|Share page||Visit Us On FB|
THE PHONOGRAPHIC METHOD 63
eral over a seventh tone away, but how often they may have been nearer the other of the two three-fourteenth notes than the one noted is a matter of some uncertainty. Besides the difficulty of comparing two closely similar pitch distances, there is another source of error peculiar to this latter class of cases, which suggested itself to me after the completion of Shiashtasha No. 2, and which I endeavored to guard against thereafter. The nature of these cases demands the repeated sounding of two adjacent harmonium semitones s'and s" with the phonograph note p' to be judged; thus p'-s', p'-s". The semitone s' being in general a clearer and dis-tincter tone than the note p', it might happen that the judgment of difference would be made between the after image of s' and the new semitone s" instead of between p' and s." In this event the result of any individual double comparison would be an estimate of p' as nearer that one of the adjacent semitones which happened to be sounded first. Although it may seem that error from this source would be most unlikely, I felt assured at the time that the danger was a real one. In both the (—) and the = class of comparisons I endeavored to eliminate the influence of fatigue by pausing after several trials, and then, before the familiarity with the notes to be judged had passed away, noting with a new stress of attention the first impression of the rested ear. But it was more difficult to get an opinion that would stand the test of repeated verification in the = class of comparison than in the (—) class. An indication of the final opinion was in both these instances sometimes given by the fact that, while as above noted it was possible by an act of will to realize in perception either of the two opposite possibilities about the relation of the notes under judgment, the resistance of the sensational material to this violence seemed greater in the one alternative than in the other.
I should perhaps hardly have attempted these minuter differentiations (—) and = of pitch relation at all had not a trial made in the course of the studies of Chinese performance above mentioned given me reason to think my ear might be trusted to report a phonographic reproduction according to this general method, in the main correctly to comparatively small fractions of a tone. Taking the mean of several