sal fin starting above the posterior margin of preopercle, slightly 

 higher than body height at its anterior end, down at the vicinity 

 of 10th spine, again up to nearly the same height posteriorly and 

 terminating just in front of insertion of second dorsal fin. Second 

 dorsal fin and second anal fin small, nearly the same in size, and 

 the latter situated anterior to the former. First anal fin 

 somewhat small and crescent-shaped. Ventral fin longer than 

 pectoral fin. 



Fin membrane of first dorsal fin deep blue in color without 

 spots. Other fins brown or dark brown. Basal part of first dorsal 

 fin and second dorsal fin silvery white. Dorsal part of body deep 

 blue, side somewhat brownish blue and belly silvery white. 



Olfactory rosette radial in form, comprising 45-47 olfactory 

 laminas. Gonad asymmetrical, Y-shaped; left lobe remarkably 

 developed. Anus situated far anterior to insertion of first anal 

 fin. 



Variation in counts of spines and fin rays as provi- 

 ded in the literature are shown in Table 1. 



Nakamura et al. (1968), who examined nine 

 specimens from the northwestern Pacific, 

 southeastern Pacific, and Indian Ocean, give the 

 following ranges of fin ray counts: first dorsal spines, 

 47-50, second dorsal fin ray, 6-7; first anal spines, 12- 

 15; second anal fin rays, 6-7; and pectoral fin rays, 18- 

 19. These all fall within the ranges given in Table 1. 



Merrett (1971) provides some morphometric 

 relations for T. angustirostris (Table 2). 



Geographic variation: Penrith (1964) suggests that 

 there are geographic variations in pectoral fin length; 

 the length of the pectoral fin forms a cline of in- 

 creasing length eastward through the Indo-Pacific 

 region. Penrith compared body proportions of a 

 specimen taken in the Indian Ocean off South Africa 

 with those of a Chilean specimen measured by Robins 

 and de Sylva (1960) and noted that the values of these 

 two specimens of T. angustirostris, taken at the ex- 

 tremes of the known range, agree very well except for 

 the length of the pectoral fin. The ratio of the length of 



Table 1. — Individual variation in counts of spines and fin rays 

 of Tetrapturus angustirostris. 



Counts (upper figure) 

 Spines and fin rays Frequency (lower figure) 



Dorsal spines _J1 JL J^ Jl_ J§_ _i?_ J9_ _51_ 52 



3 16 5 3 2 



Second dorsal fin ravs 2 _ Z ° r_ 



7 2 



Anal spines _AL ^?_ _^3_ J4_ ^5_ _16_ _17_ 



2 13 11 



Second dorsal fin rays ?_ § 1 § _ 



3 6 



Pectoral fin ravs J6_ ^J8^ J9^ J0_ _21_ 



2 5 1 



Source: Tanaka, 1914; Robins and de Sylva, 1960; Ueyanagi, 

 1962; Watanabe and Ueyanagi, 1963; Merrett, 1971. 



the pectoral fin to the body length is 9.8% for the 

 South African fish of 1,456 mm and 15.0% for the 

 Chilean fish of 1,822 mm, respectively. He adds to 

 these the value of about 11% for Japanese fish, judged 

 from the plates in Nakamura (1951) and Tanaka 

 (1914). Penrith is of the opinion that although there is 

 marked allometric growth of the pectoral fins in T. 

 pfluegeri (Robins and de Sylva, 1963), it seems doubt- 

 ful that there is any comparable allometric growth in 

 T. angustirostris, because he sees no sign of allometric 

 growth over a range of 1,470-1,857 mm (measured 

 from tip of snout to fork) in the tables given by Royce 

 (1957) for Hawaiian fish. The difference in size 

 between the South African fish and Chilean fish is 

 nearly comparable to the above range. This led 

 Penrith to believe that the large difference in the 

 ratios of the pectoral fin length to the body length 

 between these two specimens may be due to 

 geographic variation and not to the allometric growth 

 of the pectoral fins. He also shows the following values 

 for the pectoral fin length as expressed as a percentage 

 of the first dorsal height: West Indian Ocean 76.6%, 

 Hawaii (Royce, 1957) 87.4-100.4% with a mean of 94% 

 and in the extreme east, off California 100.6% (Craig, 

 1958), and Chile 100.8% (Robins and de Sylva, 1960). 

 Merrett (1971) analyzed the morphometric data on 

 T. angustirostris from the equatorial western Indian 

 Ocean from the viewpoint of specific identity and 

 compared the results with those from the Pacific 

 Ocean given by Royce (1957). His comparison of the 

 mean values of four characters, i.e., the length of the 

 pelvic fin, the anterior height of first dorsal fin to the 

 fork length, the anterior height of first anal fin to the 

 anterior height of the first dorsal fin and the length of 

 the 20th dorsal spine, shows that the values are in all 

 cases smaller for the Indian Ocean fish. Merrett (1971) 

 suggests that these probably reflect population 

 differences in the species from the Pacific and Indian 

 Oceans, since the same trend is apparent throughout. 

 He gives the mean value of the pectoral fin length to 

 body length of 11.6% for the six specimens he exam- 

 ined. According to Merrett, this tends to confirm the 



Table 2. — Various morphometric relations of Tetrapturus 

 angustirostris. 



42 



