On passing Korean peninsula the front presented itself as two, and on getting over the Japanese Islands as three lines of discontinuity as shown in fig. 1. in p. 235. Such abehaviour of the front is already pointed out by J. Bjerkness and H. Solberg. The structure of the front was nearly that shown by the model of squal line of W. Schmidt, which fact is clearly understood from the way of variation of weather elements as shown in fig. 2. in p. 236. It is noteworthy that the direction (NW) of rushing air was normal to the front (SW-NE) and after its passage the wind shifted to from the west and the air temperature increased. This fact is not quite in accordance with the model of Schmidt. It seems probable that the head of cold front is the seat of a linear vortex with horizontal axis, which proceeds by rolling as if a solitary bore. The velocity of the proceeding front was well expressed in average by the formula given by Schmidt where b is the bearing of the pressure nose in millimetre of mercury and v is the velocity in km per hour.
The magnitude of cloud particles is measured by many authorities in Europe but not yet done in Japan. Probably it will be the same in Japan as in Europe, and to make sure the author calculated it from the results of observations of the radii of coronae. The result is that the radii of cloud elements which is measured in October, 1924 in Tokyo range between 0.004mm and 0.034mm, and values near 0.01mm occurred frequently. These values are of nearly the same order of magnitude observed by Assman and Kämtz. Pernter showed that the radii of cloud elements observed by means of corona radius range between 0.01mm and 0.05mm. Dines also gives rather great value. Comparing with these results the figures obtained by the present author is rather too small. Whether this due to the effect of. locality or to the insufficiency of the number of observations must be determined afterward.