1) The polar continental air-mass originates in Manchuria and Siberia and comes to the Japan proper as the Northwest Monsoon and further reaches the Philippines as the Northeast Monsoon. The tropical air flowing out from the Pacific High Pressure comes to the Japan proper as the southerly tropical air and reaches the Philippines as the northeast trade wind. The SW Monsoon appears on the stage from April on even to November and December in the Philippines. Doubts may arise as to the exact propriety of the word “monsoon”, but in this paper we will use it with Deppermann to include all the air coming to the region apparently from the southern hemisphere. There are really such three air-masses in the equatorial or tropical region. Deppermann(1) named the front between the SW-Monsoon and the trade wind or the northers as the “Equatorial Front.” The front between the tropical air and the northers is naturally the polar front after J. Bjerknes. A'most all the typhoons originate at the Equatorial Front. They first travel westwards along the front, until they reach the “Dreimasseneck”(2), at which we find intense interaction between the SW-Monsoon, Tropical air and the Polar air. After a little beyond this point, they often recurve and travel to the northeast along the polar front. Very often they steadily travel to westwards beyond the said point along the Equatorial Front between the northers and the SW Monsoon. The location of the Pacific fronts is determined on the ground that the axis of the region of greatest cyclone frequency indicates the mean position of the front (Fig. 1). The smoothed isarithms on this map represent total typhonic frequency during the 5-year period ending 1936. These frequencies were tabulated for bimonthly intervals (September-October) and were plotted for each 2° square of latitude and longitude. The original data were obtained from the weather maps analysed by the Central Meteorological Observatory, Tôkyô, Japan. The marked relation between fronts and two types of typhoon's track, of which it often seems to be difficult to get a clear mental picture, is illustrated by Fig. 1 in quite satisfactory manners. Figure 2, depicting yearly air masses and fronts in the North Pacific, is a graphic representation of the said conclusions derived from examination of the maps of requency of typhoons. 2) Typhoons in the North Pacific sometimes recurve in the vicinity of the tropic or a little beyond it During the recurving they increase in strength and their progressive movement slackens. After the recurving they rapidly travel eastwards but decreases in strength, Depressions in the North Pacific generally recurve in the vicinity of Aleutian Islands. During the recurving they increase in strength. After the recurving they die out in the stage of occlusion. Fig. 3 is based on the model employed by the author and illustrates the application of his theory, whose underlying principle is the Kinematical Analysis of the Field of Pressure(1). 3) There are two classes of phenomena of föhn for which the aid of Ficker's theory must be invoked. On the southeastern side of the Japan proper (Pacific coasts), the föhn of relatively cold winds are observable in the colder half of the year during which the northwesterly monsoons prevail. On the northwestern side of the Japan proper, the föhn of very warm, dry winds are observable in summer during which the southerly monsoons prevail. By the usual theory of föhn, it can hardly be possible to explain the extreme hotness 40.8°C observed at the Yamagata Observatory on July 25, 1933. In order to discuss this phenomena, aerological materials on the windward side (Fig. 6) are carefully analysed. It is concluded that in the upper atmosphere a marked inversion had been observed in the layer which extends from 500 metres.
This memior is essentially a sequel to one whic one of the authors published lately in this journal. Intergrating Rayleigh's well-known thermohydrodynamical equations (Scientific Papers, Vol. VI, p. 436), the main characteristics of land and sea breezes are deduced. Stream lines, and vertical distribution of the horizontal air-current associated with land and sea breezes owing to the nature of the coast line are shown in Figs. 1-3. There is a good agreement between the theory and observations, as we could expect in such computations.
According to the statistical research by European investigators the travelling cyclone and anticyclone may be divided in two classes. One is called high type, and includes all the travelling disturbances which reach the height of the stratosphere. In this type the height of tropopause is lower over a cyclone and is higher over an anticyclone than the normal state. The other is called low type and its height is generally confined to the lower level of the troposphere. In an area of cyclone of high type air temperature is lower than the normal environment in the troposphere and this circumstance is reversed in the stratosphere. On the contrary, in an anticyclonic area of high type air temperature is higher than the normal distribution in the troposphere and reversed in the stratosphere. Thus the difference of air temperature of, a high cyclone from a high anticyclone is conspicuously marked, therefore the former may be called a cold cyclone and the latter a warm anticyclone. This relation is reversed in the stratosphere, but the three-quarters of the whole atmosphere belongs to the troposphere, therefore the air temperature of high anticyclone is definitely higher than that of high cyclone. The cyclone and anticyclone of low type, roughly speaking, have warmer and colder air temperature respectevely, whose type of structure is considered to directly induce the lower and higher pressure anomaly. In short we may say that a cold anticyclone and a warm cyclone belong to the low type and a warm anticyclone and a cold cyclone belong to the high type. The above description concerns the travelling disturbance. Here the author intends to extend the above classification concerning the travelling disturbance to the cyclone and anticyclone of stationary type, such as the equatorial low, the subtropical anticyclone, the low of middle latitudes (locates at about latitude 60°), the polar cap and monsoon high and low. The stationary pressure anomaly may also be classfied as high and low type, and warm and cold type. The classification due to the present scheme is as follows: the equatorial low is cold and high; the subtropical anticyclone is warm and high; the polar cap (if exist) is cold and low; the low of middle latitudes is cold and high; the low and high of monsoons are warm and cold respectively and belong to the low type. We have already learned that over the travelling disturbance of high type there occurs the fluctuation of height of the tropopause. It is generally supposed that such a change of tropopause-level has some important relation with the mechanism of pressure anomaly from the study of travelling disturbance. The extension of this analogy to the stationary cyclone and anticyclone may naturally lead to the provisional conclusion that the inclination of the tropopause has some important relation with the pressure anomaly of high type. In the light of observational fact we see that over the subtropical anticyclone the height of tropopause is about two times greater than that of the low of middle latitudes and the distribution of air temperature has the same tendency as that of travelling disturbance both in the tropopause and in the stratosphere. Thus the analogy between the travelling and stationary pressure anomaly holds fairly good within the scope of the present aim.
The diurnal variation of the barometric pressure has already been investigated by many authors. The present author investigated the diurnal variation of the barometric pressure at Misima and Hakoneyama and its realtion to that of air temperature.
The Present author investigated the variations of the terrestrial magnestism, the atmospheric electricity and the earth-current using the data obtained at the Kakioka and the Toyohara Magnetic observatory. The diurnal variation of the atmospheric potential gradient indicates the existence of the longitudial current flow in the lower strata of the aimosphere, and the variations in the in this currcntsystem may cause the induction current in the earth's crust. The existence of this induction effect was verifed with magnetic records. From the short period variation, we can also deduce some similarities between the variation of the terrestrial magnetism and the earth-currert.The bay-disturbance in the terrestrial magnetic field with sufficient results.
The present authors have analysed the records of 30 magnetic storms of moderate intensity recorded at the Kakioka magnetic observatory, Japan, following the method uscd by S. Chapman. The horizontal force clearly shows the initial increase of small intensity and then it decreases far greater and about 40 hours after the commencement of the storm it begins to restore the normal value. The vertical force undergoes a similar changes as that of the horizontal force inverse in direction and smaller intensity, while the declination seems to show no such a marked change. Moreover we have calculated average diurnal variations on the calm days and the additional mean diurnal variations during the three epochs, namely the first day (6h-30h) after the commencement) the intermediate day (18h-40h) and the second day (30h-54h).