Abstract
Frequency curve as shown in fig. 1 p. 3. is constracted using the data from 444 ascents of pilot balloon. The authors conclude that (1) prevailing directions of upper winds differs from a stratum to another and this seems to be caused by some disturbing action coherent to each statum; (2) the direction of maximum frequency in the layer of lower handreds maters is determined mainly by the orographic conditions. At Kobe the prevailing direction of wind at that layer is from the east. (3) The direction of gradient wind becomes prevailing at and above the height of about 2000m.
