QUANTITATIVE ESTIMATES OF SENSORY EVENTS 281 



of experimental accuracy and those of the comfort of the subject. Only 

 one standard intensity was investigated, but both time-orders were used. 

 The rather naive conclusion was drawn that certainty of judgment grew 

 with increasing difference of intensities. Individual differences between 

 the results of the two experimenters, who acted as their own subjects, also 

 received comment. 



Fechner in the Elemente (12) derived his data on sound intensity from 

 results obtained by Volkmann, who carried out two series of experiments. 

 In the first, he used a simple improvised sound-pendulum, consisting of a 

 strong knitting-needle as axis, and a wooden hammer which struck against 

 a four-sided glass flask. Two heights were found such that in the majority 

 of cases the observer could tell which gave the stronger sound, and observa- 

 tions were made at four distances, varying from i\ to 18 paces. It was 

 found that judgment remained as sure and correct at all distances, and from 

 this it was concluded that the difference threshold was independent of the 

 absolute value of the stimulus. Experiments with freely-falling bodies 

 gave for two subjects out of three a ratio of intensities 3 : 4 for which a 

 difference could be accurately judged, while with a ratio 6 : 7 considerable 

 uncertainty occurred. 



A much fuller investigation was reported by Norr (42), who introduced 

 a number of refinements into his falling-bodies technique. A much wider 

 range of intensities was used, and catch-experiments were introduced, in 

 which the standard was presented with itself. Unfortunately, the numerical 

 results are such as to make the calculation of difference thresholds by any 

 of the ordinary procedures practically impossible. Norr, however, con- 

 cluded that differential sensitivity remained constant from the weakest to 

 the strongest sounds. 



(ii) Dissatisfaction with Norr's results seems to have been one of the 

 contributing causes of the work of the Leipzig group. This series of 

 researches is of particular interest in that the writers had first-hand contacts 

 with one another, even to the extent of opportunities of working with the 

 same subjects and the same instruments. One of their main interests was 

 the exact measurement of the sounds produced by the falling bodies, 

 discussed above. 



Tischer (56) and Lorenz (31) used Hipp's fall-apparatus. Tischer's 

 results show wide individual differences among his five subjects, and 

 considerable variation over a fairly small range, together with a progressive 

 improvement of discrimination with practice. Nevertheless, the results 

 are described as ' so gut wie constant.' Lorenz, on the other hand, was 

 aware that his results were insufficient for generalisation. He characterises 

 the constancy obtained as fairly satisfactory, and states that it might have 

 been better with greater care. 



Starke (50, 51), Merkel (36), and Mosch (39) used Wundt's improved 

 fall-apparatus, and introduced further experimental refinements. Mosch 

 laid particular stress on the ' error ' aspect of variations in the difference 

 threshold, and introduced further categories of judgment (' much greater,^ 

 ' much less '). Kampfe (23) and Ament (3) reverted to the use of the 

 sound-pendulum, making considerable improvements on the model used 

 by Volkmann. Ament's work shows increased recognition of individual 

 differences, and also a decided drop in the value of the difference threshold 

 after the weaker intensities had been passed. In both these respects 

 Ament anticipates the results of later experiments. 



A slightly different approach to the stimulus-sensation problem is seen 

 in the work of Merkel (37) and Angell (5). Merkel and Angell used the 

 method of ' Mean Gradation ' to find an estimated mid-point between two 



