In 1889 A. Schuster showed that daily magnetic variation could be divided into the magnetic variation coming from an external origin and that of internal origin. S. Chapman has concluded that the latter can be perfectly explained as an electro-magnetic induction of the former into a model earth having the following property: a uniform distribution of electric conductivity throughout a sphere concentric with the earth, but of radius 4 per cent. less than of the earth, the material in the outer 4 per cent. layer, about 250km thick, being non-conductive. The probable cause of the daily, magnetic variation of external origin has been believed to be some electric current flowing at some level above the earth's surface. The features of this current system were investigated by Bartels with great success, but the location of this current system still remained unknown. Recent investigations of the radio fade out showed us that this phenomenon is generally accompanied by the augmentation of the amplitude of the S field and that the radio fade out is due to the sudden increase of the ionization of the lower part of E layer, and possibly also in F1 layer. The present author deduced, with considerable success, new formulae to calculate the electric current and the vertical components of the magnetic force from the horizontal components. The horizontal distribution of the electric current obtained by the author's method coincides very well with that obtained by Bartels as shown in the figures. The vertical components calculated by the author's method explain the observed vertical component in a rate of about 75% and 55% for the external and internal magnetic variation respectively. Thus referring to the values obtained by Chapman of the location of the internal current system, we can evaluate the location of the external current system, with the result that it can be considered, without large error, to be E or F1 layer. Namely the situation of the external current was, first, given in this paper from the observed values of the magnetic forces and this result coincided very well with that estimated on the basis of wireless researches. Thus the result of this paper seems to be a valuable contribution and a strong support to the reconstruction of the theory of daily magnetic variation and also to the physical investigation of the ionosphere itself.
The places the data of which were analysed are Peking, Nanking, Hong-Kong, Manila, Batavia and Melbourne. Information regarding the sources of data is set out in the foot notes. In figs. 1-3 are shown the monthly and annual normals of temperatures at 0.5, 1, 2, 3, and 4km. above sealevel over Eastern Asia. In figs. 4-5 are shown the sketches of monthly and annual normal temperature distribution in the atmosphere of Eastern Asia. The contrast in mixing ratio is more vivid than the contrast in upper-air temperature. In northern summer, mixing ratio is highest at all levels in Hong-Kong and Nang-king and this is caused by the summer monsoon which is hot and humid.
A. Grimes once discussed the movement of air across the equator, assuming that the earth is flat and that the approximation sinφ=φ (φ=latitude) holds good. The present author treats here the same problem without such assumtions. The equations of motion on the rotating earth are, in customary notations, The above equations may be much simplified by making the assumptions such that the vertical velocity υr is nil, the motion is steady (∂/∂t=0) and zonal (∂/∂λ=0) and the air is incompressible; thus Here the radius vector r is replaced by the radius of the earth a. The meridional component of velocity υθ is given by, from Eq. (3), and the longitudinal component of velooity υλ is derivable from Eq. (2): Eq. (5) expresses the law of conservation of angular momentum and plays an important role in the veering of wind direction in traversing the equator. In order to inspect the motion of air near the equator, θ0 is taken as 95° (5°S) or 100° (10°S) and the motion towards the northern hemisphere is treated, therefore the first term in the righthand side of Eq. (5) is always negative in the neighbourhood of the equator. Thus, when υλ0 is negative (υλ0<0 means the easterly wind), the wind is easterly up to the latitude φ0 (θ0=90°+φ0) in the northern hemisphere and then becomes westerly, while, when υλ0 is positive, two different cases occur according to the magnitude of υλ0. Namely, when υλ0 is small, the wind veers from westerly to easterly in the southern hemisphere and traverses the equator with easterly component, but soon tends to westerly over a certain latitude φ (φ<φ0). When υλ0 is large, the wind maintains its westerly component and suffers no essential change across the equator. The above three types of track are given in Figs. 3_??_6.