Kishinouye (1923: 322) stated that the anterior head 

 is "more or less roundish in cross-section in Thunnus; 

 but more or less flattened in Parathiinnus and 

 Neothunnus. . . ." We can find no significant dif- 

 ference among any of the species in this character. 

 Furthermore, we cannot confirm de Sylva's conten- 

 tion (1955:14) that the process is oblicjue to the 

 vertical limb in T. atlanticus but forms a right angle 

 in the other species. 



In T. alalunga the metapterygoid (fig. 7) is nar- 

 rower than in other species. This condition can be 

 indicated by the proportion of the length of the 

 anteroventral margin to the posteroventral margin 



Figure 8. — Quadrate of (left) Thunnus thynnus, (right) 

 T. alalunga, same specimens as in fig. 6, showing measure- 

 ments of E — total width, F — width of horizontal edge, G — 

 length. 



Figure 7. — Metapterygoid of (left) Thunnus thynnus, (right) 

 T. alalunga, .same specimens as in fig. 6, showing measure- 

 ments of C — anteroventral margin, D — posteroventral 

 margin. 



(measured in each case to the midpoint of the arc of 

 the posteroventral edge.) In our material, this pro- 

 portion is as follows: T. alalunga {N=32) 1.1-1.8, 

 x=lA; T. thynnus (A^=43) 1.6-2.6, x=2.0; T. 

 maccoyii {N=17) 1.4-2.7, x= 1.9; T. obesm {N=37) 

 1.5-3.1, x=2.0; T. atlanticus {N= 19) 1.4-2.1, x= 1.7; 

 T. albacarcs {N=58) 1.5-3.1, x=2.1; T. tonggol 

 {N=4) 1 .8-2.2, x= 2.0. These proportions are simi- 

 lar to those given by Godsil and Byers (1944: 86) 

 for Pacific tunas, but provide even better distinction. 

 Godsil and Byers measured each margin to the 

 "most ventral point," which seemed more nebulous 

 to us than the midpoint of the arc. Their figures of 

 1.0-1.5 for T. alalunga and 1.3-1.9 for T. albacarcs 

 and the other species include slightly lower propor- 

 tions than were found in our specimens, but the 

 conclusions are nevertheless similar. 



Again, in T. alalunga the quadrate (fig. 8) is 

 slightly narrower than in the other species. The 

 proportion of length to total width in our specimens 

 is as follows: T. alalunga (A^=35) 1.5-2.1, 5=1.8; 

 T. thynnus (.V=44) 1.2-1.6, x=lA; T. maccoyii 

 (iV=16) 1.3-1.5, x=1.4; T. obesus (iV=37) 1.4-1.7, 

 x=1.5; T. atlanticus {N=20) 1.3-1.8, 5=1.6; T. 

 albacarcs {N=57) 1.4-1.8, 5=1.5; T. tonggol (iV=4) 

 1.4-1.6, x= 1.5. The proportions of the same length 

 to width of the horizontal dorsal edge are : T. alalunga 

 (A^ = 35) 2.1-2.7, 5=2.5; T. thynnus (Af=44) 1.6-2.1, 

 5=1.9; T. maccoyii (iV=16) 1.6-2.0, 5=1.9; T. 

 obesus (N=37) 1.8-2.2, 5=2.0; T. atlanticus (N=20) 

 2.0-2.3, 5=2.1; r. albacarcs (A^=57) 1.8-2.2, 5=2.0; 

 T. tonggol (A^=4) 1.8-2.2, 5=2.0. We are not 

 certain where Godsil and Byers (1944: 87) measured 

 the width, but they gave proportions of 1.8-2.3 for 

 Pacific T. alalunga and 1.6-1.8 for other species. 

 Thus there is agreement in order of magnitude, but 

 their proportions are generally higher than our first 

 and lower than our second. 



The subopercle of T. thynnus and T. maccoyii 

 (fig. 9) differs, with few exceptions, from that of the 

 other species in being relatively narrow and in having 

 the anterodorsal margin almost vertical in its lower 

 two-fifths to one-half, followed by a decided change 

 in slope of the upper portion. In the other species 

 there may be a very short perpendicular portion, less 

 than one-fifth of the length, before the oblique slope 

 begins, or, most often, there is an almost straight or 

 very slightly convex oblique edge. This finding 

 conforms with the observations of Godsil and Byers 

 (1944: 101) for the Pacific forms and presumably 

 also with the observations of Kishinouye (1923 : 325), 

 although his statement is less clear. 



ANATOMY AND SYSTEMATICS OF TUNAS 



75 



