The main part of the variation recorded every day of earth-magnetism. or earth-current may be generally supposed to be an intermixed result by all agencies from the sun of which ultraviolet light is said by some authors to cause the diurnal variation and charged or neutral corpuscular rays mainly to relate to the occurrence of the disturbance. Then the disturbance must deform the diurnal variation of usual days. To know this effect of the disturbance of earth-current, we constructed the Difference-Curve_??_[Diurnal Variation on Disturbed Days]-[Diurnal Variation on Calm Days] for the east-component-larger variation direction at this place-in the period of Sept. 1932 to Dec. 1934. The Difference-Curve-call it the Curve for short hereafter-is obviously a single oscillation against the double one of the diurnal variation, having max. at about midnight and min. at early afternoon when the westward current is taken as positive. The portion from the negative part of the Curve to the positive part; that is, from day to night, is steeper than the opposite portion. The transition points between the negative and positive parts of the mean Curve in the full period are 5.0h and 18.5h (135° E. M. H.) which are about 40 minutes earlier, or later than the sun-rise and sun-set on the ground at Tokyo. This may suggest some relationships between the solar effect upon the upper air, say ionized layers, and the charge of the diurnal variation on disturbed days. A seasonal variation of the Curve is seen by the largest amplitude in Summer (May, June, July, Aug.) and smallest in Winter (Nov., Dec., Jan., Feb.), while the diurnal variation has the max. amplitude in Spring (Mar., Apr., Sept.. Oct.) at this place. Next, supposing the most disturbed interval of each disturbed day is T hours, of which N hours belong to the night part (18h to 6h) of the day, we classify the disturbed days into two cases; The Curve of (ii) has a smaller amplitude with 70% of that of (i), and a small phase difference which decided conclusion desires more data. This result means in the other words that the diurnal variation on days having many hours of disturbance in the night part of the days is not so differed from that on calm days, compared with that of disturbed days of which larger part of the disturbed hours belong to the daytime. The similar Curve for 1933 of earth magnetic horizontal component shows also a single wave such as at some other places, and the same value of R=amp of the Curve/amp. of the diurnal variation on calm days=0.7with that of earth-current for the same days, but larger phase difference compared that of the diurnal variation on calm days. he vector difference between the resultant diurnal variation current on calm and disturbed days are calculated for 1934 and expressed by a fish-bone figure. Its absolute value and the angle θ between the diurnal variation current on calm days and this difference current are also shown. This result seems to show that the effect in daytimes by the disturbance upon the diurnal variation is somewhat differ from that in night part. The interpretation of the resuls is not so easy in the present state, but the author proposes a possible suggestion th_??_t the main cause of the effect of the disturbance obtained above may not be hidden so far outside the earth's atmosphere, but in the upper parts of the ionized regions, of which lower part, for example the E-layer, is now supposed to be the main domain of the diurnal variation on calm days. The ionization of the supposed parts may be greatly increased in the mean state on the disturbed days by some agencies from the sun, w_??_ether radiations, or corpuscles may they be; and also hea_??_el or dissociated by the sun's radiation, inducing an anomalous current variation.
This memoir is essentially a sequel to one which I published recently in this journal with the same title. The following articles are studied:- (1) Atmospheric cross sections. A picture of atmospheric conditions in the vertical is urgently needed to supplement the present practice of representing the separate aerological soundin_??_s individually on one forms of diagrams. For such a cross section we choose a line, along which there lies a maximum number of airplane stations, as the base line i.e., the line of intersection of the vertical plane with the ground surface. In the long east-west section, we have Kasumigaura, Tatikawa, Yôkaichi, Oomura and Nanking. As for example, we illustrated the atmospheric cross section of the modified polar continental air mass, which los_??_s its original col'ness and dryness in the lower layers and is usually characterised by marked inversions, probably due to subsidence, above which the accumulated pollution and moisture cannot rise. The surfaces of subsidence inversion are extensive domes which may at times re_??_ch beyond 4 km level and often practically intersect the ground surf_??_ce along their pberiphery in rear of the polar continental air mass. (11) Atmospheric turbidity. In Fig. 10, the turbiditiesfor Ootomari, Zinsen, Morioka, Siomisaki, Hôkotô, Ogasawara, and Palao have been arranged. At all stations the polar continental air has the lowest turbidity but the turbidity increases rapidly as the polar air proceeds to the south-wards. And at all stations the tropical maritime air has the highest turbidity and the turbidity is nearly constant everywhere. These results are in agreement with our knowledge concerning the source regions and vertical structure of the air masses. Again the all stations show a seasonal trend in turbidity. The polar air has the lowest turbidity in the winter and highest turbidity in early spring, probably due to originating and moving over comparatively bared ground surface regions. (111) Potential gradient (Atmospheric electricity.) The polar continental air has generally higher potential gradient and the maritime air lower potential gradient. But the po_??_ential gradient show a marked seasonal change, so that the above conclusion can not be applied for indivisinl cases. It is found that the polar continental air has the highest potential gradient in winter, during which the NW-ly monsoons prevail over the Japan proper. And the maritime air has the highest potential gradient in summer, during which the tropical maritime air masses prevail over the Japan proper. Generally the potential gradients observed show a gradual increase as a front or a cyclone passes away, corresponding to the transformation othe air mass type from the fresh one to the mof dified one. This phenomena is remarkable in the colder half of the year. As the Polar continental air is modified, it is characterized by marked inversions below which the pollution and moisture is accumulated. Probably the gradual increase of the potential gradient is due to the existence of such haze layer.
