120 



SCIENTIFIC RECREATIONS. 



of the electric spark, when the box is perfectly dark. The box is open, and 

 rests on the table, B D, to allow of changing the object The conducting 

 wires of electricity are at h and i ; in the centre of the box is a strip of 

 cardboard, white on the side facing the spark, the light of which it shelters 

 from the eye of the observer and throws back again on the object. With 

 the electric light the illusion was completely perceptible with fig. 118, 

 while it disappeared altogether in fig. 120; with fig. 121 it was not 

 entirely absent, but when it showed itself, it was much more feeble and 

 doubtful than usual, though the intensity of light was quite sufficient to allow 

 of the form of the object being very distinctly examined. Thus two different 

 phenomena have to be explained ; first, the feeble illusion which is produced 

 without the intervention of movements of the eye; and secondly, the strengthen- 

 ing of the illusion in consequence of these movements. The law of contrast 

 sufficiently explains the first ; that which one perceives most distinctly 



Fig. 122. Observation of electric spark. 



with indirect vision is the concordance of directions with dimensions of the 

 same kind. We perceive more distinctly the difference of direction pre- 

 sented at their intersection by the two sides of an acute or obtuse angle, 

 than the deviation that exists between one of the sides and the perpendicular 

 which we imagine placed on the other side, but which is not marked. By 

 being distributed on both sides, the apparent enlargement of the angles gives 

 way to displacements, and changes of direction of the sides. It is difficult 

 to correct the apparent displacement of the lines when they remain parallel 

 to their true direction ; for this reason, the illusion of the figure is relatively 

 more inflexible. Changes of direction, on the contrary, are recognised more 

 easily if we examine the figure attentively, when these changes have the 

 effect of causing the concordance of the lines (which accord in reality) to 

 disappear ; it is probably because of the difference in aspect of the numerous 

 oblique lines of figs. 120 and 121 that the concordance of these lines escapes 

 the observer's notice. As regards the influence exercised by the motion of 



