^2/a.

The curves showing the relation between current and distance of metal traversed should thus be parabolic with respect to d. This is approximately the case for a simple substance like radio-tellurium. The curve for a thick layer of radium would be difficult to calculate on account of the complexity of the rays, but we know from experiment that it is approximately exponential. An account of some recent investigations made to determine the range of velocity over which the [Greek: alpha] particle is able to ionize the gas is given in Appendix A. The results there given strongly support the theory of absorption of the [Greek: alpha] rays discussed above. PART IV.

The γ or very penetrating Rays.

105. In addition to the [Greek: alpha] and β rays, the three active substances, uranium, thorium, and radium, all give out a radiation of an extraordinarily penetrating character. These γ rays are considerably more penetrating than the X rays produced in a "hard" vacuum tube. Their presence can readily be observed for an active substance like radium, but is difficult to detect for uranium and thorium unless a large quantity of active material is used.

Villard^[57], using the photographic method, first drew attention to the fact that radium gave out these very penetrating rays, and found that they were non-deviable by a magnetic field. This result was confirmed by Becquerel^[58].

Using a few milligrams of radium bromide, the γ rays can be detected in a dark room by the luminosity they excite in the mineral willemite or a screen of platinocyanide of barium. The [Greek: alpha] and β rays are completely absorbed by placing a thickness of 1 centimetre of lead over the radium, and the rays which then pass through the lead consist entirely of γ rays. The very great penetrating power of these rays is easily observed by noting the slight diminution of the luminosity of the screen when plates of metal several centimetres thick are placed between the radium and the screen. These rays also produce ionization in gases and are best investigated by the electrical method. The presence of the γ rays from 30 mgrs. of radium bromide can be observed in an electroscope after passing through 30 cms. of iron.

106. Absorption of the γ rays. In an examination of the active substances by the electrical method, the writer^[59] found that both uranium and thorium gave out γ rays in amount roughly proportional to their activity. An electroscope of the type shown in Fig. 12 was employed. This was placed on a large lead plate ·65 cm. thick the active substance being placed in a closed vessel beneath. The discharge due to the natural ionization of the air in the electroscope was first observed. The additional ionization due to the active substance must be that produced by rays which have passed through the lead plate and the walls of the electroscope. The following table shows that the discharge due to these rays decreases approximately according to an exponential law with the thickness of lead traversed.

Thickness of lead Rate of discharge
 ·62 cms. 100
  " + ·64 cms. 67
  " + 2·86 " 23
  " + 5·08 " 8

Using 100 grs. of uranium and thorium, the discharge due to the rays through 1 cm. of lead was quite appreciable, and readily measured. The results showed that the amount of γ rays was about the same for equal weights of thorium and uranium oxides. The penetrating power was also about the same as for the radium rays.

Fig. 44. The writer showed that the absorption of the γ rays from radium was approximately proportional to the density of the substance traversed. A more detailed examination of the absorption of these rays in various substances has been recently made by McClelland^[60]. The curve (Fig. 44) shows the decrease of the ionization current in a testing vessel due to the β and γ rays with successive layers of lead. It is seen that the β rays are almost completely stopped by 4 mms. of lead; the ionization is then due entirely to the γ rays.

In order to leave no doubt that all the β rays were absorbed, the radium was covered with a thickness of 8 mms. of lead, and measurements of the coefficient of absorption λ were made for additional thicknesses. The average value of λ was calculated from the usual equation I/I_{0} = e^{-λd}, where d is the thickness of matter traversed. The following table shows the value of λ, (I) for the first 2·5 mms. of matter traversed (after initially passing through 8 mms. of lead), (II) for the thickness 2·5 to 5 mms., (III) for 5 to 10 mms., (IV) 10 to 15 mms.

TABLE A.

+——————-+—————+————+————+———-+
| Substance | I | II | III | IV |
+——————-+—————+————+————+———-+
| Platinum | 1·167 | | | |
| Mercury | ·726 | ·661 | ·538 | ·493 |
| Lead | ·641 | ·563 | ·480 | ·440 |
| Zinc | ·282 | ·266 | ·248 | ·266 |
| Aluminium | ·104 | ·104 | ·104 | ·104 |
| Glass | ·087 | ·087 | ·087 | ·087 |
| Water | ·034 | ·034 | ·034 | ·034 |
+——————-+—————+————+————+———-+

In the above table, the absorption in aluminium, glass and water was too small to determine with accuracy the variation of λ with distance traversed. It will be observed that, for the denser substances, the coefficient of absorption decreases with the distance through which the rays have passed. This indicates that the rays are heterogeneous. The variation of λ is more marked in heavy substances. Table B gives the values of λ divided by density for the above numbers. If the absorption were directly proportional to the density, the quotient would be the same in all cases.

TABLE B.

λ divided by density.

+———————+————+————+————-+————+
| Substance | I | II | III | IV |
+———————+————+————+————-+————+
| Platinum | ·054 | | | |
| Mercury | ·053 | ·048 | ·039 | ·036 |
| Lead | ·056 | ·049 | ·042 | ·037 |
| Zinc | ·039 | ·037 | ·034 | ·033 |
| Aluminium | ·038 | ·038 | ·038 | ·038 |
| Glass | ·034 | ·034 | ·034 | ·034 |
| Water | ·034 | ·034 | ·034 | ·034 |
+———————+————+————+————-+————+

The numbers in column I vary considerably, but the agreement becomes closer in the succeeding columns, until in column IV the absorption is very nearly proportional to the density.

