zincite; Al; Mg. These results all well correspond with those obtained by the writer's experiments with crystal detectors and also with the order arranged after the electron affinities of metals determined by many previous investigators. By these experiments the following facts are now established; (1) rectification is possible by two separate electrodes of different electron emissions in the cold state; (2) the direction of rectified current in this case is quite the same with that rectified by the contact rectifier composed of the same electrode materials. The essential differences between these two rectifiers are the distance between the electrodes and the surface area from which electron emissions occur. In the vacuum tube there is always a layer of air between the two electrodes and ionization of gas molecules necessarily Occurs. In the crystal rectifier, it will be natural to think of the distance between the electrodes as varying from the real contact to the wide space involving the ordinary atmosphere. As the latice constant for metals and crystals has a value of 3.2-5.8 × 10-8 cm. and the mean distance between gas molecules under normal pressure is 3.33 x 10 cm., the free space of nearly from 6 × 10-8 to 60 x 10-8 cm. will not allow the existence of gas molecules, and at least free electronic emissions may be possible in the space of this dimension. But as the free mean path of electrons in the atmosphere at normal pressure is nearly 9 x 10-5 cm., the electrons are able to reach the opposite electrode without collision with gas molecules in the wider space far beyond 60 × 10-8. It may be, therefore, possible to say that the emission at the free space of a contact rectifier occurs in the highest vacuum, and no ionization of gas molecules is involved in an ordinary case of the wireless detector. 7 But in the case of the battery charger, which is another type of contact rectifiers consisting of two plate electrodes of metal and metallic oxide or sulphide and which is thought by the present writer to act on the same principle as ordinary crystal detectors, the electron emissions seem to occur even in a wider space involving an ordinary atmosphere in virtue of a higher potential. Another difference between the crystal detector and the vacuum tube used in the preceding experiments is the surface from which electrons are emitted. In the vacuum tube the electrode surface works as a whole including any heterogeneous parts if present, while in a crystal detector only a special portion is brought to work and heterogeneous parts will work as such. For example, in an argentiferous synthetic galena, silver grains between PbS crystals will always exert their influence upon the electron emissions in a vacuum tube, but in a crystal detector their effects will be quite different according to the position of the needle point. But there will be no difference in a homogeneous substance. At the first sight the surface area emitting electrons appears to be very different in a needle point and in a crystal. But this is not true, because the emissions only occur in the limited portion of the electrodes which come face to face within a certain limited distance and other parts will remain idle. In short, there is no substantial difference between a crystal detector and the vacuum tube except the metallic conduction at the real contact points in the former. This analogy is illustrated by figs. 1 and 2. Suppose A and B in fig. 1 to be two components of a crystal detector which are in contact really at C and D. The distance between M and N is thought to be so small that any molecules of gases are not admitted to enter between this space, that is to say, the space is vacuum. In fig. 2, A' and B' are the the electrodes and C'-D' denotes a glass bulb of a vacuum tube. The real contact portions denoted by C and D will serve as the supports of the electrodes and correspond to the glass bulb C-D' of the vacuum tube. C-D and C'-D' will equally make a route of leakage current if they are not good insulators and if this leakage attains to a certain value the potential difference between M and N as well as M' and N' will drop to such a value that electron emissions between Phil. Mag. S. 7. Vol. 6. No. 34. July 1928. N the electrodes will become impossible. As C and D in the crystal detector are composed of the same materials as the electrodes, it is impossible to ensure an insulation of high degree, more or less leakage as metallic conduction being inevitable. Thus it will be clear that the proper electrical resistance is of the prime importance in a crystal to be used as a detector. In conclusion, a crystal detector is nothing else than a cold vacuum tube which works as a rectifier by the difference of electron emissions from two electrodes badly insulated. Electrotechnical Laboratory, XI. The Determination of the Atomic Scattering Power for X-Rays from Powders of Gold, Silver, and Aluminium for Cu Ka Radiation. By J. BRENTANO, D.Sc., Lecturer in Physics, Manchester University *. [Plate I.] Summary. IN the present paper experiments are described intended to obtain comparative values for the scattering power of gold, silver, and aluminium. The measurements are made with small powder particles, and a method is employed in which the intensities are measured from composite layers. Some points concerning this method, which makes it possible to overcome certain difficulties encountered in measuring the intensities of X-ray reflexions from powers, are discussed, and the procedure is indicated for evaluating the photographic records. The results of the experiments indicate that, for the elements of high atomic weight examined in the state of very fine powders, the scattered intensities increase considerably less rapidly than F2 and that better agreement is obtained by assuming the scattered intensity proportional to F. These results are discussed. 