We have picked up the 1053 data of pilot balloon observation which reached above the height of 2000 meters at the Maebashi Weather Station from July 1932 to April 1942, and carefully divided the types of their paths into 21 classes. We have, also, chiefly examined the form of the winding of paths, statistical characteristics of wind direction and velocity above or below the discontinuous surfaces, and further arranged the corresponding weather frequency tables for to-day, tomorrow and day-after-tomorrow in three classes-fair, cloudy, rainy-after four seasons in the year. Considering the local, topograph, or daily change effects for the paths, we have obtained the following characteristics:- a) clockwise paths of balloons largely occur in the region of rising barometric pressure behind cyclonic area and the attending weather becomes better step by step, but the sharp curve on a discontinious surface may result in bad weather on tomorrow and day-after-tomorrow. b) anticlockwise paths generally result in bad weather excepting the case of summer, when, sometimes easterly winds prevail up to higher levels. c) when the southly components of wind in 2-4km layers are increasing, bad weather is expected. d) it is interesting to remark that the same weather map does not always correspond to the same path.
Using weather charts published during 1940-1942 the relation between the barometric gradient and the wind direction or wind velocity was investigated by dividing the case that a cyclone is over the Japan Sea in five and the case that a cyclone is over the Pacific in two.
We investigated the sea fogs in the Okhotsk Sea side and the Pacific Ocean aide of the South Kurile Islands and it was made clear clear that these fogs differ entirely in characters. The Eastern side (the Pacific): 1. The fogs resemble neatly with those at Nemuro and Kushiro. 2. The size of fog drop is generally large and is felt like rain. 3. The relative humidity is high, and the fogs are wet. 4. The fogs come from the direction between NW and SW, except the local one which comes from the direction between N and W. 5. The duration time of fog is exceedingly long from which it is supposed to be very stable: 6. They appear from May to October. 7. The height of fog is 1000m at most and it becomes more higher on land than on sea. The Western side (the Okhotsk): 1. The fogs resemble nearly with those at Abashiri. 2. The size of fog drop is generally small and in most cases it can not be perceived as a drop when it is falling. 3. The fog is dry, and in many times the air is not saturated, even in case of the densest fog. 4. The fog comes from the direction between W and N, except the local one from the Pacific Ocean. 5. The scale of fog is comparatively small and the duration time is short and it becomes an unstable Si Cu cloud when it goes on land from sea. 6. They appear from May to July. 7. The height of fog is below 200m, and lower on the sea. Physical consideration The Pacific Ocean side: 1. The fog is made on few cases over 20°C. 2. In 48 examples tested, 25 ones have duration time of fog more thap. 24 hours, and the longest one was 5 days. 3. The densest fog is perfectly saturated. The Okhotsk Sea side: 1. The fog is not made over 22°C. When the temperature is abnormally high on land, the Log on the sea does not go up on land and stagnates on the neighbouring sea. 2. The size of fog drop is 5-20 μ on the whole. 3. In 64 examoles tested, 6 ones have duration time longer than 12 hours and 37 ones 6 hours. Forecasting consideration Forecasting of sea fog is possible by noticing the variation of pressure, temperature and sea breeze. There are few cases of fog due to sea or land windy but the sea fog made by wind only is small in scale and disappears soon. Land breeze fog made at one mid-night of October is stated simply.