It is seen that the absorption of all three types of rays from radio-active substances is approximately proportional to the density of the substance traversed—a relation first observed by Lenard for the cathode rays. This law of absorption thus holds for both positively and negatively electrified particles projected from the radio-active substances, and also for the electromagnetic pulses which are believed to constitute the γ rays; although the absorption of the [Greek: alpha] rays, for example, is 10,000 times greater than for the γ rays. We have seen in section 84 that the value of the absorption constant λ for lead is 122 for the β rays from uranium. The value for the γ rays from radium varies betwen ·64 and ·44, showing that the γ rays are more than 200 times as penetrating as the β rays.

107. Nature of the rays. In addition to their great penetrating power, the γ rays differ from the [Greek: alpha] and β rays in not being deflected to an appreciable degree by a magnetic or electric field. In a strong magnetic field, it can be shown, using the photographic method, that there is an abrupt discontinuity between the β and γ rays, for the former are bent completely away from the latter. This indicates that, as regards the action of a magnetic field, there is no gradual transition of magnetic properties between the β and γ rays. Paschen^[61] has examined the γ rays in a very intense magnetic field, and, from the absence of deflection of these rays, has calculated that, if they consist of electrified particles carrying an ionic charge, and projected with a velocity approaching that of light, their apparent mass must be at least 45 times greater than that of the hydrogen atom.

It now remains for us to consider whether the γ rays are corpuscular in character, or whether they are a type of electromagnetic pulse in the ether similar to Röntgen rays. They resemble Röntgen rays in their great penetrating power and in their absence of deflection in a magnetic field. Earlier experiments seemed to indicate an important difference between the action of γ and X rays. It is well known that ordinary X rays produce much greater ionization in gases such as sulphuretted hydrogen and hydrochloric acid gas, than in air, although the differences in density are not large. For example, exposed to X rays, sulphuretted hydrogen has six times the conductivity of air, while with γ rays the conductivity only slightly exceeds that of air. The results obtained by Strutt, in this connection, have already been given in section 45. It is there shown that the relative conductivity of gases exposed to γ rays (and also to [Greek: alpha] and β rays) is, in most cases, nearly proportional to their relative densities; but, under X rays, the relative conductivity for some gases and vapours is very much greater than for the γ rays. It must be remembered, however, that the results obtained by Strutt were for "soft X rays," whose penetrating power was very much less than that of the γ rays. In order to see if the relative conductivity of gases produced by X rays depended upon their penetrating power, A. S. Eve^[62] made some experiments with a very "hard" X ray bulb, which gave an unusually penetrating type of rays.

The results of the measurements are shown in the table below, where the conductivity for each type of rays is expressed relative to air as unity. The results obtained for "soft" X rays by Strutt and by Eve for γ rays are added for comparison. It is seen that the hard rays show a much closer agreement than the soft rays with the density law found for the γ rays. The high values previously obtained for the vapours of chloroform and carbon tetrachloride are greatly reduced, and are very nearly the same as for the γ rays. On the other hand, the vapour of methyl iodide is an exception, and still shows a high conductivity. The γ rays were, however, forty times as penetrating as the hard X rays, and it is probable that the value of methyl iodide would be reduced with still more penetrating X rays.

Relative conductivities of gases.

+———————————-+————-+————+————-+———-+
| |Relative | "Soft" | "Hard" |] |
| Gas | Density | X rays | X rays | rays |
+———————————-+————-+————+————-+———-+
| Hydrogen | ·07 | ·11 | ·42 | ·19 |
| Air | 1·0 | 1·0 | 1·0 | 1·0 |
| Sulphuretted Hydrogen | 1·2 | 6 | ·9 | 1·23 |
| Chloroform | 4·3 | 32 | 4·6 | 4·8 |
| Methyl Iodide | 5·0 | 72 | 13·5 | 5·6 |
| Carbon Tetrachloride | 5·3 | 45 | 4·9 | 5·2 |
+———————————-+————-+————+————-+———-+

The hard X rays were found to give far more secondary radiation than the γ rays, but this effect is probably also a function of the penetrating power of the primary rays. It will be seen later (section 112) that γ rays give rise to a secondary radiation of the β ray type. This has also been observed for the X rays.

Considering the experimental evidence as a whole, there is undoubtedly a very marked similarity between the properties of γ and X rays. The view that the γ rays are a type of very penetrating X rays, also receives support from theoretical considerations. We have seen (section 52) that the X rays are believed to be electromagnetic pulses, akin in some respects to short light waves, which are set up by the sudden stoppage of the cathode ray particles. Conversely, it is also to be expected that X rays will be produced at the sudden starting, as well as at the sudden stopping, of electrons. Since most of the β particles from radium are ejected from the radium atom with velocities much greater than the cathode particles in a vacuum tube, X rays of a very penetrating character will arise. But the strongest argument in support of this view is derived from an examination of the origin and connection of the β and γ rays from radio-active substances. It will be shown later that the [Greek: alpha] ray activity observed in radium arises from several disintegration products, stored up in the radium, while the β and γ rays arise only from one of these products named radium C. It is found, too, that the activity measured by the γ rays is always proportional to the activity measured by the β rays, although by separation of the products the activity of the latter may be made to undergo great variations in value.

Thus the intensity of the γ rays is always proportional to the rate of expulsion of β particles, and this result indicates that there is a close connection between the β and γ rays. Such a result is to be expected if the β particle is the parent of the γ ray, for the expulsion of each electron from radium will give rise to a narrow spherical pulse travelling from the point of disturbance with the velocity of light.

