Several arbitrary scales for specific purposes had already been constructed by selecting a colour depth which could easily be distinguished, calling it a unit, and scaling it by duplicating and subdividing. This course was adopted with a coloured glass which approximately matched Ales and Malt solutions, and another which matched Nesslerized Ammonia solutions. No insuperable difficulty occurred in constructing scales available for quantitative work in these two instances. The intensity of the colour unit for these arbitrary scales, was that which appeared to be most convenient for the purpose required, but the several scales had no common basis. The unit was physiological, and the exactitude of the scales depended entirely on the skill of the vision for discriminating small differences. As the writer’s experimental work progressed, it became evident that red, yellow, and blue were the only colours suitable for systematic work. The superimposition of any two, developed a third colour which apparently had no relation to either. The superimposition of the third glass modified or destroyed all colour and reduced the amount of light. This suggested the idea that if the three colours could be so balanced that the light transmitted was colourless, it would be evidence of equivalence of intensity in the individual colours. The real difficulty was in obtaining this equivalence, because a balance which transmitted a neutral tint by one light developed colour by another. This necessitated the selection of a standard light. The light finally selected was that of a so-called sea fog, away from the contaminating influence of towns. The white fog of Salisbury Plain was used as being most available. It required two years’ work to establish equivalence in the unit. [Illustration: PLATE II NINE CIRCLES ILLUSTRATING THE ANALYSIS OF A BEAM OF WHITE LIGHT INTO THE SIX COMPOSING COLOURS BY THE ABSORPTIVE METHOD A. WHITE LIGHT B. DIVIDED INTO CHROMATIC EQUIVALENTS C. WHOLLY ABSORBED COLOURS OF THE FIRST DIVISION DEVELOPED BY THE ABSORPTION OF THE THREE COMPLIMENTARIES 1 2 3 COLOURS OF THE SECOND DIVISION DEVELOPED BY THE ABSORPTION OF THE FIVE OTHER RAYS 4 5 6 _To face page 11._] [_Lovibond, Colour Theories._ ] CHAPTER IV. Derivation of Colour from White Light. The method of analysing white light into its colour constituents by means of coloured glass absorbents of known intensity and purity, is illustrated by the set of nine circles in Plate II, which demonstrate that colour is developed by the absorption of the complementary colour rays. The ratios of transmission are equal. In this set of illustrations the circles represent light of 20 units luminous intensity, and the absorptive value of the three glass colours is each of 20 units, therefore the whole of the light and colour energies are presumed to be dealt with.