294 CHEMICAL SENSES 



sequence, solutions stored in the container, at controlled rates. Thus, release 

 of a chemical trail can be established "upstream," "downstream," parallel or 

 at angles with the stream, in stagnant water or in currents of various flow 

 rates. 



The above outline of the experimental approach to the analysis of loco- 

 motor behavior in general and oriented behavior specifically, is to serve as 

 the background for a discussion of some of the results obtained in this 

 laboratory in the study of that behavior in some sharks. 



OBSERVATIONS ON THE GENERAL LOCOMOTOR 

 BEHAVIOR OF THREE ELASMOBRANCHS: 

 SCYLIORHINUS, MUSTELUS, AND 

 GINGL YMOSTOMA 



In the late fifties and early sixties, an extensive effort was made in the 

 author's laboratory to identify the main chemical components of the "body 

 odor" of trout and other teleosts (Kleerekoper and Mogensen 1959) olfac- 

 torily attractive to the sea lamprey, Petromyzon marinus (Kleerekoper and 

 Mogensen 1963). One component, first code-named amine F, later identified 

 as isoleucine methylester, when perceived in very low concentration by 

 either adult or larval Petromyzon, elicited general and oriented locomotor 

 activity, resulting in efficient localization of the stimulus source by olfaction 

 only (Kleerekoper and Mogensen 1963, Kleerekoper 1963). 



Later, a study was begun to verify the attractiveness of the substance to 

 teleosts and elasmobranchs. Early in that study, by long-term monitoring of 

 locomotion, it was established that, contrary to expectations, the spatial 

 characteristics of the locomotor patterns of neither the elasmobranchs 

 (Scyliorhinus and Mustelus) nor the teleost (Diplodus) under investigation 

 were randomly distributed in the symmetrical environment of the cylindrical 

 monitor, even in the absence of experimentally controllable cues (Kleere- 

 koper 1967). Both elasmobranchs, because of biases in their locomotor be- 

 havior, displayed a preference for certain pathways in the tank. One of these 

 was "handedness," that is, turning predominantly to either the right or the 

 left on emerging from a compartment. In a typical experiment, Scyliorhinus 

 stellaris might display a 1.6 ratio (n = 2056) and Mustelus mustelus a 0.6 

 ratio (n = 1492) of left/right turns. Obviously, such handedness alone 

 strongly biases the locomotor pattern, but in addition both species, on 

 emerging from a compartment, manifested a strong preference to bypass or 

 "jump" a set number of compartments before making the next entry 

 (Figures 16 and 17). That behavior, however, might be different following 

 left or right turns, as in Figure 18, in which the distributions of the number 

 of bypassed compartments are illustrated for Scyliorhinus. The differences 

 between left and right turns were much smaller in Mustelus (Figure 19). 



Numerous experiments, representing many thousands of events, con- 

 finned, in the above species, nonrandom locomotor patterns, which could be 

 accounted for almost entirely by the two biases mentioned. At the same 