108. There is another possible hypothesis in regard to the nature of these rays. It has been shown (sections 48 and 82) that the apparent mass of an electron increases as the speed of light is approached; theoretically it should be very great when the velocity of the electron is exceedingly close to the velocity of light. In such a case, a moving electron would be difficult to deflect by a magnetic or electric field.

The view that the γ rays are electrons carrying a negative charge and moving with a velocity nearly equal to that of light has recently been advocated by Paschen^[63]. He concluded from experiment that the γ rays like the β rays carried a negative charge. We have seen (section 85) that Seitz also observed that a small negative charge was communicated to bodies on which the γ rays impinged, but the magnitude of this charge was much smaller than that observed by Paschen. I do not think that much weight can be attached to observations that a small positive or negative charge is communicated to bodies on which the γ rays fall, for it will be shown later that a strong secondary radiation, consisting in part of electrons, is set up during the passage of the γ rays through matter. It is not improbable that the small charge observed is not a direct result of the charge carried by the γ rays, but is an indirect effect due to the secondary radiations emitted from the surface of bodies. There is no doubt that a thick lead vessel, completely enclosing a quantity of radium, acquires a small positive charge, but this result would follow whether the γ rays carry a charge or not, since the secondary radiations from the lead surface consist of projected particles which carry with them a negative charge.

On this corpuscular theory of the nature of the γ rays, each electron must have a large apparent mass, or otherwise it would be appreciably deflected by an intense magnetic field. The energy of motion of the electron must, in consequence, be very great, and, if the number of the electrons constituting the γ rays is of the same order of magnitude as the number of the β particles, a large heating effect is to be expected when the γ rays are stopped in matter. Paschen^[64] made some experiments on the heat emission of radium due to the γ rays; he concluded that the γ rays were responsible for more than half of the total heat emission of radium and carried away energy at the rate of over 100 gram calories per hour per gram of radium. This result was not confirmed by later experiments of Rutherford and Barnes^[65], who found that the heating effect of the γ rays could not be more than a few per cent. of the total heat emission of radium. These results will be considered later in chapter XII.

The weight of evidence, both experimental and theoretical, at present supports the view that the γ rays are of the same nature as the X rays but of a more penetrating type. The theory that the X rays consist of non-periodic pulses in the ether, set up when the motion of electrons is arrested, has found most favour, although it is difficult to provide experimental tests to decide definitely the question. The strongest evidence in support of the wave nature of the X rays is derived from the experiments of Barkla^[66], who found that the amount of secondary radiation set up by the X rays on striking a metallic surface depended on the orientation of the X ray bulb. The rays thus showed evidence of a one-sidedness or polarization which is only to be expected if the rays consist of a wave motion in the ether.

PART V.

Secondary Rays.

109. Production of secondary rays. It has long been known that Röntgen rays, when they impinge on solid obstacles, produce secondary rays of much less penetrating power than the incident rays. This was first shown by Perrin and has been investigated in detail by Sagnac, Langevin, Townsend and others. Thus it is not surprising that similar phenomena should be observed for the radiation from radio-active substances. By means of the photographic method, Becquerel^[67] has made a close study of the secondary radiations produced by radio-active substances. In his earliest observations, he noticed that radiographs of metallic objects were always surrounded by a diffuse border. This effect is due to the secondary rays set up by the incident rays at the surface of the screen.

The secondary rays produced by the α rays are very feeble. They are best shown by polonium, which gives out only α rays, so that the results are not complicated by the action of the β rays. Strong secondary rays are set up at the point of impact of the β or cathodic rays. Becquerel found that the magnitude of this action depended greatly on the velocity of the rays. The rays of lowest velocity gave the most intense secondary action, while the penetrating rays gave, in comparison, scarcely any secondary effect. In consequence of the presence of this secondary radiation, the photographic impression of a screen pierced with holes is not clear and distinct. In each case there is a double photographic impression, due to the primary rays and the secondary rays set up by them.

These secondary rays are deviable by a magnetic field, and in turn produce tertiary rays and so on. The secondary rays are in all cases more readily deviated and absorbed than the primary rays, from which they arise. The very penetrating γ rays give rise to secondary rays, which cause intense action on the photographic plate. When some radium was placed in a cavity inside a deep lead block, rectangular in shape, besides the impression due to the direct rays through the lead, Becquerel observed that there was also a strong impression due to the secondary rays emitted from the surface of the lead. The action of these secondary rays on the plate is so strong that the effect on the plate is, in many cases, increased by adding a metal screen between the active material and the plate.

The comparative photographic action of the primary and secondary rays cannot be taken as a relative measure of the intensity of their radiations. For example, only a small portion of the energy of the β rays is in general absorbed in the sensitive film. Since the secondary rays are far more easily absorbed than the primary rays, a far greater proportion of their energy is expended in producing photographic action than in the case of the primary rays. It is thus not surprising that the secondary rays set up by the β and γ rays may in some cases produce a photographic impression comparable with, if not greater than, the effect of the incident rays.

On account of these secondary rays, radiographs produced by the β rays of radium in general show a diffuse border round the shadow of the object. For this reason radiographs of this kind lack the sharpness of outline of X ray photographs.

110. Secondary radiation produced by [Greek: alpha] rays. Mme Curie^[68] has shown by the electric method that the [Greek: alpha] rays of polonium produce secondary rays. The method adopted was to compare the ionization current between two parallel plates, when two screens of different material, placed over the polonium, were interchanged.

