Nov. 2.-;, 1881.] 



KNOWLEDGE 



71 



Fig. 9. — A Xest of Illusions. 



Fi_'. ^ has been regarded hj Mr. Proctor as having possibly 

 .soiu- btariiig on the irregular shape, which the outline of 

 the planet sometimes seems to have. It is referred to by llr. 

 Webb in his " Celestial Objects for Common Telescopes," 

 ■ivho illustrates it, however, by a figure in which the dis- 

 tinctive characteristic is lost ; the straight line in his figure 

 <loes not cross the curved lines, and the suggestion of 

 curvature, if it exist at all, is too slight to be readily 

 recognised. 



The eflect of position on illusions of this sort can hardly, 

 I think, be better illustrated than it is in Fig. 9, which 

 may be described as " a nest of illusions." 



Thus, First, the sides of the square ABCD, which are 

 1 eallv straight, appear to be cuned inwards. 



Secondly, the sides of the square abed, which are really 



straight, appear to Ije curved outwards, as (in less degree) 

 do the sides of the square 1 2 .3 4. 



Thirdly, the space between the two large heavy circles 

 mir and »:n, appears to be superposed on the rest of the 

 figure. 



Fourthly, arc no (lighter curve) appears to belong to a 

 smaller circle, instead of belonging, as it really does, to the 

 same circle as the arcs Im and pr, the same holding with 

 the other corresponding circular arcs of the figure. 



Fifthly, it is difficult to realise that the space between 

 the arcs st and ij is as broad as that between the arcs rv 

 and zz, &c. 



Sixthly, the circles between no and zz appear to draw 

 closer together, though in reality concentric, as they 

 approach the heavj- circle noz, lirc. 



