simple examples, 



0-)N'zf{0^d^y9) . As a practical problem there is the forecasting of the 

 locations of the skipjack fishing grounds, which every year in May, June, 

 and July shift rapidly from the waters off Nojimazaki in Boshu through the 

 seas off Choshi to the area off Kinkazan„ Hitherto only superficial 

 investigations depending on charts of surface water tempera tiare s have 

 been made, however, if we try examining the correlation between the posi- 

 tions of the fishing grounds and the distribution of the 19° -> 22°C warm 

 water zones (represented on the chart by the 20° isotherm) in each ten-day 

 period of the four years from 1935 to 1939, we can see (Figure 1) thnt 

 there is a rather close relationship between the movements of the fishing 

 grounds and the form of the isotherms„ The grounds which in May were 

 stationary in the warm current system surrounding the edges of the cold 

 water mass southeast of Nojimazaki, have shifted in June to the northward 

 projecting peaks of the undiilating line of the isotherms, which have moved 

 to the north and northeast. In July there is further sharp* northward 

 shift and two peaks in the isotherms are clearly apparent where there are 

 two projecting branches of the warm currents As a consequence the fishing 

 grounds are also split into twoo At the same time the center of the 

 fishing grounds shifts gradually from the coast to the offshore waters, 

 where the main fishing grounds are located from the latter part of June 

 to the early part of July. After the middle of July a conspicuous fishing 

 ground appears off Kinkazan comparatively close to the coast. This means 

 that the general outlines of the fishing grounds can be deduced from the 

 water temperature^, the distribution of the isotherms (7(95 and the rate of 

 change and consequent shifting of the isotherms 0^ however, since there 

 is a need for even more accurate forecasting of the "fishway", other 

 factors must be taken into consideration. Now if we show (Figure 2) the 

 locations of the appearance of feeding skipjack schools as studied by Mr. 

 Gorozo Okamoto from the data of 1936, being able to assume a route of 

 movement generally corresponding to the shifts in the main fishing grounds 

 and to the deduced migration routes of marked skipjack, we car believe 

 that the skipjack schools which move northward in the Northeastvsrn area 

 are on a feeding migration, following a path between places where natural 

 foods (sardines, mysids) are abundant. Consequently it can be deduced 

 that the movements of these schools of natural food animals also follow 

 the branches of the warm currents. 



{.2)7^ ^ f{0,^j,6) ^ As is shown in Figure 3, the results of statistical 

 studies indicate that in the Northeastern area in 1936 and 1937 both the 

 number of fish taken and the frequency of catches reached their maximum 

 value when the difference between the surface temoerature {&<>) and the 

 temperature at the 100-meter level {e,<,o^ was 10° = 16°C (^ijof 20° - 2A°C)o 

 Furthermore, in late June and early July in the southern part of the 

 Northeastern area another mode appeared where ^^ 22° - 23°C and (<S-,-(^..J 

 r 2 - 4.°C. Figiare 4 \ras m^de from charts of the distribution of water 

 temperatures at the surface and the 100-meter level obtained through the 



* With regard to this rapid shift in the isotherms and the positions of 

 branches of the warm current, a close correlation can be seen nlth 

 meteorological conditions preceding and following the start of the 

 spring rainy season^ 