From the observations of the potential gradients in the atmosphere we see that there exist the positively charged space charges and their quantities diminish with height very rapidly. The seasonal variation of the space charge near the earth's surface was treated by K. Kähler and others. According to the existence of the space charge in the atmosphere there should flow the ion currents of conduction and convection natures; the former its produced by the potential gradient and the conductivity of the air, while the latter is due to the flow of the net quantity of the space charge by the wind. The formula showing these relations is given by where i is the current; λ, the conductivity; _??_, the wind velocity; _??_, the potential; p, the space charge and these two quantities are related by the Poisson's formula: In this paper the present author considered, first of all, the variation of the space charge with season and height using many observed data by various authors: where a and b are given by (4) and T indicate_??_ the air temperature, being assumed to be expressed by a series of spherical harmonics of various orders as is given by (6). (p. 340) Then the Poisson's equation is easily solved as was given by (8). Now we a_??_sume that the negative charge is distributing on the earth's surface and the same quantity of the positive charge is wholly contained in the earth's atmosphere and hence at some altitude from the earth's surface the potential should be zero. This assumption is expressed by (9) if R1 indicates the height of the zero potential. This height is assumed to be the upper boundary of the troposphere. Thus we have the complete solution of the Poisson's equation as (10) and (11). Whence we can express the ion current by the expressions (12). In this paper the conductivity of the atmosphere was assumed as (13) and the thickness of the atmosphere was considered to be very small as compared with the radius of the earth. By the above assumptions and the next algebraic relations: if kn=A1K+A2k(k-1)…+Ank(k-1)…(k-n+1), we can solve the ion current at various cases after the elaborate calculations as (18) for the case when the atmosphere is at rest (21) for the case when the temperature is expressed by (22) for the case when the temperature is expressed by By the above expressions for the ion currents we can theoretically deduce the genraal feature of the current as follows using the numerical values of the air temperature on the earth's surface. (1) Ion current due to the convection process; this is expressed by the components (24) indicating that: (a) The horizontal componeat of the current is very small as compared with the vertical component and the same result is deduced as to the potential gradient. (b) The vertical component is generally of the same order of magnitude coinciding with the observations and this is the greastest at the polar regions and the smallest at the equatorial belts so far as we consider the values of the same altitude concerns, or in other words, it is the greatest in winter, while the smallest in summer in the northern hemisphere and in the southern hemisphere the feature is reversed. The same conclusion is deduced as to the potential gradient also. (c) The ion current of horizontal direction is zero at the earth's surface and the upper boundary of the atmosphere and maximum in the intermittent altitude. This flows to the thermal equator from the polar regions and its intensity is the greatest in the middle latitudes of both hemispheres. (2) Ion current due to the conduction process; this is expressed by the components (27) indicating that:
On Oct. 1st, 1918 a severe typhoon passed through the western side of the Tôkyô city. On this account some parts of the coast of the Bay were attacked twice by high waters of Tunami(1) and the time interval between the first and second attack was observed as 90 minutes. These Tunami waves were already considered by some authors as the proper oscillation of this bay, for instances Prof. S. Nakamura as the secondary undulation of the 3rd order and Prof. F. Omori, a sort of a secondary undulation of the Sinagawa Bay, which is one of the inlet of the Tôkyô Bay. But the author's opinion is somewhat different from the above two, for he considers it as a lateral oscillation of the Tôkyô Bay. This consideration is made by following facts that The tunami waves attacked Yokohama and the mouth of the Yedogawa River at the same time where both situate at the western coast of the Tôkyô Bay, and moreover the author has obtained theoretically the period of the lateral oscillation to be about 90 minutes, assuming the bay to be a rectanglar lake with length of 53'8km. breath of 22km and depth distribution to be h=h0(1+x/a) where h0=26m, a one half of length. The period thus obtained just coincides with the observed one. Besides the lateral oscillation, it seems to exist then a longitudinal one, which came as a solitary wave at the northern coast of the bay. Next, the author has tried to obtain theoretically the hight and the time of occurrence of Tunami at some places of western coast, assuming that these oscillations be cansed by daynamical effect of suction due to travelling atmospheric depression. The results thus obtained show that 1) when the typhoon passes the western side of the rectangular bay parallel to its length, the arrival of the longitudinal wave at the northern coast is delayed by 1/2(Ta+a/c) compared with the time of the entrance of the center of a typhoon in or beside the bay, where Ta is the period of the longitudinal oscillation of the bay. 2) and the arrival of the first lateral wave to the western coast is delayed by 1/4T_??_ compared with the time of entrance of the typhoon center in or beside the bay, where T_??_ is the period of the lateral oscillation. These results also agree considerably well with observed fac's.
It is generally noticed, that there is seven days period in the atmospheric temperature, say, the local word “Sankan-Sion.” (the meaning of it is “Three cald days and four warm days”) The existence of such a period and the mechanism of it were researched by means of the correlation method proposed by one of authors and K. Husimi, and the result of which summerized as follows. In practise there exists a free oscillation of seven days period in the atmospherfc temperature caused by the coupling of temperature and pressure of the atomsphere, and the above period is recalled by the resonance of such oscillation and the over harmonics of inner effect on pressure.