These results show that the [Greek: alpha] rays of polonium are modified in passing through matter, and that the amount of secondary rays set up varies with screens of different material. Mme Curie, using the same method, was unable to observe any such effect for the β rays of radium. The production of secondary rays by the β rays of radium is, however, readily shown by the photographic method. We have already seen (section 93) that very low velocity electrons accompany the [Greek: alpha] rays from radium or radio-tellurium spread on a metal plate. These electrons are probably liberated when the [Greek: alpha] rays escape from or impinge upon matter, and the number emitted depends upon the kind of matter used as a screen. The differences shown in the above table when the screens were interchanged are explained simply in this way.

+—————————-+—————-+—————+
| | Thickness | Current |
| Screens employed | in mms. | observed |
|—————————-+—————-+—————+
| Aluminium | 0·01 | |
| Cardboard | 0·005 | 17·9 |
| | | |
| Cardboard | 0·005 | |
| Aluminium | 0·01 | 6·7 |
| | | |
| Aluminium | 0·01 | |
| Tin | 0·005 | 150 |
| | | |
| Tin | 0·005 | |
| Aluminium | 0·01 | 126 |
| | | |
| Tin | 0·005 | |
| Cardboard | 0·005 | 13·9 |
| | | |
| Cardboard | 0·005 | |
| Tin | 0·005 | 4·4 |
+—————————-+—————-+—————+

Fig. 45.

111. Secondary rays produced by β and γ rays. An examination of the amount and character of the secondary radiation emitted by various substances, when exposed to the β and γ rays of radium, has recently been made by A. S. Eve^[69]. The general experimental method employed is shown in Fig. 45.

The electroscope (Fig. 45) was placed behind a lead screen 4·5 cms. thick, which stopped all the β rays and absorbed the greater proportion of the γ rays from the radium tube placed at R. On bringing near a plate of matter M, the primary rays fell upon it and some of the secondary rays, emitted in all directions, passed into the side of the electroscope, which was covered with aluminium foil of thickness ·05 mm. Before the plate M was placed in position the rate of discharge of the electroscope was due to the natural leak and the γ rays from R, and the secondary radiation from the air. On bringing the radiator M into position, the rate of discharge was much increased, and the difference between the rate of movement of the gold-leaf in the two cases was taken as a measure of the amount of secondary rays from M. The absorption of the secondary rays was tested by placing an aluminium plate ·85 mm. thick before the face of the electroscope.

The secondary rays were found to be fairly homogeneous, for the intensity fell off according to an exponential law with the distance traversed. The value of the absorption constant λ was determined from the usual equation I/I_{0} = e^{-λd}, where d is the thickness of the screen. The table given below shows the results obtained when thick plates of different substances of the same dimensions were placed in a definite position at M. The secondary radiation from fluids was obtained by a slight alteration of the experimental arrangements.

Thirty milligrammes of radium bromide were used, and the results are expressed in terms of the number of scale divisions passed over per second by the gold-leaf.

It will be noticed that the amount of secondary radiation follows in most cases the same order as the densities, and is greatest for mercury. The value of (secondary radiation)/density is not a constant, but varies considerably, being greatest for light substances. The absorption constant of the secondary rays from different radiators is not very different, with the exception of substances such as granite, brick, and cement, which give out secondary rays of nearly twice the penetrating power of other substances.

β and γ rays.

+———————-+————-+——————+——————————+——————+
| | | Secondary | | Aluminium |
| Radiator | Density | Radiation |(Sec. Rad.)/Density | ·085 cm. |
| | | | | λ |
+———————-+————-+——————+——————————+——————+
| Mercury | 13·6 | 147 | 10·8 | |
| Lead | 11·4 | 141 | 12·4 | 18·5 |
| Copper | 8·8 | 79 | 9·0 | 20 |
| Brass | 8·4 | 81 | 9·6 | 21 |
| Iron (wrought)| 7·8 | 75 | 9·6 | 20 |
| Tin | 7·4 | 73 | 9.9 | 20·3 |
| Zinc | 7·0 | 79 | 11·3 | |
| Granite | 2·7 | 54 | 20·0 | 12·4 |
| Slate | 2·6 | 53 | 20·4 | 12·1 |
| Aluminium | 2·6 | 42 | 16·1 | 24 |
| Glass | 2·5 | 44 | 17·6 | 24 |
| Cement | 2·4 | 47 | 19·6 | 13·5 |
| Brick | 2·2 | 49 | 22·3 | 13·0 |
| Ebonite | 1·1 | 32 | 29·1 | 26 |
| Water | 1·0 | 24 | 24·0 | 21 |
| Ice | ·92 | 26 | 28·2 | |
| Paraffin solid| ·9 | 17 | 18·8 | 21 |
| " liquid| ·85 | 16 | 18·8 | |
| Mahogany | ·56 | 21·4 | 38·2 | 23 |
| Paper | ·4? | 21·0 | 52 | 22 |
| Millboard | ·4? | 19·4 | 48 | 20·5 |
| Papier-mâché | . . . | 21·9 | | |
| Basswood | ·36 | 20·7 | 57 | 22 |
| Pine | ·35 | 21·8 | 62 | 21 |
| | | | | |
| X ray screen | | 75·2 | | 23·6 |
+———————-+————-+——————+——————————+——————+

The secondary radiation not only comes from the surface of the radiator but from a considerable depth. The amount of secondary rays increases with the thickness of the radiator, and, in the case of glass and aluminium, reaches a practical maximum for a plate about 3 mms. thick.

