2 T. Terada: 



of the patli tlian on tlie right, and the heaviest falls occur in ad- 

 vance of the centre" regardless of the direction of progression of 

 the depression. He also remarks that the distribution of cyclonic 

 rains seems to have no apparent relation to the physical feature of 

 the country, though data were wanting to decide the point. J. A. 

 Udden^"^ who attacked the proldem for Davonport, Rock Island, 

 111., U.S.A. and also for a number of other districts, found that 

 the position of tlie area of the greatest rainfall-frequency relative to 

 the depression, varies largely for difïerent localities ; he suggested 

 that the relation may vary w^ith the hours, the seasons and the 

 -courses taken ]\y the depression. W. G. Reed^-\ following the 

 method adopted by Mill, made an extensive investigation of the 

 " smear " for a number of cyclones passing over the U.S.A. His 

 results are not so simple as that of Mill. He notes rightly the con- 

 siderable influence of large water bodies, such as the Atlantic Ocean 

 .and the Great I^akes and also the fact that the rain area forms a 

 series of patches, the heavier ones being connected by lighter ones 

 ■or rainless areas. The latter fact is also attril)uted to the influence 

 •of important water bodies, though he concludes : " every relation 

 ■expected occurs, l)ut there seems to be as yet no classification 

 which will reduce the relation to a system." Hann^'^ discusses 

 these results in his Lehrbuch and emphasizes the influence of the 

 •districts over which the wind comes. 



One of the authors^'^ has discussed the influence of the distri- 

 bution of land and water in modifying the meteorological feature of 

 a region and proposed a simple theory which may explain many 

 regional irregularities of barometric distribution. As an illustration 

 of this theory, he drew attention to the difference of the Pacific and 

 Japan Sea Coasts, ^vith regard to the relative position of the rain area 

 and the centre of Ijarometric depression. The present paper may 

 be regarded on one hand as the continuation of the above cited one, 

 in so far as the results of the statistical investigation are reviewed 

 under the light of the theory, but on the other hand as a contribu- 

 tion to the literature on the general precipitation problem. 



1.) J. A. Udden, Symon's Met. Mag. 41, 1906. 



2.) W. G, Reed, Monthly W^eather Review, Oct. 1911. 



3.) Hann, Lehrbuch d. Met. 3te Auf. S. 524., 552. 



4.) Terada, Proc. Tùkyù Math.-Phys. Soc, 7, 1914.. 



