Page:Philosophical Review Volume 1.djvu/363

No. 3.]

Experiments undertaken, in the first instance, to test Stumpf’s assertion that high tones have an especial sensation value as regards intensity. The test was made by finding reaction-times for different intensities of the same tone. In light of the experiments, the author also reviews the opinion of Wundt, Exner, and others, that, in general, reaction-times decrease with the increase in intensity of stimuli. M. found that, with a proper amount of practice and training of the attention, the reaction-times for strong impressions of sound arising from a series of stimuli differing considerably in intensity were equal, but that for very weak impressions the reaction-times were longer, notwithstanding practice and training of the attention. The cause of the increase of the reaction-times with weak impressions lies in the difficulty of perceiving them, and in the slower co-ordination of impression and motion, and not to purely physiological processes of transmission.

Author criticises the 'proportional' hypothesis of the relation of stimulus to sensation, finds there are as many forms of the hypothesis as there are advocates of it, shows that in any form it is meaningless, self-contradictory, and not capable of experimental proof, and supports the 'difference' hypothesis in the sense of a law of Apperception. Merkel's results drawn from experiments with several psycho-physical methods are not in agreement with one another, nor with any known psycho-physical law. The methods of doubled stimuli and mean gradations are tested experimentally, and the following conclusions are arrived at. 1) The method of doubled stimuli cannot be considered as of any value in psycho-physics. 2) The sources of error in the comparison of sound-intensities according to the method of mean gradations are numerous enough, when the variable stimulus is moved regularly in one direction, to obscure the actual relation of stimulus to sensation. 3) Irregular change of the variable in the method of mean gradations does not prevent a reliable comparison of differences of sound-intensities, and with this irregular change, the method is to be considered valid. 4) Freed from the disturbing influences of expectation, and stereotyped ways of judging, the comparison of broad differences of sound-intensities gives results corresponding to the requirements of Weber's law.