Table 5. — Position of longitudinal and latitudinal first moments at the beginning of Decem- 

 ber and at the end of March for each year during the study period. The average movement 

 per day was calculated by converting the December-March differences into miles and divid- 

 ing by 120 



Resultant = V (1.65)^ + (6.25)^ = 6.46 miles (11.97 km.) day" 



tical motions of their environmental medium to 

 maintain equilibrium and thus do not need to 

 sustain a continual threshold velocity to avoid 

 sinking. We felt that some of these possibilities 

 could be explored by studying the apparent ve- 

 locities of the tagged fish reported by Otsu and 

 Uchida (1963). Accordingly, we have plotted 

 distance between the location of marking and 

 recapture for the albacore tagging data sum- 

 marized by these authors as a function of time 

 at liberty (fig. 16). We can see from these data 

 that the points fall into essentially two groups. 

 The first is a group of relatively short-term 

 recoveries (time at liberty < 100 days), and the 

 second is a group of relatively long-term re- 

 coveries (time at liberty > 200 days). The 

 latter group shows that there is a tendency for 

 the fish that are at liberty the longest to be re- 

 captured nearer to the site of tagging. We in- 

 terpret this tendency to reflect a circulation 

 pattern of the albacore in the North Pacific; 

 hence low velocities as deduced from the dis- 

 tance traveled by tagged fish could result from 

 the fish actually moving a short distance or 

 this returning of the fish to the site of tagging 

 after a sojourn to some other area. We should 

 be cautious, therefore, in interpreting any but 

 the maximal velocities as actual swimming 

 speeds. These maximal velocities in the tag- 

 ging data are about 15 miles (28 km.) day"' , a 



velocity which is about double that deduced 

 from the moments. We conclude that if the in- 

 ferred transpacific velocity of albacore from 

 tagging studies is equivalent to the velocity of 

 the albacore on the longline grounds, as in- 

 ferred from the motion of the moments (under 

 the assumption that apparent movement is 

 equivalent to actual movement), then the alba- 

 core are, on the average, constrained to move 

 each day within an area circumscribed by two 

 intersecting lines; the length of each being 15 

 miles (28 km.) and the distance between the 

 intersections being 7 miles (13 km.). A con- 

 siderable discrepance, between 15 miles (28 

 km.) day"! and 86 km. day'l, still remains. We 

 conclude that the most likely cause for this 

 discrepance is that the albacore generally 

 follow tortuous routes between any two, even 

 closely spaced points. If tunas actually do soar 

 in the same fashion as the birds studied by 

 Cone (1962), then perhaps they also follow the 

 trochoidal path described by Cone. Such a path 

 would account for a considerable difference 

 between a velocity on a straight line connecting 

 a starting point and a terminal point and a ve- 

 locity computed on the actual nonlinear path 

 between the two points. A nonlinear path is 

 very likely considering that a tuna in some 

 fixed time must swim a considerable distance 

 to maintain hydrostatic equilibrium, especially 



22 



