are sufficient reasons to assume that the 

 initial factor would come in effect at an earlier 

 stage of the fish under development. This as- 

 sumption appears in an agreement with hypoth- 

 esis (i). Even if the frequency would start 

 once at an early developmental stage, however, 

 the ecological mechanism accountable for the 

 phenomenon must have been of connplex rather 

 than simple one. For this reason, when prop- 

 agation of the fish is to be discussedas acom- 

 ponent of the mechanism influencing the fre- 

 quency, it is necessary to take into consider- 

 ation such various environmental factors as 

 to be supplied into different parts of the fishing 

 regions, i.e., the northern hemisphere of the 

 Pacific, its counterpart in the south, and the 

 Indian Ocean. 



5. So far as the present study goes, it is 

 prennature to decide whether or not the biennial 

 frequency will continue to make an appearance 

 in the future. But if it does, knowledge and 

 data we obtained so far will certainly serve 

 as a n important suggestion from which ad- 

 vancement of study on the big-eyed tuna is 

 expected. 



/end of English summary/ 



Nakamura, Kamimura, and Yabuta (1953) 

 have already made a connparison of the length 

 frequency composition of the catch of bigeye 

 tuna from the North Pacific fishing grounds in 

 the two periods of December 1948-March 1949 

 and October 1949-March 1950, and they have 

 pointed out clearly discrepancies in the posi- 

 tions at which the modes appeared in these two 

 seasons and in the manner of their appearance, 

 the former having modes centered at 85-88 cm. , 

 121-124 cm. , and 149-152 cm. , while the lat- 

 ter had conspicuous modes centered at 93-96 

 cm. and 137-140 cm., with another somewhat 

 obscure mode centered at 105-108 cm. They 

 have considered the following three possible 

 causes of these discrepancies: 



1) Differences from year to year in 

 growth rates 



2) Differences from year to year in 

 recruitnnent 



3) Differences fronn year to year in 

 migrational phenomena 



The present authors have made a compar- 

 ison of the length measurement data for the 

 three periods of October 1950-March 1951, 

 October 1951-March 1952, and October 1952- 

 March 1953, together with the older data 



mentioned above, and have found some 

 extremely interesting phenomena, which they 

 report here. 



Collection of the data used in this study 

 was all planned and executed by the High Seas 

 Resources Section of the Nankai Regional 

 Fisheries Research Laboratory. The term 

 "North Pacific fishing ground" used in this 

 paper means the area from 130 E. longitude to 

 165 W. longitude and north of 26 N. latitude. 

 The authors herewith express their thanks to 

 Dr. Nakamura, director of this laboratory, and 

 to Dr. Yabe, chief of the High Seas Resources 

 Section, for their guidance in this study. 



Data 



The measurements, taken with wooden 

 calipers, are the shortest distance from the tip 

 of the snout to the bottom of the fork of the tail. 

 Measurements were read to the nearest centi- 

 meter, units below 1 cm. being dropped. In 

 this study the data were arranged in 4 cm. 

 groups. As far as sampling is concerned, the 

 first unit of selection was the vessel and the 

 second unit of selection was the fish. The per- 

 centage of selection at the first level did not 

 follow any particular standard, the greatest 

 number of boats possible under the circum- 

 stances being sampled. At the second unit of 

 selection the original objective was to measure 

 all fish, but where this was not possible the 

 percentage of selection was nnodified to suit the 

 circumstances. 



Table 1 shows the numbers of fish measured 

 by month and area: 



I. Length Composition 



Figure 1 shows the length composition by 

 years* and by sea area. The most striking 

 features of this figure are the appearance in 

 each year of what clearly can be considered 

 nnodes, and the fact that a tendency can be de- 

 tected for the positions at which the modes ap- 

 pear to coincide generally in the different sea 

 areas. The positions of the modes were: in 

 1948, 69-80 cm., 81-96 cm. , 113-132 cm. , and 

 145-164 cm. ; in 1949, 89-100 cm. , 101-116 cm., 

 and 132-148 cm.; in 1950, 85-100 cm., 109-128 

 cm., and 141-164 cm.; in 1951, 89-104 cm., 

 105- 124 cm., and 129-152 cm. ; and in 1952, 89- 

 104 cm., 113-136 cm., and 141-160 cm. As for 



♦Hereafter the periods of measurement will be 

 referred to as "years". 



22 



