to each representative ring^ that iSj the radius (r) of the 

 annulusj and of the distance from the center to the outer edge 

 of the centrum^ that is, the total length (T) of the centrum 

 (table 2)„ There is room for individual error in the selection 

 of the representative annuli., but the fact that this error is 

 very small is clear fr-om a consideration of the standard devia- 

 tion of the radius of each ring« as shown in table 2o There iSj 

 as shown in figure 1^ a definite correlation between the body- 

 length 0) of the skipjack and the total length of the centrum 

 (T) at the time,, Consequently every ring can be tentatively 

 taken as an annuluso Each ring is completed when the total 

 length T is equal to ro In other wordSj it can be thought that 

 when T^ - ify-^y the body length is l^^s, and at this time the 

 skipjack has completed n years since hatching » Skipjack complete 

 the first ring (r^) at a body length of 26 mos, the second (rg) 

 at 3U ciUos the third (r-^) at k3 cmo, and the fourth (r|^) at Sk cmo 

 Skipjack with a body length of less than 26 cmo are first year 

 fish (0-year groiip)^ those from 26=3ii cmo are second=year fish 

 (I-year group), :^-U3 cmo are third-year fish (Il-year group), 

 k3~^k cmc are fourth=-year fish (Ill-year group), and those over 

 5ii cmo arai/ fifth=-year (IV-year group )» In the catch from the 

 pole and line fishery there are no'fish under 30 ctoo nor over 

 80 cmo J and the m^ n part of the catch is from slightly over 

 li3'cmo to slightly over 6? cmo The group of small fish with its 

 mode at i|6 cmo is between UO-SO cmo and clearly consists of fourth- 

 year fish, and the large-sized group between 55-70 cmo are fifth- 

 year flsho On the average in the three years 19314.-1936 fourth- 

 year fish and fifth-year fish were taken in the proportion of 

 l|8o9^ and 6lol^ respectively (table 3).v which is in approximate 

 agreement with the ratios of large and small skipjack in the 

 catch as given in table lo The medium-sized skipjack include fish 

 which must belong to either the fourth-year or the fifth^year 

 classo From year to year there is more or less of a change in the 

 average body length and the mode of the fotsrth-year and fifth-year 

 fish (table k) o Furthermore;, there are differences in the age 

 composition in different years (table 3)o Since this is considered 

 to be one of the iirportant causes of fluctuations in the amount of 

 the catchy hereafter accurate observations must be made of the 

 course of changes in the age composition in addij-tion to accurate 

 investigations of the amount of the catch,, 



=/ Large skipjack are occasionally taken on tuna longlines in the 

 wintero A skipjack taken in the middle of December 1936^ h3 

 miles SE/n|E of Nojima Saki (water temperature 19ol°C) at a 

 depth of 30 fathoms had a body length of 8lo3 cmc (total length 

 88o5 cmo) and a body weight of liio5 kgo Since it was possible 

 to measure up to the 6th ringj it is presumed io have been a 

 seventh-year fisho 



36 



