x'm»i>Wvl \y> vn the Hypothesis ofUndulationsiJl'Ati' 337 



of sunlight, is one of the components. The explanation given 

 by the theory of there being no prismatic colour complementary 

 to green is, that there is no colour the ratio of whose wave-length 

 to that of green can be expressed by sufficiently low numbers. 



(8) Again, it may be remarked, that, in sound-sensations, the 

 combination of waves whose leng-ths are in the ratio of 4 to 3 

 produces harmony; whilst, according to this theory of colour- 

 sensations, the same ratio of wave-lengths of yellow and blue 

 results in whiteness, which is analogous to unmusical noise. 

 There is, however, a remarkable fact bearing on this point which 

 must be taken into consideration. On proceeding to still greater 

 differences of wave-lengths, when. we come to a combination of 

 red and violet colours, the ratio of whose wave-lengths is nearly 

 that of the low numbers 3 and 2, the result is purple, a decided 

 colour, in which the eye seems to be able to distinguish the com- 

 ponents, as the ear distinguishes the components of a harmony. 

 The analogy to sound here returns, with the difference that light- 

 harmony does not appear to have so wide a range as sound-har- 

 mony. Possibly rose colour may be a harmonious result from a 

 mixture of wave-lengths of red and blue in the ratio of 4 to 3. 



(9) The method of combining colours by a painted revolving 

 disc, and that of receiving in the same direction colours trans- 

 mitted through, and reflected from, plate-glass (Helmholtz, Phil. 

 Mag. p. 530), are, theoretically speaking, methods of varying 

 the relative intensities of the two kinds of rays, and the value of 

 the difference /jl for each pair. In the case of the revolving disc, 

 in which different kinds of impulses act in succession, the degrada- 

 tion of the vibrations of the particles of the eye in the compara- 

 tively large interval between one kind of impulse and the next 

 succeeding one, is equivalent to a difference in the values of m 

 and m' for two kinds of impulse which take effect simultaneously. 

 The amount of difference, or the value of fju, depends on the velo- 

 city of the motion of the disc. It is probable, however, that suc- 

 cessive impulses can in no case be exactly equivalent to simul- 

 taneous impulses, because in the former the individual rays admit 

 of being combined so that the intervals between the impulses 

 will be different for each pair. It is also to be observed, that in 

 the other experiment, when the intensities of the two parcels of 

 light are so adjusted that neither colour is predominant, the 

 values of m and m' for the individual rays will generally be un- 

 equal, because the intrinsic intensity of a reflected ray is, ceteris 

 paribus, less than that of a transmitted ray. Hence a tendency 

 to whiteness in the compound may be expected. 



Several other inferences drawn by M. Helmholtz from his expe- 

 riments are in accordance with this theory. For instance, that the 

 extreme colours of the spectrum, red and violet, cannot be com- 



