Some experiments of the optomotor reaction using the Japanese killifish, Oryzias latipes TEMMINCK et SCHLEGEL, were carried out in the apparatus slightly modified from the one used by CRONL Y -DILLON and MUNTZ (1965) consisting in rotating pattern of alternating light and dark stripes around the glass vessel in which fish are held. The results obtained can be summarized as follows.
1. Wild fish do not show any clear positive optomotor reaction until they are well trained by 3 to 5 trial experiments. It takes 30 to 60 seconds even for welltrained fish to come to the state of following completely the moving pattern of stripes.
2. Fish show irregular optomotor reaction when they are reared in a small glass container, even if they have been duly trained. Besides, when reared in a small opaque container, they begin to react several minutes later than they do when normally reared.
3. Fish that have been shown the pattern of stripes prior to the experiment begin to react to the moving pattern of stripes earlier than those which have not been shown.
4. The starting time of the optomotor reaction in fish kept in darkness (darkadapted) is delayed correlatively to the time at which the process of light adaptation begins in retinas of fish held in the working apparatus.