In the above table, the secondary radiation arises from both the β rays and γ rays together. When the β rays were cut off by a layer of lead 6·3 mms. thick, placed between the radium and the radiator, the effect on the electroscope was reduced to less than 20 per cent. of its former value, showing that the β rays supplied more than 80 per cent. of the secondary radiation. The following table shows the relative amount of secondary rays from different substances when exposed to β and γ rays together and to γ rays alone. The amount from lead in each case is taken as a standard and equal to 100. The amount of secondary radiation found by Townsend from soft X rays is added for comparison.

Secondary Radiations.

+—————-+————————————-+———————+————-+
| Radiator | β and γ rays | γ rays | Röntgen |
+—————-+————————————-+———————+————-+
| Lead | 100 | 100 | 100 |
| Copper | 57 | 61 | 291 |
| Brass | 58 | 59 | 263 |
| Zinc | 57 | . . . | 282 |
| Aluminium | 30 | 30 | 25 |
| Glass | 31 | 35 | 31 |
| Paraffin | 12 | 20 | 125 |
+—————-+————————————-+———————+————-+

It will be observed that the relative amounts are about the same for the γ rays alone as for the β and γ rays together. On the other hand, the amount of secondary radiation set up by X rays is very different, lead for example giving much less than brass or copper. The secondary rays from the γ rays alone are slightly less penetrating than for the β and γ rays together, but are far more penetrating than the secondary radiation from the X rays examined by Townsend.

The amount of secondary radiation set up by the β and γ rays is mainly independent of the state of the surface of the radiator. About the same amount is obtained from iron as from iron filings; from liquid as from solid paraffin; and from ice as from water^[70].

Becquerel has shown that the secondary rays set up by the β rays are deflected by a magnet and consist of negatively charged particles (electrons). It has been pointed out in section 52 that the cathode rays are diffusely reflected from the metal on which they fall. These secondary rays consist in part of electrons moving with about the same velocity as the primary, and in part of some electrons with a much slower speed. The secondary rays set up by the β rays of radium have on an average less penetrating power than the primary rays, and consequently less velocity than the primary rays. It must be remembered that the β rays from radium are very complex, and consist of electrons projected with a considerable range of velocities. The secondary rays are, on an average, certainly more penetrating than the most easily absorbed β rays emitted from radium, and probably move with a velocity of about half that of light.

It is still uncertain whether the secondary rays are produced by the action of the primary rays on matter, or whether they consist of a portion of the primary rays whose direction of motion has been deflected in their passage through matter, so that they emerge again with diminished velocity from the surface.

112. Magnetic deflection of secondary rays from γ rays. It has been seen that the secondary rays set up by the γ rays alone are very similar in character to those caused by the β rays. This result was still further confirmed by Eve, who showed that the secondary rays produced by the γ rays are readily deflected by a magnetic field. The experimental arrangement is shown in Fig. 46.

A small electroscope was mounted on one side of a lead platform 1·2 cms. thick, which rested on a lead cylinder 10 cms. high and 10 cms. in diameter. The radium was placed at the bottom of a hole reaching to the centre of the cylinder.

On applying a strong magnetic field, at right angles to the plane of the paper, so as to bend the secondary rays from the platform towards the electroscope, the rate of discharge was much increased. On reversing the field, the effect was much diminished. Since the γ rays are not themselves deflected by a magnetic field, this result shows that the secondary radiation is quite different in character from the primary rays, and consists of electrons projected with a velocity (deduced from the penetrating power) of about half the velocity of light. We have already pointed out that the emission of electrons from a substance traversed by the rays will account sufficiently well for the charge observed by Paschen, without the necessity of assuming that the γ rays carry a negative charge of electricity.

Fig. 46.

The secondary radiation set up by Röntgen rays, like that due to the β and γ rays, consists in part of electrons projected with considerable velocity. These three types of rays seem about equally efficient in causing the expulsion of electrons from the substance through which they pass. We have seen that the X and γ rays are, in all probability, electromagnetic pulses set up by the sudden starting or stopping of electrons, and, since these rays in turn cause the removal of electrons, the process appears to be reversible. Since the β rays pass through some thickness of matter before their energy of motion is arrested, theory would lead us to expect that a type of

soft X rays should be generated in the absorbing matter.

PART VI.

113. Comparison of the ionization produced by the [Greek: alpha] and β rays. With unscreened active material the ionization produced between two parallel plates, placed as in Fig. 17, is mainly due to the [Greek: alpha] rays. On account of the slight penetrating power of the [Greek: alpha] rays, the current due to them practically reaches a maximum with a small thickness of radio-active material. The following saturation currents were observed^[71] for different thicknesses of uranium oxide between parallel plates sufficiently far apart for all the [Greek: alpha] rays to be absorbed in the gas between them. Surface of uranium oxide 38 sq. cms.

+—————————————+————————————+
| Weight of uranium oxide | Saturation current |
| in grammes per sq. cm. | in amperes per sq. cm. |
| of surface | of surface |
+—————————————+————————————+
| ·0036 | 1·7 × 10^{-13} |
| ·0096 | 3·2 × 10^{-13} |
| ·0189 | 4·0 × 10^{-13} |
| ·0350 | 4·4 × 10^{-13} |
| ·0955 | 4·7 × 10^{-13} |
+—————————————+————————————+

The current reached about half its maximum value for a weight of oxide ·0055 gr. per sq. cm. If the [Greek: alpha] rays are cut off by a metallic screen, the ionization is then mainly due to the β rays, since the ionization produced by the γ rays is small in comparison. For the β rays from uranium oxide it has been shown (section 86) that the current reaches half its maximum value for a thickness of 0·11 gr. per sq. cm.

