838 EIGHTH PACIFIC SCIENCE CONGRESS 



September and in October. Perhaps the ratio in August can be con- 

 jectured to be large from the tendency of the figure, but observations 

 were not done before September. 



In 1949 the Kurosiwo had average or somewhat stronger intensity 

 in summer, and the warmer anomaly of surface water temperature was 

 largest in October. The figure shows that the ratio of the Kurosiwo 

 had maxima in June and July, and that their value was usual 

 lor summer maxima. Another maximal value appeared in Septem- 

 ber, after the rainy season. This was smaller than that of the preced- 

 ing year. This fact indicates that the Kurosiwo was weaker than in 

 the preceding year. In these years, the available data were scanty due 

 to the imperfection of our equipment. 



In 1950 the Kurosi'wo became stronger rapidly from April, with an 

 extremely ^\arm surface temperature in summer. The warmer anomaly 

 of surface temperature, ^vhich was largest in May and August, was 

 limited to the somewhat upper layers in summer. The Oyasiwo be- 

 came stronger rapidly in April. Aveakening from middle August and 

 growing stronger again in September. But it w^as still warmer in the 

 warm current region and colder in the cold current region than in the 

 average year.^ The ratio of the Kurosiwo grew larger from April and 

 May, and it was very lage in August and September, growing smaller 

 rapidly in October and tending to zero in December. The ratio of the 

 Oyasiwo Avas large in spring and in early summer also, but it was small 

 in summer, growing large again from September, and reaching a maxi- 

 mum in December. 



In 1951 the Kurosiwo became stronger rapidly from April, but 

 weakened from middle June to July; the temperature anomaly Avas low- 

 er than the average. But its temperature became higher again from 

 late July to August. It stopped to rise, maintaining the average or 

 somewhat lower A'alue in September and October. But it became strong- 

 er from November especially and kept the intensity till February 1952; 

 so the Avarmer anomaly of temperature, Avarmer than those of 1950 and 

 1951, Avas observed in February, and this vigorous KurosiAvo weakened 

 in March and April. The OyasiAvo had less intensity than that in the 

 preceding year in early Spring, but it became stronger in early July, 

 Aveakening in August and again becoming stronger in September, then 

 Aveakening from October till February.""' The ratio of the Kurosiwo 

 grew large rapidly from April and was largest in Jime, growing smaller 

 from July but larger again from NoA'ember, becoming largest in Feb- 

 ruary, 1952. The ratio of the OyasiAvo had a maximum iiij December, 

 1950, and alter that it grcAV small and A\as smaller than in the preced- 