1. Na previous paper † a discussion was given of some tained from extremely fine powders of rock-salt. Communicated by the Author. † J. Brentano, Phil. Mag. iv. p. 620 (1927). The object of these measurements, which were recorded with a photographic method, was not to obtain data of a higher degree of accuracy than had been obtained from measurements on large crystal faces, but to verify whether the measurements made on large crystals were not affected to any considerable extent by extinction. It resulted that in the particular case of rock-salt extinction effects played a very small part, but it was found that in other cases where these effects were large, the breaking up of a crystal into a powder could, in general, not be considered a reliable means for eliminating extinction, unless the particles are of the order of 10-6 or 10-5 cm. in diameter, so that only a limited advantage is obtained by using coarser powders in preference to large crystals. 2. In the present paper some measurements are described, directed to verify the general assumptions on the scattering of X-rays from atoms in crystal lattices. A problem which presents itself in this connexion is in a certain way similar to the problem discussed in the case of rock-salt. A number of factors which determine the intensity of X-ray reflexions and which involve a certain amount of uncertainty in their numerical evaluation, assume definite values or become negligible for small angles of deflexion or, more exactly, for small values of sin A, where is the glancing angle of the reflexion and λ the wave-length. One of these factors is the quantity F, which we can define as the effective negative charge of the atom, which situated at its centre would be equivalent to the actual charge distributed in space for scattering radiation of the particular wave-length at the particular glancing angle *. F becomes equal to the actual charge for small values of sin x. Other factors are terms accounting for heat motion and for the Compton effect, which become negligible for small angles of deflexion. On the other hand, extinction effects are greatest for the strong reflexions at small angles. They depend on the dimensions of the regular crystal units and on their distribution in the crystal and cannot be determined in a direct way. *A more general definition of F can be given in which F is associated with the amplitude scattered from a plane of atoms, but for the purpose of the interpretation of the present experiments it is of greater interest to give to F the more special significance, rather than to introduce it in a phenomenological way. It seemed, therefore, that a considerable simplification in the interpretation of the results could be obtained by measuring the reflexions at small glancing angles from a powder consisting of particles so small as to satisfy the conditions of negligible extinction. In this way we can avoid the superposition of too many factors, which render the interpretation of X-ray intensity measurements so difficult, and attempt to verify the fundamental assumptions before introducing them in more complex cases. The general evidence derived from the analysis of structures supports the classical relation making the contribution of each atom to the scattered intensity proportional to F2, but most of this evidence refers to light atoms, and not so much information is available with respect to the scattering from atoms of higher atomic number. Measurements for the intensity of scattering were therefore made for gold, silver, and aluminium, which all belong to the cubic face-centered type of crystals. In pursuing these determinations a method for measuring the intensities had to be evolved so as to be adapted to the particular conditions of a powder. We have to refer to a few points in this connexion to account for the particular way in which the determinations were carried out. Šimilar conditions present themselves when quantitative intensity measurements are required in connexion with the determination of structures; we discuss them, therefore, in a more general way than would be strictly necessary for the purpose of our experiments. 3. The effect of extinction on the intensity of X-ray reflexions from the individual particles of a crystal powder depends mainly on the extinction in the particular crystal unit which is contributing to the reflexion, which Darwin calls primary extinction. Darwin has given an approximate expression indicating the relative reduction of the intensity of the reflected radiation owing to the extinction in the reflecting unit. If m is the number of the reflecting planes and g=N(e2/mc2)Fx2/sin20 measures the amplitude reflected from one plane, N being A general exposition of the phenomena of extinction in a single perfect crystal and in a crystal of "mosaic" structure has recently been given by Bragg, Darwin, & James, Phil. Mag. i. p. 897 (1926), and by P. Ewald, Handbuch der Physik, xxiv. (1926). |