Meyer and Schweidler^[72] have found that the radiation from a water solution of uranium nitrate is very nearly proportional to the amount of uranium present in the solution.

On account of the difference in the penetrating power of the [Greek: alpha] and β rays, the ratio of the ionization currents produced by them depends on the thickness of the radio-active layer under examination. The following comparative values of the current due to the [Greek: alpha] and β rays were obtained for very thin layers of active matter^[73]. A weight of 1/10 gramme of fine powder, consisting of uranium oxide, thorium oxide, or radium chloride of activity 2000, was spread as uniformly as possible over an area of 80 sq. cms. The saturation current was observed between parallel plates 5·7 cms. apart. This distance was sufficient to absorb most of the [Greek: alpha] rays from the active substances. A layer of aluminium ·009 cm. thick absorbed all the [Greek: alpha] rays.

+————+—————————-+—————————+——————————————+
| | Current due to | Current due to | Ratio of currents |
| |[Greek: alpha] rays|[Greek: beta] rays|[Greek: beta]/[Greek: alpha]|
+————+—————————-+—————————+——————————————+
|Uranium | 1 | 1 | ·0074 |
|Thorium | 1 | ·27 | ·0020 |
|Radium | 2000 | 1350 | ·0033 |
+————+—————————-+—————————+——————————————+

In the above table the saturation current due to the [Greek: alpha] and β rays of uranium is, in each case, taken as unity. The third column gives the ratio of the currents observed for equal weights of substance. The results are only approximate in character, for the ionization due to a given weight of substance depends on its fineness of division. In all cases, the current due to the β rays is small compared with that due to the [Greek: alpha] rays, being greatest for uranium and least for thorium. As the thickness of layer increases, the ratio of currents [Greek: beta]/[Greek: alpha] steadily increases to a constant value.

114. Comparison of the energy radiated by the [Greek: alpha] and [Greek: beta] rays. It has not yet been found possible to measure directly the energy of the [Greek: alpha] and β rays. A comparison of the energy radiated in the two forms of rays can, however, be made indirectly by two distinct methods.

If it be assumed that the same amount of energy is required to produce an ion by either the [Greek: alpha] or the [Greek: beta] ray, and that the same proportion of the total energy is used up in producing ions, an approximate estimate can be made of the ratio of the energy radiated by the [Greek: alpha] and β rays by measuring the ratio of the total number of ions produced by them. If λ is the coefficient of absorption of the β rays in air, the rate of production of ions per unit volume at a distance x from the source is q_{0}e^{-λx} where q_{0} is the rate of ionization at the source.

The total number of ions produced by complete absorption of the rays is

[integral]_{0}^[infinity] q_{0}e^{-λx}dx = q_{0}/λ.

Now λ is difficult to measure experimentally for air, but an approximate estimate can be made of its value from the known fact that the absorption of β rays is approximately proportional to the density of any given substance.

For β rays from uranium the value of λ for aluminium is about 14, and λ divided by the density is 5·4. Taking the density of air as ·0012, we find that for air

λ = ·0065.

The total number of ions produced in air is thus 154q_{0} when the rays are completely absorbed.

Now from the above table the ionization due to the β rays is ·0074 of that produced by [Greek: alpha] rays, when the β rays passed through a distance of 5·7 cms. of air.

Thus we have approximately

(Total number of ions produced by β rays)/(Total number of ions produced by [Greek: alpha] rays) = ·0074/5·7 × 154 = 0·20.

Therefore about 1/6 of the total energy radiated into air by a thin layer of uranium is carried by the β rays or electrons. The ratio for thorium is about 1/22 and for radium about 1/14, assuming the rays to have about the same average value of λ.

This calculation takes into account only the energy which is radiated out into the surrounding gas; but on account of the ease with which the [Greek: alpha] rays are absorbed, even with a thin layer, the greater proportion of the radiation is absorbed by the radio-active substance itself. This is seen to be the case when it is recalled that the [Greek: alpha] radiation of thorium or radium is reduced to half value after passing through a thickness of about 0·0005 cm. of aluminium. Taking into consideration the great density of the radio-active substances, it is probable that most of the radiation which escapes into the air is due to a thin skin of the powder not much more than ·0001 cm. in thickness.

 

An estimate, however, of the relative rate of emission of energy by the [Greek: alpha] and β rays from a thick layer of material can be made in the following way:—For simplicity suppose a thick layer of radio-active substance spread uniformly over a large plane area. There seems to be no doubt that the radiations are emitted uniformly from each portion of the mass; consequently, the radiation, which produces the ionizing action in the gas above the radio-active layer, is the sum total of all the radiation which reaches the surface of the layer.

 

Let λ_{1} be the average coefficient of absorption of the [Greek: alpha] rays in the radio-active substance itself and σ the specific gravity of the substance. Let E_{1} be the total energy radiated per sec. per unit mass of the substance when the absorption of the rays in the substance itself is disregarded. The energy per sec. radiated to the upper surface by a thickness dx of a layer of unit area at a distance x from the surface is given by

(1/2)E_{1}σe^{-λ_{1}x}dx.

