Visual Sensation

These two systems, the chromatic and the achromatic, can obvi ously be combined to form a three-dimensional figure which em braces, simultaneously, the properties of brilliance, saturation and hue. The re sulting scheme, as diagrammed in fig. r, provides locations for all conceivable col ours, and thus enables us to specify any colour numerically in terms of three de terminants or attributes. Chromatic col ours exist, or are conceivable, at all levels of brilliance, so that the figure becomes a psychological colour solid. The exact boundaries of this solid cannot be speci fied at the present time, although we may safely affirm a tendency for the number of possible chromatic colours to show a maximum in the middle range of brilliance, and to be reduced practically to zero at the ideal black and white.

The Psychophysiology of Colour.

Experimental studies of visual sensation have been concerned, for the most part, with the laws which connect brilliance, saturation and hue with features of the stimulus and sense organ process. The electromagnetic waves which constitute the stimulus impinge upon the corneas of the eyes, pass through the pupils, and are refracted by the various ocular media so as to form images upon the retinas. The ele mentary properties of the stimulus, from the physical standpoint, are its intensity and its relative wave-length composition. The latter can be subdivided into two features, such as dominant and purity. The simplest psychophysiological rela tions are found when homogeneous, or single wave, stimuli are employed, and, in this case, the stimulus is specified by its wave length and intensity.

It is found that all three attributes of colour depend upon each possible aspect of the stimulus. Beginning with the relationships of brilliance, we note, firstly, that at constant intensity this attribute varies with wave-length, in accordance with a function which is approximately of the probability integral type. This symmetrical, single maximum curve corresponds in general form with the so-called visibility function, although, strictly speaking, the latter represents the reciprocal intensities which are required at different wave-lengths, to yield equal brilliances, rather than delineating a direct psychophysical relationship. The visibility curve gives values by which radiometric intensities must be multi plied in order to secure the corresponding photometric or light intensities.

The position of the maximum of the function which connects brilliance with wave-length varies with the stimulus intensity and with the state of adaptation of the eye. At intensities corre sponding to daylight, the maximum lies at about 554 mil, but under twilight illumination it shifts to the neighbourhood of 5 I I mkt. The exact limits of the curve, which determine the ex

tremes of the visible spectrum, naturally vary with the intensity level, but may be taken as 400 mp, and 76o naµ for daylight con ditions, shifting with the maximum at lower intensities. The high intensity curve is characteristic of pliotopic (or cone) vision, while the low intensity one features scotopic (or rod) vision. The transition from the former to the latter is responsible for the Purkinje and other similar phenomena.

When a stimulus, of fixed wave-length composition, is varied in intensity, the brilliance is found to be a logarithmic function of the latter in the range of ordinary daylight intensities. At both higher and lower intensities, the brilliance changes less rapidly with respect to the intensity than is characteristic of the loga rithmic section of the law. The total law may be designated as the Fechner function. In the logarithmic region, the just notice able increment of intensity is approximately one one-hundredth part, depending upon conditions. The Fechner function is sub stantially independent of wave-length composition, if intensities are expressed in photometric terms, i.e., if the physical intensi ties are multiplied by the corresponding visibility values. This is true for either homogeneous or heterogeneous stimuli, and in the latter case implies the principle that the light values of the components in a mixture add arithmetically.

Hue is determined primarily by the wave-length of the stimu lus, when the latter is homogeneous, and by a quantitative balance between component wave-lengths and intensities, when it is hetero geneous. The hue changes progressively from one end of the visible spectrum to the other, although at a variable rate. Maxi mal rates of change of hue with respect to wave-length are found at 589 mil, 507 mp, and 489.5 m,u, at each of which points a differ ence of i mkt is just noticeable. At the long-wave (red) end of the spectrum, there is no change of hue beyond about 700 mkt and there is only one just noticeable step between 700 mi.c and 678 mil. The psychologically primary hues, within the spectrum, are the yellow at 574.5 mil, the green at 505.5 mil and the blue at mil, the primary red requiring the admixture of a small quantity of short-wave radiation to the spectral rays between 700 mµ and the long-wave end. Saturation also varies with the wave-length of a homogeneous stimulus. It shows maxima at either end of the spectrum, with a minimum at about 575 mp..