In Ser. 2, Vol. 3, No. 3 and No. 7 of this journal, the present author has found that the velocity of condensational seismic wave propagating in the direction perpendicular to the arc of the Japanese Islands is greater than that of the wave propagating along it. He explained it from the difference of elastic properties in these two directions. For brevity's sake, we call this phenomena “elliptical propagation of seismic wave” at Prof. Okada's suggestion. In the present paper, the author confirmed the reality of this phenomena of elliptical propagation by the investigation of some local shocks occurred on the plain of Kwantô. He has found that this is also the case with P isochronals. The isochronals of a few local shocks and of the great destructive earthquakes of Sept. 1, 1923 are shown in Figs. 1 to 6. Nextly, the P isochronals and iso-PP lines i.e. the lines which connect the stations having the equal durations of PP are drawn and shown in Fig. 8. Also the time curve of the great earthquake of Sept. 1, 1923 is drawn with the most accurate data reported to the Central Meteorological Observatory from 96 meteorological stations and is shown in Fig. 7. Froin the above considerations the following conolusions have been drawn:- 1. In this country, P wave itself shows the phenomena of elliptical propagation when propagating in a portion of the earth crust. 2. Elliptical propagation is the phenomena due to the anisotropic nature of the earth crust under consideration. It originates by the bending of our great island toward the Pacific Ocean. According to this bending our great island is stretched along the Pacific coast side and consequently compressed on the Japan sea side. This causes an increase of elastic constants perpendicular to the arc of the island than that along to it. 3. From the examination of the P isochronals and iso-PP lines, it seems that in the portion of the lithosphere lying under the well-known discontinuous surface at sixty km. beneath the surface of the earth, the phenomena of elliptical propagation does not exist or if exist, it may be in a lesser degree. This is due to the fact that the layer has the higher values of elastic constants.
According to Professor Okada there are three circumstances giving birth to travelling anticyclones in the Far East, i. e. (1) Two or more cyclones appear simultaneously and high pressure develops in their intermediate region. (2) Nocturnal cooling on the land surface. (3) Byproducts of the general circulation of the Earth's stmosphere. In the present note the above circumstances are explained with the aid of same weathermaps. The travelling anticyclones are numerous in spring and in autumn and those appear in these seasons are more persistent than those in winter and summer. In the latter seasons according to the strong contrast between the continent and ocean of their respective temperature, the high pressure always situates on one of them and the belt-form is not remarkable. On the contrary in spring and in autumn high pressure belt becomes distinct and detached highs are liable to be formed. Figs. 1 and 2 show the development of detached highs cut off at the neck. of wedge-shaped high area. In Figs. 3 and 4 we see nearly the same circumstance as in Figs. 1. and 2, and also the outbreak of cold wind from the western Continent dropping off as a detached high. Fig. 5 shows the development of an auticyclone between two cyclones and Fig. 6 that of anticyclones surrounding a cyclone.