A very cold stratosphere is found often in front of typhoon, the tempereature of which is generally below -70°C. It is supposed that this cold air mass may be brought by strong wind around the typhoon from equatorial region. From those phenomena, the following rules for typhoon movement and the distribution of temperature in the stratosphere are derived. 1. A typhoon moves towards a cold area, avoiding a warm area. 2. A typhoon moves towards the tip of cold tongue in front of typhoon. 3. A typhoon has a tendency to move towards -70°C area. 4. A typhoon moves along a cold trough of temperature. 5. A typhoon recurves or fills up, if a warm air invades in its front. 6. If the temperature of the stratosphere in Japan is higher than -60°C, typhoons will not attack Japan.
It is already known that the observed terminal velocity of falling raindrops with a diameter larger than 1mm is considerably smaller than that which may be expected from the drag coefficient of a sphere. This fact is generally attributed to its deformation due to the non-uniform distribution of pressure upon the surface of the drop. In this paper, the author shows theoretically that, if the flow around the drop is a potential one and the deformation is small, the shape of a falling raindrop is an oblate spheroid whose axis is vertical. He then computes the fall velocity, on the assumption that the increase in cross area due to the deformation only causes the decrease in fall velocity, and that the drag coefficient of an oblate spheroid with small eccentricity is approximately equal to that of a sphere whose radius is equal to the major radius of the former. The computed values of fall velocity agree very well with the recent measurements of R. Gunn and J. D. Kinzer in the range of diameters less than 2mm. A more complete one of this paper in English is published in the Geophysical Magazine, Vol. 21, No.3, Tokyo.
On the rate of evaporation of water drops, the following semi-empirical formula was derived: where -dm/dt=rate of evaporation; D=diffusion constant; r=radius of the drop; Cs=saturation vapour density at the temperature of the drop; C∞=vapour density in the air; Re=Reynolds number; K=β/1-β/2√RTw/2πM; β=condensation coefficient of water; R=the universal gas constant; M=molecular weight; Tw=temperature of the drop; σ=D/υ; υ=kinematic viscosity of air. The formula is nearly equal to the Frössling's formula when r is large, and to the author's former formula when r is small.
In the atmosphere, there are condensation nuclei of various sizes, and the large hygroscopic nuclei do not need the supersaturation of vapour for condensation which occur on them. Therefore the supersaturation would be very small, if the rate of condensation is where R1 is the gas constant of dry air of 1 gram, x the mixing ratio of the air below the condensation level. The formulas (1) and (4) were solved numerically under the conditions (3') and (5), and the following initial conditions: r=r0, T=Tw=T0 at h=h0. (6) As an example, four cases in table 1 were calculated. The numbers 105 and 106 of n in Table 1 correspond to 107 and 1070 cloud particles in 1cm3 at the cloud base. In all the cases, h0=1000m, r0_??_0, T0=20°C. The result in the case II is given in Fig. 2. In this figure, abscissa is the height h, and the temperature Tw, T, the lapse rate of air temperature-dT/dh, the supersaturation C-Cs/Cs (Cs is the saturated vapour concentration at the temperature of air), and the radii of cloud particles are represented by full line curves. The results of all cases I-IV are given in Fig. 3. There are the curves of air temperature, supersaturation and radii of cloud particles. It is seen from the figures that the supersaturation is tabulated in Table 2. They should be in order of 3% in extreme cases, but may be less than 1% usually.