that weighed 27.7 kg and was 226 cm TL had 7 

 ridges (Figure 2), while the sagitta of a sailfish 

 that weighed 12.7 kg and was 183 cm TL had 3 

 ridges (Figure 4). These ridges are distinctive and 

 easily counted, and of a sample of 65 sailfish only 1 

 specimen had uncountable ridges on both sagittae. 



The convex or lateral surface of the sagitta (Fig- 

 ure 5) was moderately smooth and the two lobes 

 were joined in the core region. The convex surface 

 did not have the distinctive configuration seen on 

 the side of the rostral lobe. A medial section of the 

 sagitta revealed internal structures which indi- 

 cated incremental otolith growth and progressive 

 deposition (Figure 6). 



The asteriscus and lapillus (Figure 7) are much 

 smaller than the sagitta and also show internal 

 features that appear to have progressive deposi- 

 tion. The asteriscus is very fragile and was often 

 broken during dissection, but its growth features 

 support the sagittal ridge counts. The lapillus was 

 much heavier in structure than the asteriscus and 

 also showed internal growth features that appear 

 to be age related (Figure 8). The lapillus showed 

 layers that correspond to the number of ridges 

 found on the sagitta when viewed under transmit- 

 ted light (Figure 9). The lapilli age counts were 

 identical to sagittal ridge counts 85% of the time 

 and ±1 yr in the other counts. The possibility 

 exists that age estimates could be obtained by 

 light microscope techniques, although the sagittal 



Figure 4.— The rostral lobe of the left sagitta from a 12.7 kg 

 sailfish with three ridges (85 x 1. Numbers indicate ridges. Bar = 

 0.1 mm. 



ridges were very distinct in SEM preparations 

 which we feel would be more accurate. 



The ridges on the sagitta of 64 sailfish were 

 counted and compared with body weight and a 

 strong correlation was found (Figure 3). The 

 ridges appear to be annual events and fish with 

 four ridges were the most numerous in our sample. 

 We did a log transformation of the allometric 

 growth curve equation or power curve fit of the 

 data (Table 1) so we could compare our data di- 

 rectly with that of Jolley (1977) for which sum- 

 mary statistics only are available. The data used 

 for statistical analysis were selected from the Jol- 

 ley (1977) regressions. An F-test for difference 

 between the regression coefficients gave P>0.05, 

 so there is no apparent difference between the 

 slopes of Jolley 's data and ours. The data indicate 

 that the sailfish develop a large muscle mass in a 

 short period of time (<4 yr), and our results are 

 similar to Jolley's over the same time span. 



Table l.— Age-weight relationships ( Y = a.v'') for the Atlantic 

 sailfish as expressed by predictive power equations, logio trans- 

 formed for purposes of testing difference in regression coeffi- 

 cients between the data presented in this paper and those of 

 Jolley (1977). 



3 4 5 



AGE (TEARS) 



Figure 3. — Relationship between the weight of the sailfish and 

 the number of ridges (years) on the sagitta. 



363 



