FISHERY BULLETIN: VOL. 70. NO. 3 



with scales collected from wild fish and three ex- 

 perienced scale readers read the scales (Table 1) . 

 The percent error was as large as 100%, which 

 indicates that accessory marks can indeed be 

 easily misidentified as annuli. 



TIME OF FORMATION 



As many as f oiir marks were observed on some 

 scales. The percentage of fish with various num- 

 bers of marks was tabulated for each sample 

 (Table 2) . The results show that the first mark 

 was formed from August to October (4-5 months 

 old), the secbnd from November to March (6-10 



ANNULUS 



|< 1 



months old) , the third from May to August (IS- 

 IS months old), and the fourth sometime after 

 November (18 months old) but probably before 

 January (20 months old). This January date 

 for the fourth mark was deduced from two scales 

 collected in late December from the bottom of the 

 pool, but the data are not given in Table 2, The 

 two scales had four distinct marks, but it is not 

 known whether the scales were from one or two 

 fish. The bottom of the pool was cleaned daily. 

 Based on the above criteria of an annulus and 

 accessory mark, the second and fourth marks 

 are annuli, and the first and third marks are ac- 

 cessory marks. The interval between annuli 

 was about 12 months, and that between acces- 

 sory marks was only 9 months. 



ACCESSORY 



< n 



C0NDITI0r4 FACTOR 



RELATIVE GROWTH RATE 



GONAD INDEX 



WATER SALINITY 



WATER TEMPERATURE 



Figure 1. — Four possible factors that may be associated with mark formation on scales 

 of Pacific sardines. Periods of mark formation are delineated by vertical lines. 



1046 



