Authors investigated barometric gradient for 60 cases when north wind was blowing at 14 o'clock. at Kumagaya. They calculated the gradients from the barometer readings at Kumagaya and four surrounding stations. They arrived at a very queenr result shown in Fig. A and B (page30) in which the angle of deviation is reckonned possitively from North to East. This perhaps shows that the word “barometric gradient” in such mountainous country as Japan is not self evident. It depends on the magnitude of area concerned just as the case with motion.
It is explained that along the steering line, there exists a roll of horizontal vortex, which is maintained by the difference of temperature and velocity of air currents at the sliding surface. Near a cyclone there is gradual increase of ascending current as approaching the centre and the end of the said roll is lifted up and winds in the centre, and feeds the vorticity of the centre. The vorticity and cosequantly the kinetic energy of a storm is mainly maintained by this process. This view stands quite in harmony with the theory of Margules, the fact of W. H. Dines and the discovery of J. Bjerknes and H. Solberg.