Fishes possess most of the senses which belong to the human mind. The visual and olfactory senses are well developed, being the chief factors in the detection of prey. The tactile senses are exceptionally acute, as is proved by the ease with which fishes detect disturbances in the water. This acuteness probably accounts for most cases of supposed audition. The extent to which fishes are able to sense temperature, pain, and muscular changes is as yet doubtful, and it has been supposed that they have certain senses which men do not possess. It has been conclusively shown that fishes have some capability for education. The apparently intelligent instincts of fishes, e.g., the spawning habits of the salmon, are to be explained as very simple reactions to immediate external conditions, and do not presuppose any high degree of mental development. It may be conjectured that the fish mind possesses a very simple form of perception, that it associates these percepts according to recency, frequency, and vividness, that it has the power of involuntary memory, and perhaps even some glimmerings of consciousness.
These articles describe an experimental investigation of the possibility of the distribution of attention. Previous experiments to determine this point are criticised as inconclusive in that their conditions did not really make distribution necessary. It was found that, in counting simultaneous series of sensations, the rate of counting decreased as the number of series increased. This decrease was much greater when the sensations to be counted were from disparate senses. These results, together with the introspection of the subjects, were interpreted as pointing to the conclusion that the attention was not really distributed, but fluctuated rapidly from one stimulus to another. In order to test this conclusion, an experiment was devised by which two series, differing only in the concentration and attempted distribution of attention, could be compared. Again the attempted distribution caused an increase in the reaction time, a result to be interpreted in favor of fluctuation rather than distribution. These results led to an investigation of Wundt's tachistoscopic experiments, which constitute the strongest evidence for distribution. The question is: Was the attention really divided in Wundt's experiments? Elaborate tachistoscopic experiments showed that conscious perception did not take place during the application of the stimulus, but came to consciousness in the form of a mental after-image. It was found that the reaction time again increased in proportion to the number of objects counted. This indicated that the perception was characterized rather by separate acts of attention than by its
