FISHERY BULLETIN; VOL. 74, NO. 3 



An exceedingly interesting and simple exper- 

 iment was undertaken by Radakov (1972) with 

 21 young Pollachius virens (Linnaeus) of 8 to 9 

 cm. These were placed in a tank measuring 

 1.6x7x0.3 m. It was divided into two equal com- 

 partments by a clear glass partition. All the fish 

 were placed in one compartment. The experiment 

 consisted of transferring the fishes, one at a time, 

 to the other compartment. With 20% or less of the 

 fishes transferred, the smaller group tried contin- 

 uously to swim through the glass partition in their 

 efforts to rejoin the others. Above that percent- 

 age, the two larger groups, between 30 and 40% of 

 the fish on both sides tried to form a common 

 school with the glass partition cutting through it. 

 Continuing the transferring, a reverse series of 

 the attitudes described above was obtained. 



Movements of Individuals 



The study of travel by individual fishes within a 



school has difficult and tedious aspects, as is 



evident from the preceding. The subject has not 



attracted many investigators as witness the 



paucity of comments on it in earlier papers. An 



examination of Figure 12 shows quickly that such 



internal traveling is neither negligible nor slight, 



at least in very loosely organized schools, but is 



probably much less so in very tight schools. 



Because of this, the geometrical properties of 



schools have been considered chiefly in a single 



layer of fishes, i.e., in terms of plane geometry. 



Schools of greater depth present special difficulties 



in obtaining adequate field data, as it is necessary 



to invoke the complications of the third dimension 



while the fishes are often so closely packed that 



visual perception within the school is severely 



restricted. In addition, there are further problems 



incident to the fishes' continual activity. This is 



particularly difficult in efforts to recognize the 



rhombohedron of Figure IB. The present efforts 



have yielded some hints that suggest support to 



our thesis. 



The vertical structure of schools and vertical 

 mixing within them is much more diflScult to 

 handle. This is evidently owing partly to the 

 greater inherent difficulties in three dimensional 

 plotting and partly in the nature of fish mor- 

 phology and methods of propulsion. The influences 

 of each fish on the others in the same horizontal 

 plane are greater than in any other direction 

 because both vision and locomotor mechanics 

 operate primarily in that plane. That is, optical 



axes of schooling fishes lie in that plane and the 

 propulsive mechanism produces forces operating 

 in it.** 



It is consequently less difficult to compare the 

 relative amount of shifting about in the horizontal 

 plane as compared with that in the vertical. 

 Although we have no clear observations or photo- 

 graphs of a fish sinking to the layer below it or 

 rising up from one below, there are many in- 

 stances of evidently "uncertain" fishes seen 

 between distinct layers or ones dropping slightly 

 below, as in Hunter's (1966) figure 2. 



Shape and Size of Schools 



The closed figure that forms the outline of a 

 school is a remarkably flexible boundary subject to 

 continual transformation. These changes are 

 produced by a large variety of influences both 

 intrinsic and, by a vastly greater number, extrin- 

 sic. Obviously, the most important intrinsic factor 

 in holding a school together is the impulse that 

 causes fishes of one kind to assemble, respecting 

 each others necessary swimming room and ac- 

 cepting a common polarization. 



The fishes that are outermost along the sides of 

 a school do not form a special boundary layer any 

 more than do those at the front form "leaders." 

 Those at the side surfaces differ from the rest only 

 in that they lack fellows on one side. Like those at 

 the front, they are continually changing as their 

 aggregating tendency apparently moves them 

 toward a more central position. 



Aside from temporary weakening of the bonds 

 by such things as vigorous feeding, reproduction, 

 the coming of a suflRciently dark night, or a 

 particularly violent disturbance, the basic school 

 structure is continuous in obligate schoolers. In 

 facultative schoolers, the school is periodic or of 

 occasional occurrence. True semipermanent inter- 

 mediates between these two ordinarily distinct 

 modes are not easy to find and are uncertain at 

 best. 



The intrinsic influences divide naturally into 

 two groups, the first being those of nonorganic 

 elements. Common examples of these are light, 

 water currents, shoreline, sharply mottled bottom 

 patterns, and obstructions. Sharp discontinuities 

 of any of these are especially influential. Organic 



*A comparison of fish schools with those of cetaceans should be 

 illuminating because the propulsive efforts of the latter operate 

 in the vertical plane. 



482 



