184 ANNUAL OP SCIENTIFIC DISCOVERY. 



ends of the spectrum : one side, from red up to a yellow shade, a 

 little greenish, on the other side, from violet up to a blue shade, also 

 a little greenish. The shades, however, in the middle of the spectrum, 

 in which the green preponderates, can not give white any other homo- 

 geneous color. Their complement is purple, and must be com- 

 pounded by violet and red. The complementary colour of red 

 is greenish blue, of orange, sky blue, of yellow, indigo, of 

 greenish yellow, violet. The author found, moreover, that the com- 

 plementary colors are arranged in the spectrum in a most irregu- 

 lar manner. As the breadth of the differently colored bands in pris- 

 matic spectra depends not only on the wave-length, but on the sub- 

 stance of the prism, he refers the following results to interferential 

 spectra, where the distance of two colors is proportional to the differ- 

 ence of their respective wave-lights. If you pass with an equal velocity 

 through the different colors of such a spectrum, the shade is altered 

 very slowly at both its extremities, in the red and violet, but in those 

 parts where the complements of red and violet are placed, in the 

 greenish yellow and the greenish blue, the shade alters very rapidly, 

 so that the distance of extreme red and golden yellow is about ten 

 times greater than the distance of their complementary colors, green- 

 ish blue and sky blue. The author observed two circumstances in these 

 experiments which had prevented him in his former experiments from 

 finding other complementary colors than yellow and indigo, At first, 

 according to the peculiar distribution of complementary shades in the 

 spectrum, the said colors were able to give a larger white spot than 

 the others. Secondly, it appeared to be very difficult to the human 

 eye, which is not quite achromatic, to find and to keep the right focal 

 length for objects illuminated by two kinds of homogeneous rays of 

 very different refrangibility. Indigo aud yellow are of less dif- 

 ferent refrangibility than any other pair of homogeneous comple- 

 mentary colors, and are therefore easily combined. Others, as red 

 and greenish blue, on the contrary, are united in the same field of the 

 retina with great difficulty. Finally, the author gave some remarks on 

 the best method for bringing the whole variety of colors into a system. 

 He stated that Newton's colored disc appeared to be the most simple 

 and complete manner. Some points, however, are to be changed. First, 

 not only the seven principal colors of Newton must be arranged on the 

 margin of the disc, but the whole definite number of them existing in 

 the spectrum, so that complementary colors are placed in the opposite 

 ends of the same diameter. Secondly, the two ends of the spectrum 

 cannot meet together, but must be separated by an interval, where the 

 complementary color of the green shades, namely, purple, is to be in- 

 tercalated. The commonly-received theory of three principal colors 

 includes a restriction of Newton's method contradictory to the author's 

 former experiments. 





