means of interchange and transition. See pages 311 and 314. Every chord, including, of course, its bass note, contains, at least, one note that is sounded in the chord following; and thus, as applied to any two consecutive chords, the principle of putting like with like is fulfilled.

Harmony in music, therefore, may be said to be an effect of unity produced when complex wholes are grouped by putting together those that have like partial effects. The main result of this, as intimated on page 310, is consonance. When consonance is not complete, its general effect is secured through using such methods as those of interchange, gradation, and transition, which, nevertheless, cause all the divergent parts of a composition to assimilate. (See Appendix, page 387.) Because, too, all the methods in the chart on page 277 are, more or less, connected, music, at times, reveals traces of the influence of every one of these.

not so.

Some may suppose that, in poetry, there are no effects corresponding to those of musical harmony. But this is Inasmuch as poetry uses words, the articulation of these renders them more clearly distinguishable from one another than are musical notes; and there is not the same necessity, as in the latter, for merely tonal distinctions of quality and pitch. But science has ascertained that in addition to the pitch on which a vowel or consonant is apparently sounded, it has, at least, one partial tone peculiar to itself, which tone is always at the same pitch. For this reason, alliteration, assonance, and rhyme all involve the use of like pitch; consecutive syllables produce different consecutive degrees of pitch, i. e., melodies, or what are termed tunes of verse; and every syllable containing a vowel and a consonant, like an, for instance, contains two tones that may or may not har

monise. For these reasons, the words of poetry, though in a very subtle, but, at the same time, suggestive way, fulfil the same methods as those of musical harmony. See the author's "Rhythm and Harmony in Poetry and Music," Chapters V. to XII.

CHAPTER XVIII.

HARMONY OF COLOUR IN THE ARTS OF SIGHT.

Production of the Colours of the Spectrum-Effect of Light upon ColoursDefinition of Terms-Complementary Colours-As Produced by Light and by Pigments-The After-Image in Consecutive Contrast-Simultaneous Contrast-All Colours Impart about them Tints of their Complementaries-Principles Determining Use together of Two ColoursOf Three Colours-Of Four Colours-Consecutive and Simultaneous Contrast Due to Physiological Action of the Eye-Correspondences between Ratios of Harmonic Colours and Tones-Owing to Minuteness of Colour-Waves Nothing in Colours Corresponds to the Different Scales in Music-The Ratios of the Two Notes of a Single Musical Scale Forming the Most Perfect Consonance—This Ratio as Represented among the Colours-Colour Harmony as Actually Developed-Not from Ratios Occasioning Vibrations, but from Analysis of LightThe Field-Theory of Colour-Harmony-Theory Based on Psychological Effects-On Physiological Effects-Tone, or the Predominant Use of One Colour in a Painting-Why this May Fulfil the Same Principle of Harmony as the Use of Great Variety of Colour-Colour Harmony Results from an Application to Colour of All the Principles Unfolded in Chapters XIV. and XV.—Beauty in Art-And Suggestion-Conclusion.

T is now more than two centuries since Newton, analysing the rays of the sun, detected that all the different colours, except, perhaps, extreme purple are contained in light. Most of us know how to reproduce his analysis. By means of a mirror, the sun's rays are reflected in a small band through a narrow opening in a window-shade or blind, and sent into an otherwise darkened room. When they enter this room, they are made to pass through a glass prism. The prism turns the band of rays aside from its direction, and, at the same time, separates

it into many bands of rays which are coloured, and each of of which, after leaving the prism, continues in a straight line. If these bands fall on a white wall or screen, each produces a different colour, and all together a series of colours in which we recognise all that are in the rainbow. Nearest where the white would have fallen, if the prism had not intervened, we find red, and next to this the other colours in this order: orange, yellow, green, blue, indigo, and violet. This series of colours is called the Spectrum.

If, aside from the Spectrum, we test the effects of different degrees of light upon colours, we find that in a darkened room, blue appears to be dark blue, but as we gradually increase the light it becomes first blue, then light blue, then pale blue, then, in light of great intensity, loses its blueness almost entirely, becoming very nearly white. So, too, if in place of different degrees of light, we use black or white pigments, mixing them with coloured pigments, we find the colours becoming respectively darker or lighter.

When

The different kinds of colours are termed hues. hues are in the state in which they appear in the spectrum, they are called full or high colours. If darker than in the spectrum, the colours are termed dark, if lighter, light; if very much lighter, pale, or, what means the same thing, broken. When full colours are made darker, their different degrees of darkness are termed shades. When they are made lighter their different degrees of lightness are termed tints. The degree of colouring or of dark or light in a shade or tint determines the tone, as when we speak of a golden and gay, or a gray and sombre tone. Paintings, however, are not generally said to be distinguished by tone except when producing the effect

described on page 381. In a positive colour the tint or shade of a single hue is prominent; in a neutral colour, there is so much of a mixture that there is no predominating hue. The warm are the reds, browns, oranges, yellows, and associated colours; the cold are the greens, blues, violets, purples, and associated colours. Primary is a term formerly applied to red, yellow, and blue, becaused they were supposed to be primitives from which, when mixing pigments, the secondary colours, orange, green, and violet were derived, orange by mixing red and yellow, green by mixing yellow and blue, and violet by mixing blue and red. For reasons to be given by and by, however, these distinctions between primary and secondary are not now considered tenable.

Let us return to the spectrum. If all the colours together make white, it follows that the absence from white light of any of its constituent elements must produce a colour. This logical inference has been confirmed by the following among other experiments. Between the prism and the spectrum cast by it, according to the explanations given on page 367, a lens bounded by cylindrical surfaces is introduced. This lens is so constructed that it reunites the prismatic bundle of rays into a single band, i. e.,it restores these rays to the same condition in which they were before they reached the prism from the slit in the window. This cylindrical lens now gathers the rays together, and casts upon the wall, where the spectrum was before, merely a small white image of the slit in the window, giving thus a proof, in addition to the others just noticed, that all the colours together make white. If now between the cylindrical lens and the wall a part of the light be shut off by means of a screen, a coloured image instantly appears upon the wall. If, for