The coefficients of correlation between the yield of hemp and the weather factors (i. e. monthly mean temperature, monthly amount of rainfall, monthly total number of rainy days and monthly total hours of sunsh ne) have been calculated by the method reported in a previous number of this magazine, for each prefecture in Japan, in each month during the cultivating time. The correlation coefficients calculated are shown in tables 1, 2, 3 and 4. The correlation coefficients are generally small; only, in the north-eastern provinces, air temperature is more closely correlated with the yield than any other weather factor, and that positively, while in the southern part the closest correlation is found between the yield and rainfall. The results of this investigation show that the warm, rainy weather in the cultivation time of hemp is favourable for it in Japan proper.
In his famous paper on the energy of the storm, M. Margules proved that a system of warmer and colder air masses lying side by side liberates sufficient energy to produce a very considerable velocity in readjusting itself so that the warmer mass ascends and lies above the colder, assuming the stratification of dry adiabatic lapse-rate. Such a horizontally unstable situation as Margules postulated will be called a state of horizontal lability in S. Fujiwhara's terminology. S. Fujiwhara(1) has further pointed out various types of horizontal lability and discussed its importance in the energetics of cyclone, thunderstorm and other related phenomena. In Margules'theory the process of readjustment is dry adiabatic in a layer of dry adiabatic lapserate, therefore no energy is lost owing to vertical motion. The present paper extends the above theory so as to include a lapse-rate smaller than the dry adiabatic. Thus the stratification is stable for any dry adiabatic vertical motion. Consider the horizontal distribution of warmer and colder masses as in Fig. 1. As the lapserate is assumed to be smaller than the dry adiabatic, the entropy increases with height, therefore it can happen that the entropy at a height of π1 in the colder mass 1 is equal to that at the ground in the warmer, mass 2, and the entropy at the top of the warmer is the same as that at a height of π2 in the colder. In the final state, the lower part of the colder mass (lower shaded area in Fig. 1) spreads laterally on the ground, on which comes the mixed layer of 1 and 2 masses (unshaded area) and further upwards the upper part of the warmer mass. Thus in the final state the entropy increases continuously with height. Naturally the mass-integral (Massenintegral) becomes much complicated in comparison with that of Margules. The total potential energy E∫ in the initial state is simply given by where a is the lapse-rate and the other symbols denote the same as those in Margules' paper(1). The derivation of potential energy Ef in the final state requires laborious evaluation and the result is as follows: Here _??_1 and _??_2 denote the temperatures at π1, and π2 respectively and (from the condition of equal entropy at p1 and p2), The difference of the above two energies expresses the maximum energy releasable. Margules' computations were carried through with great accuracy, since the final net gain of kinetic energy is the relatively small difference of two large quantities. The same is the case with the present computation. The result of numerical evaluation is as follows: 1. horizontal difference of temperature: 10°C 2. horizontal difference of temperature: 7.5°C 3. horizontal difference of temperature: 5°C It is seen here that the small horizontal lability is readily supressed by a stable stratification of the atmosphere, but when it becomes larger and exceeds a certain critical value (about 8°C), the energy of horizontal lability is able to produce sufficient energy to maintain a cyclonic storm even in a very stable atmosphere. The author has added the proof that the above conclusion holds good also when the temperature varies continuously in horizontal direction.
For the purpose of taking photographs of lightning flashes, a photographic instrument with a wide visual angle was constructed by the present writer. Ten Leica cameras were so attached along a vertical steel rod that a lightning occurring in any direction might be caught by some one of the cameras. The visual angle is 360° in this case. The steel rod was rotated with a motor. In this way eighteen lightnings appearing at distances smaller than 3km, as well as many others more distant, were photographed. Contrary to the observations of Schonland, the leaders of the first strokes showed no stepped structure. Some of them were continuous like dart leaders and others were composed of several fore-discharges which resembled those occurring in the long sparks produced in the labolatory. One of the leaders showed a band-like structure. It started from the middle point of the lightning path and propagated in both the upward and downward directions. Two examples of upward branching were obtained, but they were found to be different in nature from the ordinary downward branching. Various interesting phenomena caused by space charge below the cloud were observed. Short lightnings streching out into the clear sky from the rim of the cloud were photographed, and it is supposed that they ended in the space charge outside the cloud.