I907] TITCHENER AND PYLE— JUDGMENT OF DISTANCE. 101 



If the illu-sion motive is effective, the figures of the first cohimn 

 will be less, those of the second greater, than the corresponding 

 figures of the third column. The observers Ug and Un thus give 

 negative results ; P and S give one positive and one negative ; and G 

 alone gives two positive results. But a glance at the magnitude of 

 the differences, as compared with that of the MV , shows the ab- 

 surdity of drawing any conclusion of this sort from the results.^ 



Experiment III. — In view of the negative outcome of this second 

 set of experiments, it seemed necessary to raise the question sug- 

 gested in the introduction to this paper : the question whether an 

 illusion motive, if eft'ective at all, must not be effective at full 

 strength. Have we any reason to suppose that a dim illusion-figure 

 will produce a small effect, and a clear illusion-figure a marked 

 effect upon the observer's judgment? We approached this problem 

 as follows. 



The horizontal line and its three markers were removed from 

 the white bristol-board, and their place was taken by a length of 

 fine black sewing thread. To render the shadows of the iMiiller- 

 Lyer angles intensive, we placed three lamps behind the apparatus, 



^The averages, as was stated above, have been calculated in Dunlap's 

 way. We have submitted the figures to all the other forms of methodically 

 permissible treatment known to us, and can discover no trace of the influence 

 of the illusion motive. 



Jastrow declares {op. cit., 417) that if, after experiments with the jMiiller- 

 Lyer figure, one proceeds to experiment upon the estimation of visual distances 

 in Dunlap's way, /'. e., with " the shadow-strokes reduced to such a degree of 

 faintness that the eye fails to detect their presence," the observer will judge 

 " naturally with diminished confidence " as to the relation of the two lines. 

 Why? If the shadows are not seen, how can the observer's confidence be 

 diminished? He is simply called upon to compare two horizontal lines. — 

 On the general question of ' degree of confidence,' see esp. G. E. Miiller, 

 '' Gesichtspunkte u. Tatsachen der p.sychophysischen Methodik," 1904, 21 f. 



