338 ON THE CHARACTER OF THE "8" SOUND [397 find references to wave-lengths 31-2mm.,l-304in. = 33>lmm.,l-28m. = It is true that these are of a pitch too high for Kohler's optimum, which at ordinary temperatures corresponds to a wave-length of 40'6 mm., or 1/60 inches; but 'they agree pretty well with the pitch found for actual hisses in my observations with Enock. Prof. Titchener has lately returned to the subject. In a communication to the American Philosophical Society f he writes : "It occurred to me that the question might be put to the test of experiment. The sound of a Galton's whistle set for 8400 v.d. might be imitated by the mouth,, and a series of observations might be taken upon material composed partly of the natural (mouth) sounds and partly of the artificial (whistle) tones. If a listening observer were unable to distinguish between the two stimuli, and if the mouth sound were shown, phonetically, to be a true hiss, then it would be proved that the whistle also gives an s, and Lord Rayleigh would be answered. " The experiment was more troublesome than I had anticipated ; but I may say at once that it has been carried out, and with affirmative result." A whistle of Edelmann's pattern (symmetrical, like a steam whistle) was used, actuated by a rubber bulb ; and it appears clear that a practised operator was able to imitate the whistle so successfully that the observer could not say with any certainty which was which. More doubt may be felt as to whether the sound was really a fully developed hiss. "Reliance seems to have been placed almost exclusively upon the position of the lips and tongue of the operator. I confess I should prefer the opinion of unsophisticated observers judging of the result simply by ear. The only evidence of this kind mentioned is in a footnote (p. 328) : " Mr Stephens' use of the word ' hiss ' was spontaneous, not due to suggestion." I have noticed that sometimes a hiss passes momentarily into what may almost be described as a whistle, but I do not think this can be regarded as a normal .9. Since reading Prof. Titchener's paper I have made further experiments with results that I propose to describe. The pitch of the sounds was determined by the sensitive flame and sliding reflector method, which is abundantly sensitive for the purpose. . The reflector is gradually drawn back from the burner, and the positions noted in which the flame is unaffected. This phase occurs when the burner occupies a node of the stationary waves. It is a place where there is no to and fro motion. The places of recovery are thus at distances from the reflector which are (odd or even) multiples of the half wave-length. The reflector was usually drawn back until there had been five * Scientific Papers, Vol. I. p. 407; Vol. n. p. 100. , t Proceedings, Vol. LIU. Augustó ^December,, 1,9.14, p. 32$.this quantity, it is not surprising that the influence of linear scale should fail to manifest itself. cm.2 '