Palmén (2) has shown that many “occluded” cyclones on surface maps have never gone through a real process of occlusion although they show the same characteristic structure as really occluded polar-front perturbations. Here a well-marked surface front is not so essential for the development as was generally assumed formerly. On October 8, 1951, the Central Meteorological Observatory, Tokyo, discovered evidence of a tropical disturbance in the formative stage south of the Guam Island. The disturbance soon developed into a typhoon whose center crossed Japan in the mature stage of development. Surface maps, as well as charts for the 500- and 700-mb surfaces are shown in Figs. 1 and 2, which are reproductions of charts prepared by the Central Meteorological Observatory, Tokyo. Fig. 1 presents six surface maps for October 14, 15 and 16. At the beginning of the period there is no surface front connected with this typhoon RUTH. There is one eastern front corresponding to the northern boundary of the moist warm Tropical Maritme air. On the chart for October 14, 1500 JMT a surface cold front extending to the southwest from the center is drawn, while the eastern front corresponding to the northern boundary of the Tropical air remains nearly stationary. On the chart for October 14, 2100 JMT two surface fronts have already been brought so near each other that a separation is almost impossible. At that time a structure corresponding to the classical picture of wave-shaped frontal cyclone has already started to develop. The surface typhoon of October 15, 1500 JMT has all the characteristics of a polar-front cyclone in spite of the fact that it did not develop from a wave cyclone. By comparison of the consecutive 500- and 700-mb charts in Fig. 2 it can be seen that the area of warm air at these level increases during the developement of the “occluded” surface typhoon. Therefore, in the upper atmosphere the warm air gains area over Japan, while in the lower atmosphere the cold air gains area.
The author deals with the vertical diffusion width of smoke in the wind near the ground. At first the diffusion phenomena are clearly separated in two states, i.e., the transient and the ultimate state, and also in two zones, i, _??_., the I-class and the II-class zone. These separations are shown to be possible by means of the mutual comparisons of the Lagrangian averaging-time and the life-time of the absolute largest turbulon in the flow, and also of the Lagrangian process-time and the life-time of the effective largest turbulon. As to the ultimate state of the vertical turbulent diffusion of smoke near the ground, the theoretical results are compared with recent F. A. Record's (1951) experimental ones in general agreements. On the other hand, concerning O. G. Sutton's (1951) recent question about the interrelation between the smoke width and the averaging-time the author presents a certain reasonable explanation making use of the concept of the transient state of vertical diffusion.
“Slice Method” for determining the stability or instability conditions of the atmosphere was discussed. As the result it was found that the treatments by J. Bjerknes, S. Petterssen and S. Syono were not correct. By dynamical treatment of the problem the author obtained the following results. (1) In the cases of non-saturated air and also of saturated air, stability criteria and solenoid-producing energy derived from the parcel method hold also for the slice method. (2) In the case of saturated ascent through a dry-adiabatically descending environment, stability criteria and solenoid-producing energy are slightly different from those by J. Bjerknes, etc. The force realized in the ascending slice M1, by small displacement _??_z1, was tabulated for various cases and methods.
In this paper, the author deals theoretically with the problem of ice formation, in view of extending his previous study, and the formula which approximately represents the speed of the growth of ice layer is derived as where ε0: thickness of ice layer, t: time measured from the beginning of ice formation, U( ): air temperature, at a point sufficiently far from the ice surface, k: coefficient of heat condution of ice, ρi; density of ice, L: melting heat of ice, α: heat transfer coefficient. Although this formula is derived as an approximate solution of the differential equation of heat conduction in an ice layer, it is also made clear that the accuracy of this approximation is very satisfactory. The good agreement between this theoretical result and the experimental data recently obtained by K. Neumann verifies also the validity of this formula. In the latter part of this paper, the relation between α and wind velocity is discussed and it is shown that the transfer from the surface of ice into the air can be reasonably explained, provided that we take the existence of a turbulent boundary layer near the surface into consideration.