FISHERY BULLETIN: VOL. 72. NO. 2 



Table 1. — Capture areas, species, number, size range, and mean standard 

 length for specimens studied. 



various reports, probably because of insufficient 

 sample size and population differences. I have 

 found variability in precaudal and caudal 

 arrangement, as well as in total vertebrae. Verte- 

 bral numbers ranged from 38 to 40, and total 

 variability from the normal counts of 18 + 21 = 39 

 and 19 + 20 = 39 ranged from 1.97c for T. atlanti- 

 cus to 14.69c {or Thunnus spp. Frade (1932) found 

 13.6% variability for T. thynnus compared to my 

 5.1%, and he reported eight specimens with 38, 

 six with 40, and one with 41 vertebrae. Gibbs 

 and Collette (1967) doubted Frade's (1932) high 

 counts of 40 and 41 vertebrae, but they confirmed 

 Godsil and Byers' (1944) specimen of T. thynnus 

 with 38 vertebrae. Otherwise, they report no 



variability in vertebral numbers from more than 

 200 skeletons, except for three abnormalities 

 where two adjacent centra were fused. I found one 

 such "fusion abnormality" in a T. thynnus with 

 16 + 22 = 38 from more than 400 Thunnus 

 specimens examined. 



First Ventrally Directed Parapophysis 



(Figure 3) 



Ventrally directed parapophyses are already 

 present on the anterior centra in the smallest 

 (8 mm SL) specimens. There are two parapophy- 

 ses per centrum. Posteriorly, these two structures 

 become larger and finally join to form the haemal 



DORSAL FIN PTERYGIOPHORE 



NEURAL SPINE 



HAEMAL PREZYQAPOPHY8I8 / ^ANAL FIN PTERYGIOPHORE 



1ST HAEMAL SPINE 



Figure 3. — Relationship of the axial skeleton to the fin supports and fins in Thunnus thynnus, 24 mm SL. 



566 