The total energy W_{1} per unit area radiated to the surface per sec. by a thickness d is given by

W_{1} = (1/2)[integral]_{0}^d E_{1}σe^{-λ_{1}x}dx
= (E_{1}σ/(2λ_{1}))(1 - e^{-λ_{1}d}) = E_{1}σ/(2λ_{1})

if λ_{1}d is large.

 

In a similar way it may be shown that the energy W_{2} of the β rays reaching the surface is given by W_{2} = E_{2}σ/(2λ_{2}) where E_{2} and λ_{2} are the values for the β rays corresponding to E_{1} and λ_{1} for the [Greek: alpha] rays. Thus it follows that

E_{1}/E_{2} = λ_{1}W_{1}/(λ_{2}W_{2});

λ_{1} and λ_{2} are difficult to determine directly for the radio-active

substance itself, but it is probable that the ratio λ_{1}/λ_{2} is not very different from the ratio for the absorption coefficients for another substance like aluminium. This follows from the general result that the absorption of both [Greek: alpha] and β rays is proportional to the density of the substance; for it has already been shown in the case of the β rays from uranium that the absorption of the rays in the radio-active material is about the same as for non-radio-active matter of the same density.

With a thick layer of uranium oxide spread over an area of 22 sq. cms., it was found that the saturation current between parallel plates 6·1 cms. apart, due to the [Greek: alpha] rays, was 12·7 times as great as the current due to the β rays. Since the [Greek: alpha] rays were entirely absorbed between the plates and the total ionization produced by the β rays is 154 times the value at the surface of the plates,

W_{1}/W_{2} = (total number of ions due to [Greek: alpha] rays)/(total number of ions due to β rays)
= (12·7 × 6·1)/154 = 0·5 approximately.

Now the value ofλ_{1} for aluminium is 2740 and of λ_{2} for the same metal 14, thus

E_{1}/E_{2} = λ_{1}W_{1}/(λ_{2}W_{2}) = 100 approximately.

This shows that the energy radiated from a thick layer of material by the β rays is only about 1 per cent. of the energy radiated in the form of [Greek: alpha] rays.

This estimate is confirmed by calculations based on independent data. Let m_{1}, m_{2} be the masses of the [Greek: alpha] and β particles respectively and v_{1}, v_{2} their velocities.

(Energy of one [Greek: alpha] particle)/(Energy of one β particle) = m_{1}v_{1}^2/(m_{2}v_{2}^2) = ((m_{1}/e)v_{1}^2)/((m_{2}/e)v_{2}^2).

Now it has been shown that for the [Greek: alpha] rays of radium

v_{1} = 2·5 × 10^9,
e/m_{1} = 6 × 10^3.

The velocity of the β rays of radium varies between wide

limits. Taking for an average value

v_{2} = 1·5 × 10^{10},
e/m_{2} = 1·8 × 10^7,

it follows that the energy of the [Greek: alpha] particle from radium is almost 83 times the energy of the β particle. If equal numbers of [Greek: alpha] and β particles are projected per second, the total energy radiated in the form of [Greek: alpha] rays is about 83 times the amount in the form of β rays.

Evidence will be given later (section 246) to show that the number of [Greek: alpha] particles projected is probably four times the number of β particles; so that a still greater proportion of the energy is emitted in the form of [Greek: alpha] rays. These results thus lead to the conclusion that, from the point of view of the energy emitted, the [Greek: alpha] rays are far more important than the β rays. This conclusion is supported by other evidence which is discussed in chapters XII and XIII, where it will be shown that the [Greek: alpha] rays play by far the most important part in the changes occurring in radio-active bodies, and that the β rays only appear in the latter stages of the radio-active processes. From data based on the relative absorption and ionization of the β and γ rays in air, it can be shown that the γ rays carry off about the same amount of energy as the β rays. These conclusions are confirmed by direct measurement of the heating effect of radium, which is discussed in detail in chapter XII.

