112 Sir J. F. W. Herschel on the Action of the Rays 
colouring matter, both being very speedily bleached by ex- 
posure to light. 
193. Viola odorata.—Chemists are familiar with the colour 
of this flower as a test of acids and alkalies, for which, how- 
ever, it seems by no means better adapted than many others; 
less so, indeed, than that of the Viola tricolor, the common 
purple Iris, and many others which might be named. _ It of- 
fers, in fact, another, and rather a striking instance of the si- 
multaneous existence of two colouring ingredients in the same 
flower, comporting themselves differently, not only in regard 
to light but to chemical agents. Extracted with alcohol, the 
juice of the violet is of a rich blue colovr, which it imparts in 
high perfection to paper. Exposed to sunshine, a portion of 
this colour gives way pretty readily, but a residual blue, rather 
inclining to greenish, resists obstinately, and requires a very 
much longer exposure (for whole weeks indeed) for its de- 
struction, which is not even then complete. Photographic 
impressions, therefore, taken on this paper, though very pretty, 
are exceedingly tedious in their preparation, if we would have 
the lights sharply made out. 
194. The residual tint thus outstanding, after long expo- 
sure, is turned, not green, but yellow, by alkalies; or, if 
greenish at first, a very few hours suffice for the destruction of 
the slight remnant of blue, and the consequent appearance of 
the yellow colour. Reasoning on this fact, as well as on the ac- 
tion of light above mentioned, it seems highly probable that 
the tincture in question holds in solution two distinct colouring 
principles, of which the one (greatly preponderant in quan- 
tity) is destructible by light, and either destroyed or turned 
green by alkalies; the other, indestructible by light, and either 
naturally yellow in colour or changeable into yellow by alka- 
line agency. 
195. This view of the composite nature of the colour in 
question receives corroboration from the habitudes of the al- 
coholic tincture above mentioned, when rendered green by 
admixture of carbonate of soda. On making this addition it 
becomes evident that a large amount of colour has been de- 
stroyed; the green tint imparted by it to paper being far less 
intense than might be expected from the intensity of the ori- 
ginal hue, and from the trifling dilution caused by the sinall 
quantity of alkaline liquid required to effect the change. 
What remains is a fine green; but when exposed to light, the 
blue constituent alone of that green is destroyed, and a resi- 
dual tint of pure yellow, which is very indestructible by light, 
is left. Exposure of a slip of such paper to the spectrum 
proves this change to be operated almost wholly by rays less 
