In the previous paper under the same title, the results of model experiment by Mr. Matuo concerning the tunami in the Osaka Bay was compared with results calculated from hydrodynamical theory. but it scarcely succeeded in bringing both in good agreement. In this paper the amplitude and time of occurrence of lst wave are newly measured from the records of model expriment (Fig. 1), and it is tried again to explain them theoretically. The observed travel time is in good agreement with what calculated as velocity is √gS/b, where S, b are the sectional area and the breadth of the Bay respectively. We regard the tunami as a long wave and put for its hydrodynamical equation under the influence of frictional force and put (that is velocity) then for f we have From this equation f is evaluated by numerical integration (table 1) adopting the observed values β given in Table 2. The results follows the observed curve closely enough (Fig.3). The values of B(=βT/2π) (table2) for the model experiment are very large and it is not considered to be the case in the nature. The β for the tunami which occurred in the Tokyo Bay at the great Kwanto earthquake in 1923 is evaluated from the then mareogram at Yokosuka (Table 4). We see that it is very small compared with that of Table 3. The f for the tunami in the Osaka Bay under this value β is calculated by numerical integration by the equation (1). The results are illustrated in Fig. 5 in which the mode of dissipation for the wave energy are also shown.
Winds or currents at strait of Gibraltar or Simonoseki strait, whose orientation is eastwest, are prevailingly East or Nest, seldom North or South. The most noteworthy is the marked localinecrease of the wind force along the coast lying to the right of the general trend of the wind. The deflective force of the earth's rotation causes a streamline convergence along such a coast, which must be compensated by an acceleration of the wind if ascent of air shall be avoided. The north coasts of these channels will therefore be subject to strong east winds, but will have subnormal west winds (Figs. 1a and lb). Along the southcoast conditions will be intermediate. On the other hand, local cyclonic wind system is often observable there (Fig. 1c). It is well known that cyclones weaken while crossing our islands. We could list the following possible causes: (1) The mountains may have an impeding effect on the circulation. (2) The land friction, being more than at sea, naturally would tend to decrease wind velocity. (3) The supply of water vapor must be somewhat diminished and thus slacken the speed of condensation of water vapor. Even though the Soya, Tugaru and Tusima straits are half land, still these effects must be somewhat diminished, so shat cyclones have a marked tendency to pass over these channels. The effects of local cyclonic wind system will also contribute, fascinating task to explain the real cyclone paths. A study of cyclone paths might yield results of interest. The location of the principal cyclone paths is determined on the ground that the axis of the region of greatest cyclone frequency indicates the mean position of the path (Fig. 2). The smoothed isartihms on this map represent total cyclone frequency during the period 1938. These frequencies were plotted for each 2° square of latitude and longitude. The original date were obtained from the weather maps analysed by the Central Meteorological Observatory, Tokyo, Japan. The marked tendency that cyclones pass through the Soya, Tugaru and Tusima straits is illustrated by Fig. 2 in quite satisfactory manners.
The author of the present paper studied theoretically the variation of temperature in a closed room of a building when the temperature of the outer side of the building varies periodically, and compared the result with the observed records. The form of the building in a rectangular parallelopiped and it has only one room. For the sake of simplicity we reduced the problem to a case of one dimension. The temperature in the wall υ1 and that of the air in the room υ2 was obtained by solving the following equations. provided that where α denotes the breadth of the wall, k1, k2, thermal conductivity of wall and air respectively. h an experimental constant depending upon the surface conductivity of the wall. It was also assumed that the temperature of the outer side of the building is equal to that of that of the air temperature. The observed results was obtained by a thermograph in the period from August to October, 1938. The coincidence between the results obtained mathematically and those obtained from the observation was good.