PSYCHO-PHYSICAL METHOD 129 



may be not fully prepared, or the accommodation of the 

 lens of the eye may be imperfect at the moment of 

 observation. 



Hence, in order to refine our result we must repeat 

 the comparison many times at each step of increase of 

 width of our windows. We shall then find that, when 

 the difference of brightness is very slight, we shall fail to 

 detect it in a certain proportion of cases, while in other 

 cases we do detect it. But so long as we truly detect the 

 difference we shall find in a given number of judgements, 

 say 20, an excess of judgements of difference over judge- 

 ments of no difference, and so we can go on diminishing 

 the difference of brightness of the two flashes until the 

 stage is reached at which this excess of judgements of 

 difference is abolished. 



By means of this procedure we reach a much more 

 accurate determination of the action-time than is pos- 

 sible by means of a single comparison or of comparisons 

 repeated two or three times only. This is known as the 

 method of right and wrong cases. It was devised by 

 Fechner to eliminate the uncertainties arising from the 

 variability of judgement due to the many factors by which 

 it is liable to be favourably or unfavourably affected. 



This procedure seems, then, to be satisfactory. The 

 physical conditions seem good, and we avoid the error due 

 to continued exposure of one of the circles to be compared, 

 and also that due to physiological interference or contrast; 

 the psychological conditions are good because the suc- 

 cessive flashes come at intervals of i| seconds, and that, 

 as we know from experience, is an interval very favour- 

 able to exact comparison of any two impressions ; lastly, 

 by applying the method of right and wrong cases we still 

 further refine the result. And the figure we find in this 

 way lies about midway between the extremely divergent 



st. K 



