One of the most important factors for the yearly yield of paddy rice is weather, but in Japan Islands lying lengthily from south to north climatic elements controlling the yields vary naturally with regions. Thus, in the cold district, the northeast part of Japan, low temperature in summer will be the most harmful as the paddy rice is suited for warm climate. But in the southwest part blessed with comparatively good climatic conditions it is difficult to make clear the controlling climatic elements because of their complex interactions. In this part [II], the author intends to choice climatic elements which give the most remarkable influence on the productivity of paddy rice and to analyze the degree of their interactions for Tohoku and Kyushu districts selected as the representatives of cold and warm districts respectively. /1. The climatic influence upon paddy rice productivity in the cold district (Tohoku)/ It has been shown by many investigators that the productivity of paddy rice in the cold district depends primarily upon the temperature in summer. But in order to obtain quantitative relations between the productivity and climatic elements, it seems necessary first to make clear the definition as to the limits of good and bad harvest of paddy rice. Thus the author calculates deviations of the actual yield from the estimated one due to the above logistic curve, transforming them to those given by a normal distribution, defines the positive and negative deviations as limits of good and bad harvest respectively whose appearing probability in 1/5. Figure 32 shows the relation between good and bad harvest of paddy rice and corresponding mean temperature of July to August. It is seen at once from the figure that poor yield occurs without exception when the mean temperature of July to August decreases beyond 21.3℃. On the contrary, when the mean temperature exceeds 23.8℃ poor yield does not occur and good or usual yield comes out. Next, a regression curve of yield on mean temperature in July to August is obtained and shown in Figure 34. It is seen that the damage of low temperature is very remarkable, and the regression is curvilinear whose tendency has been described as follows: (i) Usual yield will be given in mean temperature about 22.0℃ in July to August. (ii) Yield will be less than average bellow 21.0℃ and more than above 24.0℃. (iii) Yield will be poor bellow 20.0℃ and not always good above 24.0℃. /2. The climatic influence upon paddy rice productivity in the warm district (Kyushu)/ /(1) Choice of climatic factors influencing upon rice yield/ The method of multiple correlation is applied to estimate the effect of climatic elements represented in monthly units of time on rice yield for the years from 1919 to 1938 in Fukuoka, and it is shown that the sunshine-duration in September has the prime influence and the mean temperature in August has the next, and that these two factors only give a sufficient regression equation for yield. /(2) Seasonal variation of the influence of sunshine-duration/ To make clear the single effect of the sunshine-duration in detail, the 120 days from July to October, i.e. growing season, are divided in 10 days' intervals, and the effect of the sunshine-duration in each interval on the yield are estimated by Fisher's method) of orthogonal polynomials. It is shown that the sunshine-duration the second 10 days of July, and the first and the second 10 days of September have larger positive effects than that in other periods, but in October it has rather a negative effect (Figure 42). /(3) Joint effect of sunshine-duration and mean temperature on rice yield/ The joint effects of sunshine-duration and mean temperature in August and September on the rice yield in Fukuoka are analyzed, assuming that the effect of sunshine-duration in one of the months depends upon the degree of the mean temperature in the same month, and vice versa. It is found that the effect of the sunshine-duration increases as the mean temperature rises, being largest in September (Figure 45), and that the effect of high temperature is beneficial only when the sunshine-duration is large (Figure 46). /(4) Influence of high temperature on rice yield/ In warm district, too high temperature in summer often causes the decrease in rice yield. Figure 47 shows regression curves of the rice yield on the mean temperature for three prefectures in Kyushu, and it is seen at once that the mean temperature in August is most suitable at about 27℃ and the range of positive effect lies in 26.5℃ to 27.5℃. Figure 48 shows accordingly the frequency (%) of occurrence of cases of the mean August temperature in Kyushu exceeding 27.5℃ and being short of 26.5℃ respectively. /3. The relation between the early and late cultivation of paddy rice and climatic conditions/ The early or late cultivation of paddy rice for the purpose of its stability and land utilization will possibly develop more in the future. Generally speaking, accumulated temperature (above 10℃) is about 5000℃ and the period of possibility of paddy rice cultivation is very long in Kyushu. So the author investigates about temperatures in warm season when the paddy rice cultivation is possible. Assuming for convenience that the mean temperature necessary for the cultivation of paddy rice is above 10℃, isopleths of occurrence of 10℃ in Spring and Autumn are shown in Figures 50 and 51, and its relation against topography is shown in Figure 52 being classified by azimuth. Next, regarding the number of days when the mean temperature is higher than 10℃ as the period of possibility of cultivation, the distribution of the period is given Figure 53. It is seen that the area of minimum of about 200 lies inland and that of maximum of about 280 lies in the southern part of Kagoshima Prefecture. Also some considerations about the change of air temperature during the course of the period of possibility of paddy rice cultivation are made, and an example showing the difference between the coastal and inland areas is given in Figure 57.