1. In an examination of uranium the writer (Phil. Mag. p. 116, Jan. 1899) found that the rays from uranium consist of two kinds, differing greatly in penetrating power, which were called the α and β rays. Later, it was found that similar types of rays were emitted by thorium and radium. On the discovery that very penetrating rays were given out by uranium and thorium as well as by radium, the term γ was applied to them by the writer. The word "ray" has been retained in this work, although it is now settled that the α and β rays consist of particles projected with great velocity. The term is thus used in the same sense as by Newton, who applied it in the Principia to the stream of corpuscles which he believed to be responsible for the phenomenon of light. In some recent papers, the α and β rays have been called the α and β "emanations." This nomenclature cannot fail to lead to confusion, since the term "radio-active emanation" has already been generally adopted in radio-activity as applying to the material substance which gradually diffuses from thorium and radium compounds, and itself emits rays.
2. This method of illustration is due to Mme Curie, Thèse présentée à la Faculté des Sciences de Paris, 1903.
3. Giesel, Annal. d. Phys. 69, p. 834, 1899.
4. Meyer and Schweidler, Phys. Zeit. 1, pp. 90, 113, 1899.
5. Becquerel, C. R. 129, pp. 997, 1205. 1899.
6. Curie, C. R. 130, p. 73, 1900.
7. Rutherford, Phil. Mag. January, 1899.
8. Rutherford and Grier, Phil. Mag. September, 1902.
9. Becquerel, C. R. 130, pp. 206, 372, 810, 979. 1900.
10. M. and Mme Curie, C. R. 130, p. 647, 1900.
11. The activity of the radium preparation was not stated in the paper.
12. Dorn, Phys. Zeit. 4, No. 18, p. 507, 1903.
13. Strutt, Phil. Mag. Nov. 1903.
14. Wien, Phys. Zeit. 4, No. 23, p. 624, 1903.
15. Dorn, C. R. 130, p. 1129, 1900.
16. Becquerel, C. R. 130, p. 809, 1900.
17. Kaufmann, Phys. Zeit. 4, No. 1 b, p. 54, 1902.
18. Abraham, Phys. Zeit. 4, No. 1 b, p. 57, 1902.
19. Kaufmann, Nachrichten d. Ges. d. Wiss. zu Gött., Nov. 8, 1901.
20. Simon, Annal. d. Phys. p. 589, 1899.
21. Kaufmann, Phys. Zeit. 4, No. 1 b, p. 54, 1902.
22. Paschen, Annal. d. Phys. 14, p. 389, 1904.
23. Meyer and Schweidler, Phys. Zeit. pp. 90, 113, 209, 1900.
24. Lenard, Annal. d. Phys. 56, p. 275, 1895.
25. Strutt, Nature, p. 539, 1900.
26. Seitz, Phys. Zeit. 5, No. 14, p. 395, 1904.
27. It is presumed that the results were corrected, if necessary, for the discharging action due to the ionized gas, although no direct mention of this is made in the paper by Seitz.
28. Strutt, Phil. Trans. A, p. 507, 1901.
29. Crookes, Proc. Roy. Soc. 1902. Chem. News, 85, p. 109, 1902.
30. Mme Curie, C. R. 130, p. 76, 1900.
31. Rutherford, Phil. Mag. Feb. 1903. Phys. Zeit. 4, p. 235, 1902.
32. Becquerel, C. R. 136, p. 199, 1903.
33. Becquerel, C. R. 136, p. 431, 1903.
34. Des Coudres, Phys. Zeit. 4, No. 17, p. 483, 1903.
35. Becquerel, C. R. 136, p. 1517, 1903.
36. Bragg, Phil. Mag. Dec. 1904; Bragg and Kleeman, Phil. Mag. Dec. 1904.
37. Further experimental results bearing on this important question are given in an Appendix to this book.
38. Bakerian Lecture, Phil. Trans. A, p. 169, 1904.
39. Strutt, Phil. Mag. Aug. 1904.
40. J. J. Thomson, Proc. Camb. Phil. Soc. 13, Pt. I. p. 39, 1905. Nature, Dec. 15, 1904.
41. Rutherford, Nature, March 2, 1905. J. J. Thomson, Nature, March 9, 1905.
42. Crookes, Proc. Roy. Soc. 81, p. 405, 1903.
43. Elster and Geitel, Phys. Zeit. No. 15, p. 437, 1903.
44. Glew, Arch. Röntgen Ray, June 1904.
45. Becquerel, C. R. 137, Oct. 27, 1903.
46. Tommasina, C. R. 137, Nov. 9, 1903.
47. An interesting side-light is thrown on this question by the experiments described in Appendix A of this book.
48. Rutherford and Miss Brooks, Phil. Mag. July 1902.
49. In order to obtain a thin layer, the compound to be tested is ground to a fine powder and then sifted through a fine gauge uniformly over the area, so that the plate is only partially covered.
50. Rutherford, Phil. Mag. Jan. 1899.
51. Owens, Phil. Mag. Oct. 1899.
52. Rutherford and Miss Brooks, Phil. Mag. July, 1900.
53. Since the ionization at any point above the plate is the resultant effect of the [Greek: alpha
54. Townsend, Phil. Mag. Feb. 1901.
55. Durack, Phil. Mag. July 1902, May 1903.
56. Bragg and Bragg and Kleeman, Phil. Mag. Dec. 1904.
57. Villard, C. R. 130, pp. 1010, 1178, 1900.
58. Becquerel, C. R. 130, p. 1154, 1900.
59. Rutherford, Phys. Zeit. 3, p. 517, 1902.
60. McClelland, Phil. Mag. July 1904.
61. Paschen, Phys. Zeit. 5, No. 18, p. 563, 1904.
62. A. S. Eve, Phil. Mag. Nov. 1904.
63. Paschen, Annal. d. Physik, 14, p. 114, 1904; 14, 2, p. 389, 1904. Phys. Zeit. 5, No. 18, p. 563, 1904.
64. Paschen, Phys. Zeit. 5, No. 18, p. 563, 1904.
65. Rutherford and Barnes, Phil. Mag. May 1905. Nature, p. 151, Dec. 15, 1904.
66. Barkla, Nature, March 17, 1904.
67. Becquerel, C.R. 132, pp. 371, 734, 1286. 1901.
68. Mme Curie, Thèse présentée à la Faculté des Sciences, Paris 1903, p. 85.
69. A. S. Eve, Phil. Mag. Dec. 1904.
70. In a recent paper (Phil. Mag. Feb. 1905), McClelland has, in the main, confirmed the experimental results obtained by Eve. An electrometer was used instead of an electroscope. He finds, in addition, that the amount of secondary radiation depends on the angle of incidence of the primary rays, and is greatest for an angle of 45°. In a letter to Nature (Feb. 23, p. 390, 1905), he states that more recent experiments have shown that the amount of secondary radiation from different substances is a function of their atomic weights rather than of their densities. In every case examined, the amount of secondary radiation increases with the atomic weight, but is not proportional to it.
71. Rutherford and McClung, Phil. Trans. A. p. 25, 1901.
72. Meyer and Schweidler, Wien Ber. 113, July, 1904.
73. Rutherford and Grier, Phil. Mag. Sept. 1902.