VARIATION OF THE KUROSHIO NEAR THE JAPAN ISLANDS 621 



oi very low temperature (and very low salinity) of 3-4° C at 200-meter 

 layer near the coast, (This temperature minimum is a very rare phe- 

 nomenon as this appeared so far south.) These results indicate that the 

 Oyashio came down farther south than usual. On 22 April, 1 1 days aft- 

 er the former observation, the temperature minimum disappeared com- 

 pletely and the isotherms inclined so sharp that this indicated that there 

 was a strong northeasterly current, that is the Kuroshio was flowing in 

 this area. From the observation of 9 May, the temperature rose gen- 

 erally and warm Avater occupied this area. 



The variation of the sea water has close connection with the me- 

 teorological conditions. So we examined the wind observed at the Chosi 

 Weather Station. From 13 March to 11 April northern winds prevailed. 

 On 18 and 28 March and 8 April the wind velocity exceeded 20 m/s 

 (wind direction N or NNE) and on 18 March and 8 April winds of 

 more than 10 m/s continued more than 24 hours. After 11 April it 

 was generally calm and the wind direction was variable. From these 

 data it may be concluded that the southward flow of cold water from 

 March to April was caused by the strong north wind of the same period, 



3) The Translation of the Kuroshio and its Periodic Character 

 AS Observed by the Mean Sea Level 



Due to La Fond the variation of water level observed at tidal 

 stations (the effect of tide eliminated by taking suitable mean values) 

 is almost the same as the variation of dynamic height of the neigh- 

 bouring sea surface. 



The difference in the dynamic height on both sidea of an ocean 

 cunent like the Kuroshio reaches as much as about 100 cm. There- 

 fore if water level is observed at an island situated in the stream of 

 the Kuroshio, we can know the translation of the Kuroshio continuously. 

 For this purpose we put a tidal station at Hachijo-sima, and observed 

 the variation of water level, and at the same time the record of three 

 tidal stations on the coast of Honshu were also examined. As the va- 

 riation of the current was very sharp in some cases, we used daily mean 

 sea level (25-hour mean); and to eliminate the barometric effect, cor- 

 rection was performed assuming 1 mmb. of air pressure corresponds to 

 1 cm. of water level. Air pressiue was taken from the record of the 

 nearest weather stations. 



At first it is necessary to examine whether the mean sea level varies 

 parallel with the dynamic depth or not. Figure 8 shows the correlation 

 between the dynamic depth obtained from oceanographic observations 

 and the sea level. As both values have error of a few centimeters, we 

 can conclude that proportionality exists fairly well between these values. 



