FROM "ANNALEN DER PHYSIKS" (J. BARTH) Fig. 46.-DIAGRAM OF GERLACH AND STERN APPARATUS FOR THE MAG NETIC DEFLECTION OF ATOMIC RAYS plate was only 3 mm. square. In the experiments with silver, a glass receiving-plate was used. The traces obtained were generally developed by immersion in a hydroquinone silver nitrate solution, the deposition of silver being necessary to intensify the traces obtained.
In the absence of a field, a trace consisting of a single line is obtained. When the field is applied the atoms will be deflected; the deflection depending on the resolved magnetic moment 12 in the direction of the field, the velocity of the atoms (which can be calculated from the oven temperature), the length of the OH magnetic field, and the variation -- i as in the strength of the field 1144 that to the ;I,*1 re$ in '4 to a 4, tht I DLit ( eltal ts) Ind nal to 1). Tilos, 1em no I solved, tber e Malsgoetic gtely connedt of the spin, gs of Iron, large single (body centred developed, and rocts have by K. Bed titter (1921), Kays (1926) 1921). There in in the general shish are prob. ttia which may ipuritie cunt y magnetization 1 axis saturation final axis. G. emetic prop.
in Willi, of the but this ii2tion in weak For the iron rases the initial bviteresis whilt nalleoce is f diak ible to deter• restits arc not is ifs 011 ale site Lioa tt101 of coos do IDUrb t‘j ed srP in a direction at right angles to the stream of atoms. If s is the deflection, it may be shown that, with A as an apparatus constant = A s .
The resolved magnetic moment ,u may be conveniently expressed as a multiple of the Bohr unit magneton (equal to 9.23 X this being the magnetic moment on the original 47r me quantum theory for an electron moving in an orbit with the unit angular momentum). If µo is the atomic magnetic moment and the magnetic axis makes an angle 0 with the field, /..1 = The results to be anticipated on,, the classical and quantum the ories are quite different. On the classical theory, any orientation of the atoms in the field is possible, so that, when the field is applied, the single line trace should broaden out into a band. On the quantum theory, however, the re solved magnetic moment in the field direc tion can only assume certain discrete values. For an atom with unit moment, the re solved moment will be equal to or — 1, so that when the field is applied the trace should split up into two, and there should be no undeflected rays. (The theory of "space quantization," which leads to the conclusion that only certain resolved mo ments are possible, was developed by P.
Debye and A. Sommerfield in connection with the quantum theory of the Zeeman effect [q.v.]. On the basis of the theory, the possi ble resolved magnetic moments of atoms in any spectroscopic state may be determined from purely spectroscopic data ; the resolved moments being the so-called mg values.) The first experiments
(1921), which were made with silver, were inconclusive, but even tually the apparatus was so refined that, not only was a splitting clearly shown, but quantitative measurements could be made. Within the limits of error the experimental magnetic moment of the silver atom was found to be equal to one Bohr magneton. An enlarged schematic representation of the silver trace is shown in fig. 47. The attracted beam (to left) is unsuited for measure ments owing to the strong and varying inhomogeneity of the field near the wedge pole piece. The deviation is most conveniently ob tained from the repelled beam, being given by the distance b, a mean of a and c. Since the earlier experiments were made, the technique has been greatly improved. A considerable number of elements have been investigated, and experiments are being made also with molecular rays, the molecules containing more than one atom.
Copper and gold, like silver, give two traces, the deviation (AL = ± I) corresponding to one Bohr magneton, as is anticipated from the normal spectroscopic state of the atoms, in which there is one electron (the valency electron) outside a closed electronic configuration in which the magnetic moments of all the electrons balance each other. In the atoms of the alkali metals, and of hydrogen, the normal spectroscopic state is the same; sodium and potassium (investigated by J. B. Taylor and also A. Leu), and atomic hydrogen (T. E. Phipps and S. B. Taylor, and E. Wrede) all give traces corresponding to one Bohr magneton. In the hy drogen experiments the receiving plate was covered with molyb denum oxide, the atomic hydrogen producing a trace by reduction of the oxide. The atomic rays of zinc, cadmium and mercury, and also tin and lead are undeflected—the atomic magnetic moment is zero, corresponding to a balanced electronic configuration. For thallium there is a double trace, but the separation is smaller than that for silver, corresponding to p. :Li'. This value agrees with that predicted from the spectra. With bismuth and antimony the results were difficult to interpret, probably owing to the presence of molecular rays. Nickel gave a multiple trace, and iron an undeviated trace, but with these elements further experiments are needed.
In general, the magnetic deviation experiments confirm the quantum spectroscopic theory, in particular as to the deduction of magnetic moments of atoms from an analysis of the spectra. They form, perhaps, the most striking direct experimental con firmation of the quantum conception of atomic structure, and, from a purely magnetic point of view, they provide the most direct method of investigating the magnetic properties of atoms.