Held and Schneider and others have postulated two, functionally separate, modes of vision-one concerned with localization, the second with identification-and have suggested that the superior colliculus is crucial for the perception of location. Colliculectomized hamsters were tested on two localization tasks. No deficit in depth perception was observed. Perceptual-motor integration necessary to perform a search task was severely disrupted, however. These results suggest that the superior colliculus functions in the control of fine visual-motor coordination.Research by Schneider (1969) has specified anatomical loci for two functional mechanisms of vision in golden hamsters. Colliculectomized subjects in a discrimination task could discriminate simple patterns, but could not orient or locomote directly to the correct pattern. Conversely, animals with ablations of the visual cortex displayed an impaired ability to discriminate patterns, but no loss of ability to orient and localize.These results, in addition to those of Held (1968), Ingle (1967), and Trevarthen (1968, have led these authors to postulate two different modes of visionthe first concerned with "location" and the second with "identification." The term "location," however, refers to a wide range of behaviors which require the pickup of different types of stimulus information, and which demand varying degrees of visuomotor integrative ability in order to be performed. The question arises that perhaps a lesion in a specific area of the brain could interfere with an animal's performance on one task involving "location" while another behavior that can be categorized under the same heading remains intact. That this may be the case is suggested by previous studies involving ablations of the superior colliculus. For example, Lashley (1937) demonstrated that rats with lesions in the superior colliculus could negotiate jumps of between 20 and 40 cm to a platform without any difficulty, and found that the The authors wish to thank Gerald Schneider for his advice on the surgical procedures used in these experiments, P. A. Cabe for his suggestions concerning the open-field test, and E. J. Gibson for her advice and encouragement. Requests for reprints should be sent to Richard R. Rosinski,

Three experimental situations were used to assess the ability of the golden hamster to avoid collision with an approaching object. In all three situations (one optical loom and two real looms), the animals did not avoid the looming stimulus. Any responses that were made would not have resulted in successful avoidance of an approaching object. The golden hamster lacks the ability to avoid optically specified impending collision.The perception of impending collision and the ability to avoid a rapidly approaching object is of clear survival value to an organism. Perception of imminent collision is possible on the basis of optical changes that are projected to the eye of an organism. Gibson, Olum, and Rosenblatt (1955) have shown that the patterns of angular velocities (motion perspective) visually available to · a moving organism potentially specify the distance of a point from the organism and the velocity of the observer. In addition, motion perspective specifies the rate of approach of an object, time to collision with an object, and path of approach of an object relative to a stationary organism. Gibson (1958) has argued that such information is of crucial importance in the behavioral ecology of an organism since it provides the animal with a basis for capturing food, avoiding a predator, and locomoting around obstacles.Many investigators (Ball & Tronick, 1971;Bower, Broughton, & Moore, 1971 ; Schiff, 1965; rronick, 1967) have demonstrated that optical information which specifies a looming object is used by human infants, chicks, rhesus infants, frogs, and crabs to avoid impending collision. The pickup of this information is specific enough to allow these animals to distinguish between a receding object, an approaching object on a collision course, and an approaching object on a noncollision course. Since in the normal ecology of the animal, these events may signify an attacking predator, the ability to pick up visual information for collision and respond appropriately may have greater biological significance. Furthermore, the ability to avoid imminent collision is so important that it is apparently little Alexander W. Siegel sponsors this paper and takes full editorial responsibility for it. 53affected by specific experience . For example, Bower, Broughton, and Moore (1971) found that 6-day-old human neonates showed a pronounced avoidance response to looming. Hein, Gower, and Diamond (1970) found that kittens deprived of visual experience by being reared in diffuse light for 4 weeks showed triggered extension of the forepaws to impending collision with a surface. And, Schiff, Caviness, and Gibson (1962) found that avoidance responses to optically specified collision did not habituate over repeated presentations. METHOD Eight adult male golden hamsters (Mesocricetus auratus)were used as subjects in three different situations to determine the loom avoidance ability of the animals. Schiff (Note 1) has suggested that loom avoidance behavior may depend in part on the amount or kind of stimulus infonnation pro...