stock of large tuna, and in addition to this trend ^° ^ on© should probably 

 discuss the elimination of this long- term change by the year-to-year 

 variations duo to oceanographic conditions and so fortho 



This tendency can be shown approximately by the linear correlation 

 y = ax / b between the number of fish taken per boat per cruise y and the 

 year x » By the method of least squares we get a s 0o313 and b = 6 ,365 

 and can draw the solid line shown in Figure 2^ 



The coefficient of correlation between the vertical difference in 

 water temperatures and the deviation from this regression line is r s 

 — 0„738;t0o085p and fairly reliable values can bo obtamedc 



Furthermore, the catch of small fish per cruise_is correlated with 

 v^^w L^urface water temperature in Augusts IJ 

 the favorable temperatures (Figure 3)o ^'^' 



the surface water temperature in Augusts 15 oO - 16 -,5 C appearing to be 



In preparing this paper I received much kind advice from Professor 

 Okada of the Fisheries Experiment Stationj and Assistant Technician 

 Nakajima of the Hokkaido Fisheries Experiment Station and ^r .> Hisashi 

 Saga of the Kushiro Fish Market supplied me with materials.. I wish to 

 express here my deep thanks to these persons o 



(6) 



(7) 



This trend also appears to be related to the broadened scope of the 

 fishing grounds through the increase in vessels fishings but there are 

 no data on which to determine this point o 



According to the study by Mr,. Takeo Kida of th? total number of fish 

 (large, medium^, and small) taken per cruise on the grounds off Kushiro 

 in August 1935, there is a favorable surface temperature at 17°C„ 

 Kida, Takeo s On the Surface Jtfater Temperattires and the Summer Season 

 Tuna Grounds Off A-ushiro and Urakawa^. Bullo Jap „ Sqco Scio Fisho 

 5 (2), 1936, 85-90o 



13 



