1. From the observations of swells made by ships, followingl results were obtained. Seven typhoons, 73 synoptic charts, and 1193 observations of ships were picked up as the data. 2. Within the typhoon, there are two kinds of swells, one progresses according to the wind, the other propagates from the center of typhoon to radial direction. 3. The swells of the first kind, which progress according to the winds, exist in every typhoon, but the swells of the second kind, which propagate from the center, exist only in one of seven typhoons. 4. The oscillation of the atmospheric pressure within the central part of the typhoon will cause the swells of the second kind. Most of the swells, caused by winds, run along the wind-direction within the typhoon-area.
During June 28th-July5th, 1938, Central Japan experienced a severe rainfall which is the heaviest ever recorded during any month since the Meteorological Observatories at Mito, Tokyo, Yokohama, Numazu and Kobe were opened. For several days prior to June 30th there had existed a stationary front running from northeast to southwest across Southern Kwanto district between cool air to the north and warm and moist air to the south. These conditions are shown on the chart for 0600 J. M. T. June 29th. During 28th-30th a typhoon continued to move ENE on this front. The full violence of the exceptionally heavy rains occurred during 29th. Rain commenced to fall few days before and was almost continuous until morning on the 30th During the whole period on the 29th 276.6mm of rains were recorded at Mito, 278.3mm at Tokyo, 268.3mm at Yokohama and 363.5mm at Numadu. The normal monthly total of rainfall in July in Tokyo is 132.3mm and the maximum fall for any one day hitherto recorded is 193.7mm. The 1200 J. M. T. Chart for June 30th showed that the centre continued to move ENE or NE passing just south of Tokyo. After 1800 J. M. T., as the typhoon moved farther away, the wind velocity fell and it was fine weather. One day later there formed a new front of stationary type running along Japan proper between cool air to the north and warm and moist air to the south. During the second period of the exceptionally heavy rainfall July 5th 270.4mm of rain were recorded at the Kobe Meteorological Observatory. As might be anticipated serious damage was done in Japan proper, mostly due to floods, 715 lives were lost and the total momentary loss from this storm has been variously estimated at ¥1, 000, 000, 000 to ¥2, 000, 000, 000. Full report on the heavy rainfall may be seen in the “Gou Hokoku” (253 pages with 32 plates), published by the Central Meteorological Observatory.
From the analytical data of rain water in Tõkyõ, Kõbe and Hamamatu, the writer obtained the following results. §Chloride. 1. The yearly averages. Cl mg/L, Tõkyõ 1.79, Kõbe 2.33, Hamamatu 2.39. 2. There is a marked correlation between chloride content and wind velocity. 3. There are group distributions in the chloride concentration as in the case of Köhler and Israël, however, the principal values are not the same as those obtained by the above authors; Tokyo: Cl=2.10×2nmg/L Kobe: Cl=1.89×2nmg/L Hamamatu: Cl=4.0×2nmg/L 4. The ratio, Cl/S, in rain water is larger than value of Cl/S in sea water, when sulphuric acid from coal can be eliminated. The possibility of the variation Cl/S and Cl/Mg (Lipp) is discussed. §Sulphuric acid. 1. Sulphuric acid content in rain water is larger in winter than in summer. The yearly average: Tõkyõ. 1.86 S. mg/L. 2. There is a parallelism between the yearly variation of sulphuric acid in rain and that in air. §Ammonia. 1. In the yearly variation of ammonia, the maximum exists in June; this is good agreement with that in air. 2. The yearly average; Tõkyõ 0.58, Kõbe 0.28, Hamamatu 0.19 N. mg/L. §pH. The yearly average: Tokyo; 4.1, Kobe 5.2, Hamamatu 5.6. 2. It can be theoretically calculated from the difference between the equivalent of sulphuric acid and ammonia. §Nitrite. 1. The nitrite content in rain is very small when compared with that in air. 2. The yearly average. Tõkyõ 6.7, Kõbe 4.0, Hamamatu 2.7 γ/L. 3. The seasonal variation (γ/L)
The investigation of the earth's magnetic variation produced by thunderstorms is hitherto confined to its impulsive changes due to lightning discharges, A statistical investigation now performed is to find a rather slow, or stationary variations due to thunderstorms visiting near the Kakioka Magnetic Observatory in the summer months. Most of them are heat-thunderstorms and their paths of travel are more or less regular, from the north-western part of the Kwanto-District to the east, or south-east, spreading over the ocean. Considering various modes of influencing mechanism, possible magnetic variations in the present case will be more conspicuous than those due to other kinds of thunderstorms. 212 thunder storms are obtained in the three summer months of June, July and August in the course of 1924-1933. If the mean hourly values of each element of the earth's magnetism of these thunderstorm days and those of the total summer days are designated by Ks and K0 respe_??_ively, the difference ΔKs=Ks-K0 will be responsible for the investigation of the effects of thunderstorms. The results thus obtained are as follow. (1) The mean daily range of the horizontal intensity on thunderstorms days is almost equal to that of total period, for example, 49.Oγ for the former and 49.1γ for the latter in ten years' mean, indicating that the activity of the earth's magnetism on thunderstorm days is normal. (2) Various kinds of Statistics show the similar type of the diurnal variation of ΔKs which is a single wave with a maximum between 16h and 18h in local time, while the diurnal variation of the horizontal intensity H has a maximum at 14h, 13h for the declination D and 5h for the vertical intensity V. (3) The diurnal variation of ΔKs of H, or D is very similar to that of the frequency curve (P) of thunderstorms observed at the observatory, while V has generally the same tendency as the other elements, but markedly decreases near the maximum time of (P), or H and D, when the thunderstorm passes overhead. It is reasonably expected that thunderstorm effects upon the earth's magnetism become larger with increasing (P). (4) From the different periods of years, the absolute amplitudes of the diurnal variation of ΔKs vary in the range of 1γ_??_3γ for H, 2γ_??_5γ for V and 0.'2_??_0.'4 for D. (5) The absolute amplitudes of ΔKs for the sun-spots maximum years of 1926-1930 is 2 or 3 times larger than that of minimum five years of 1924, 1925, and 1931-1933. (6) Each horizontal component of ΔKs for different groups of thunderstorms are almost perpendicularly to the direction of thunderstorms most frequently observed at this observatory; that is, in the other words, if the magnetic field of ΔKs is derived from the earth-current flowing approximately uniformly around the observatory, the current flows into, or out of the thunderclouds. (7) From ten years' mean, for example, the above current flows into the cloud. It is, therefore, resulted that the earth has gained an excess of negative charges from the thunderstorm in the course of ten years, and so-called Wilsontype of thunderstorm has been frequently observed, provided that the quantity of electric charges of the individual thunderstorm is approximately equal. Under reasonable assumptions, the order of magnitude of the current flowing into the clouds is estimated as 0.2amps./km2, which is the same order obtained B. F. J. Schonland and F. J. W. Whipple from their atmospheric electric researches.