RECENT ADVANCES IN SCIENCE 539 



three mutually perpendicular and equally defined planes of 

 cleavage, and show no traces of diamagnetism or paramagnetism 

 in accordance with these views. The presence of several planes 

 of cleavage in a crystal represents that the areas of the electron 

 orbits can be projected parallel as well as perpendicular to the 

 principal cleavage, and therefore the necessary spacial arrange- 

 ment of electrons rotating in small circles or ellipses at various 

 distances from the atomic nucleus is similar to that suggested 

 by the Lewis-Langmuir cubical atom theory. On this theory 

 one of the stable arrangements of electrons takes the form of 

 eight electrons situated at the corners of a cube, and it therefore 

 offers a more reasonable explanation of magnetic phenomena 

 than the Bohr-Sommerfeld theory. 



In the Physical Review for November 1920 Messrs, A. H. 

 Compton and O. Rognley describe an experiment designed to 

 test the dependence of the intensity of a beam of X-rays on the 

 magnetic state of a ferromagnetic crystal from which it is re- 

 flected. A crystal of magnetite is magnetised {a) parallel and 

 {b) perpendicular to the III face, and the intensity of a beam of 

 X-rays reflected from this face is investigated in the first three 

 orders of reflection in case {a) and in the first four orders in 

 case {b). No detectable difference is found in the beams reflected 

 from the crystal in the magnetised and unmagnetised states, 

 although a sensitive null method capable of recording a i per 

 cent, difference of intensity is used, and the crystal is magnetised 

 to slightly over one-third its saturation value perpendicular to 

 the reflecting face. If magnetisation is due to the rotation 

 of groups of atoms, or of single iron atoms of the Rutherford 

 type, or of the type suggested by Hull, then a large difference 

 in intensity should result. Hence the conclusion previously 

 formed by A. K. Compton and A. E. Trousdale {Phys. Rev. 

 5, 191 5) that the elementary magnet is not a group of atoms is 

 confirmed ; and this is consistent with the results of A. P. 

 Chattock and F. B. Fawcett {Phil Mag., Nov. 1894). The 

 authors also conclude that, in all probability, the elementary 

 magnet is not the atom in its entirety. The experimental 

 results are, however, in agreement with the view that the 

 electron itself is the ultimate magnetic particle, although they 

 are not in themselves a definite proof that this is the case. 



The investigations of Mr. J. Chadwick on the charge on the 

 atomic nucleus, to which reference is made in the Bakerian 

 Lecture, 1920, is published in the Phil. Mag. for December 

 1920. Geiger and Marsden, by means of experiments on the 

 scattering of a particles, determined the charge on the atomic 

 nucleus to be ^ Ae, where A is the atomic wieght, and e the 

 electronic charge, within an error of about 20 per cent., the 

 inaccuracy being due to the employment of different methods 



