162 



NA TURE 



[July 29, 192: 



Earthquakes in the Region around Tokyo. — 

 Prof. Omori contributes to the second number of the 

 new journal, Seismological Notes, an interesting paper 

 on the Tokyo earthquake of December 8, 192 1, the 

 strongest experienced in that city since 1894, though 

 resulting in only slight damage to buildings. The 

 shock was notable for its extraordinary duration, 

 having been sensible for three minutes. The position 

 of the epicentre, as determined by the duration of 

 the preliminary tremors at Tokyo, Mito and Choshi, 

 is 58 kms. N. 65° E. of Hongo (Tokyo). The actual 

 distance of the focus from the latter place was, 

 however, 65 kms., leading to the conclusion that the 

 depth of the focus was about 29 kms. or 18 miles. 

 Prof. Omori also considers the distribution of Tokyo 

 earthquakes in space and time. Those of the eight 

 years 1914-1921 originated for the most part in four 

 regions, three of them at a mean distance of about 

 35 miles from Tokyo, the district surrounding the 

 capital being at present immune, and the fourth a 

 submarine zone off the eastern coast of the .Main 

 Island. The curve of annual frequency beginning 

 with the year 1876 resembles so closely the curve 

 of mean precipitation at Niigata and Akita (places 

 in the Japan Sea region with an abundant snowfall) 

 as to suggest the probability that the precipitation 

 on the north-west side of the Main Island may be 

 one of the secondary causes that determine the 

 frequency of earthquakes felt in Tokyo. 



Auroral Measurements. — In No. 8, vol. u., 

 Geofysiske Publikationer of the Norwegian Geophysical 

 Commission, Prof. Karl Stormer gives further results 

 of his auroral measurements. As a check on the 

 accuracy of his methods, he deals with simultaneous 

 observations taken on March 22, 1920, at Christiania, 

 Kongsberg and Frederikstad. Combining the statii ins 

 in pairs, he gets three sets of results derived from three 

 differently oriented bases, having lengths of approxi- 

 mately 66, 79, and 89 kilometres. The heights of 

 three selected auroral points are calculated in two 

 different ways. As the accuracy seems much the 

 same for the two methods, it will suffice to give the 

 results of the first. The means of the heights 

 calculated for the three points were 143, 177, and 

 214 km., and the differences between the highest and 

 lowest of the three estimates from the three bases 

 were respectively 5, 5, and 7 km. A second chapter 

 deals with observations on the height and position 

 of an auroral arc determined from observations with 

 a base of 258-5 km. As seen from the southmost 

 station, near Christiania, the arc was near the horizon. 

 It was overhead at a point situated about 700 km. 

 to the north-west. Besides diagrams in the text 

 there are eight large plates, the first two referring 

 to the cases described above. The others are selected 

 as among the finest auroras which Prof. Stormer 

 has yet photographed. In some cases difficulty may 

 be experienced in deciding which is the top and which 

 is the bottom of the picture. This might with 

 advantage have been stated in words on the plates. 



The New Magnetic Atom and its Properties. — 

 In order to improve his original theory of magnetic 

 induction in ferromagnetic materials, Sir J. A. 

 Ewing published in February his theory that the 

 magnetic element is not the atom itself but something 

 smaller which can rotate within the atom. It ma\ 

 be likened to a wheel with a number of like poles 

 on its rim and the poles of opposite kind at its centre. 

 Around it is a fixed ring provided with magnetic 

 poles, under the action of which the wheel may 

 take up a number of positions of equilibrium. Two 

 months later, Prof. E. T. Whittaker showed that a 



model of this type when approached by an electron 

 with kinetic energy exceeding a certain amount U, 

 would take from it an amount U depending on the 

 electronic charge and on the constants of the model, 

 and that if n is the frequency of the oscillations oi the 

 model and of the radiation it emits, then U =hn 

 where /; is the constant of the quantum theory. More 

 recently Dr. H. S. Allen has pointed out that such a 

 vibrator would not lead to the equipartition law, and 

 Dr. R. A. Houston has shown that for the model to 

 vibrate with the frequency of sodium light its 

 diameter must be 2-7 x io~ 8 cm. and with X-ray 

 frequency 35 x 10"" cm. which agree well with the 

 known diameter, 2 x 10-* cm. of the hydrogen atom. 

 The four papers mentioned will be found in the 

 Proceedings of the Royal Society of Edinburgh. 



Structure of Abraded Glass Surfaces. — A 

 series of papers issued from the Research Department 

 of Messrs. Taylor, Taylor and Hobson, Ltd., and 

 published in the current number of the Transactions 

 of the Optical Society, provide a valuable contribu- 

 tion to our knowledge of the physical nature of the 

 processes involved in the workshop operations of 

 cutting and grinding, particularly of hard brittle 

 substances. Detailed photomicrographic studies of 

 the structure of ground glass surfaces and of the flaws 

 produced in glass by various means lead the author, 

 Mr. F. YV. Preston, to the conception of a flaw and 

 fissure complex rather than the current view of a 

 hill and hollow structure as characterising a ground 

 surface in brittle materials. His experiments show 

 that in polishing as well as in grinding, the forces 

 of mechanical abrasion are active, although in the 

 former case they tend to be obscured by surface 

 tension effects. During polishing the whole of the 

 Haw (omplex previously formed is broken away from 

 the roots and a slight molecular rearrangement of 

 the new surface then takes place, as observed by 

 Beilby, who has demonstrated the existence on all 

 polished surfaces of a flowed (surface tension) layer. 

 The enhanced solubility of ground surfaces in hydro- 

 fluoric acid, and also the appearance of the structure 

 of polished surfaces as developed by etching, suggest 

 a low value for the thickness of this surface tension 

 layer — in the case of glass some two or three millionths 

 of an inch at the most. 



Repeating Patterns as E'ecorations. — The 

 Journal of the Royal Society of Arts for June 30 

 contains an interesting paper by Major P. A. Mac- 

 Mahon on the design of repeating patterns for 

 decorative work. Major MacMahon has recently 

 investigated the modes of dividing flat space into 

 identical figures without gaps or overlapping. Such 

 divisions are familiar to every one in wall paper 

 designs, which are generally formed by cutting up 

 the plane first into identical parallelograms. More 

 complex types of repeating pattern were designed in 

 very early days and some of them still survive in 

 ancient tessellated pavements. The method of 

 classification obtained by Major MacMahon has 

 proved wide enough to place every such design 

 brought to his notice, and he has reasons for thinking 

 that it is exhaustive. He reports that at least 90 

 per cent, of the few hundreds of categories have 

 apparently never been drawn upon for practical 

 application. The paper is illustrated by a few 

 specimen patterns, some of which are sufficiently 

 remarkable to merit close attention from those 

 engaged in decorative work. An interesting fact, 

 which does not appear to have been generally recog- 

 nised before, is that a plane can be completely tilled 

 with congruent quadrilaterals of any shape. 



NO. 



2752, VOL. I IO] 



