1 873.] Colours and their Relations. gi 



pure red, is turned on a pure white, the complementary- 

 green which it sees is not pure, but is that mixture of blue 

 and green which is needful to complement the pure red, in 

 order to constitute a perfect white. So, when the eye first 

 dwells on a pure green, the red which it subsequently sees 

 is not pure, but that mixture of red with blue which, in 

 order to constitute a perfect white, is needful to complement 

 the pure green. From these observations it follows, that in 

 Newton's rings, the reflected and refracted tints, being com- 

 plementary to each other, cannot be pure colours, such as 

 are those of the diffracted spectrum ; but there must be at 

 least three pure colours in every opposing pair of the New- 

 tonian rings. 



From the foregoing sketch it will be perceived what an 

 additional charm has been thrown around the subject of 

 colour by the discoveries of Natural Philosophy. By the 

 appliances of which that science avails itself we are, as it 

 were, furnished with additional organs of vision, and enabled 

 to contemplate natural beauties, of which the human mind 

 had, before those discoveries, hardly formed a conception. 

 And then there returns upon us the startling fact, that all 

 these wonderful and beautiful phenomena are nothing more 

 than mere variations in the rates of certain minute vibra- 

 tions, — just as are the notes of various musical instruments 

 in the case of sound, whose melodies and harmonies have 

 thus, to a certain extent, their analogies in those of colour. 

 The nature and scope of these analogies will be considered 

 in the remaining part of this paper. 



Part III. 



The analogy between colours and musical tones has pre- 

 sented itself to many minds, and there has been among 

 scientific men much discussion as to its nature and extent. 

 The grounds on which those who have contended for a 

 perfect correspondence between the colours of the spectrum 

 and the notes of the musical scale have based their argu- 

 ment, were at one time supposed to be stronger than they 

 actually are. 



The case is greatly complicated by the uncertainty which 

 prevails in regard to what really constitutes the true musical- 

 scale. The mathematical idea of a perfectly musical scale 

 is one that- should divide the octave into twelve equivalent 

 semitones, forming a regular geometrical progression. For 

 the purpose of comparison with the actual musical scales, 

 this ideal scale is here given, with the relative number of 



