GOETHE'S ' tfAKBESLEHRE.' 59 



bid media come into play ? If they exist, where are 

 they ? The poet's answer to this question is subtle in 

 the extreme. He wanders round the answer before he 

 touches it, indulging in various considerations regarding 

 penumbrse and double images, with the apparent aim of 

 breaking down the repugnance to his logic which the 

 mind of his reader is only too likely to entertain. If 

 you place A white card near the surface of a piece of 

 plate-glass, and look obliquely at the image of the 

 card reflected from the two surfaces, you observe two 

 images, which are hazy at the edges and more dense 

 and defined where they overlap. These hazy edges 

 Goethe pressed into his service as turbid media. He 

 fancied that they associated themselves indissolubly 

 with his refracted rectangles that in every case the 

 image of the rectangle was accompanied by a secondary 

 hazy image, a little in "advance of the principal one. 

 At one edge, he contended, the advanced secondary 

 imnge had black behind it, which was converted into 

 blue ; while at the other edge it bad white behind it, 

 and appeared yellow. When the refracted rectangle is 

 made very narrow, the fringes approach each other and 

 finally overlap. Blue thus mingles with yellow, and 

 the green of the spectrum is the consequence. This, 

 in a nutshell, is the theory of colours develorei in 

 the ' Farbenlehre.' Goethe obviously regarded the 

 narrowing of the rectangle, of the cylindrical beam, or 

 of the slit from which the light passed to the prism 

 according to Newton the indispensable requisite for the 

 production of a pare spectrum as an impure and 

 complicated mode of illustrating the phenomenon. 

 The elementary fact is, according to Goethe, obtained 

 when we operate with a wide rectangle the edges only 

 of which are coloured by refraction, while the centre 

 remains white. His experiments with the parchment 



