SPECIFIC COLOUR.
It has sometimes been assumed that colour increase was in direct ratio to intensity increase, but this is never the case, each substance has its own rate, specific to itself. It is conceivable that the colours of two substances may coincide at one point, but as their densities increase, or decrease, their rates of change vary.
The term “Specific Colour” is based on the experimental fact, that the colour of a given substance is constant, so long as the substance itself and the conditions of observation, remain unaltered. During experimental work a sufficient number of instances have accumulated to warrant the writer in advancing and using the term “Specific Colour” as describing a new natural law, as rigid in its application as that of “Specific Gravity” or “Specific Heat.”
PLATE V
ABSORPTION CURVES OF FIVE COLOUR CONSTANTS.
To face page 33.[Lovibond, Colour Theories.
When this principle is applied to the measurement of regularly increasing thicknesses, curves of colour changes can be established, which are specific for the substance in question, and afford a certain means of identifying similar substances in future. This is effected by varying the nature of the co-ordinates, making the ordinates to represent the tintometrical scale of colour units irrespective of colour, whilst the abscissae represent the scale of increasing thicknesses. Then by plotting the separate factors of each measurement according to their unit values, a series of curves is established, specific to the substance in question, and applicable to none other.
We have now two systems of charting colour, in the first, the complete sensation is represented by a single point, as in Plate IV. In the second, each factor is represented by a separate point, and by connecting points of similar colours, a series of curves is established which represents a quantitative analysis of the progressive colour development, as in Plate V.
CHAPTER X.
Representations of Colour in Space of Three Dimensions.
The relations of the different colours to one another, and to neutral tint are, perhaps, best represented to the mind by a solid model, or by reference to three co-ordinate axes, as employed in solid geometry (see [Fig. 2]).
Fig. 2.
Let the three adjacent edges OR, OB, OY, of the above cube be three axes, along which are measured degrees of Red, Yellow and Blue respectively, starting from the origin O. Every point in space on the positive side of this origin will then represent a conceivable colour, the constituents of which in degrees of red, yellow and blue are measured by the three co-ordinates of the points. Pure reds lie all along the axis OR, pure yellows on the axis OY, and pure blues on the axis OB.
All normal oranges, normal greens, and normal violets lie on the diagonals of the faces of the cubes OO1, OG, OV respectively.
Pure neutral tints lie on the diagonal ON of the cube, equally inclined to the three principal axes.
Red violets will be found on the plane ROB, between OV and OR.
Blue violets on the same plane between OV and OB.
“Saddened” red violets all within the wedge or open space enclosed by the three planes, whose boundaries are OB, OV, ON.
The other colours, red and yellow oranges, blue and yellow greens, pure and saddened, are found in corresponding positions in relation to the other cases.[4]
CHAPTER XI.
The Spectrum in relation to Colour Standardization.
The spectrum has naturally been considered as a suitable source for colour standards, but the power of analysing has disclosed some difficulties, which have yet to be overcome.
Concerning the prismatic spectrum, there has always been a difficulty in apportioning the different colours to specific areas, and further, before this spectrum is available for colour standardization, some method of correction for the unequal distribution of colours must be devised.
Neither of these difficulties occur in the use of the diffraction spectrum, where the pure colours are apportioned by Professor Rood from A to H in the manner shown in table on next page.
Professor Rood further divides the spectrum from A to H into 100 equal divisions, allotting 20 unit divisions of 72,716 wave lengths to the space between each two colour lines. This allots a space of 3,635 W.L. to each unit division, as shown in Table III.
TABLE III.
| Wave Length Position. | No. of Wave Lengths from Colour between each. | Division | W.L.K. per Division | ||
| 760,400 A. | Red | 760,400 —— | 72,717 | == 20 | 3,635 |
| Orange | 687,683 —— | 72,716 | == 20 | 3,635 | |
| Yellow | 614,967 —— | 72,716 | == 20 | 3,635 | |
| Green | 542,251 —— | 72,716 | == 20 | 3,635 | |
| Blue | 469,535 —— | 72,716 | == 20 | 3,635 | |
| 396,819 H. | Violet | 396,819 | |||
| 363,581 | Total W.L. between A. & H. | 363,581 | 100 | ||
Having provided equal wave length positions for the six pure colours, the intermediate colours are necessarily binaries in definite proportions, accounted for by a regular overlapping of two bounding colours in opposite directions from zero to 20, as shown in the following table from Red to Orange, representing the space between these two pure colours.
| Red W.L 760,400 | 20 | 19 | 18 | 17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 | W.L 687,683 Orange. |
| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | ||
| 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 | 20 |
It follows, that apart from the six monochromes, all spectrum complex colours in a single wave length must be binaries, whose united values equal 20.
On comparing Professor Rood’s scales of divisions with those of the tintometrical scales already described, they appear to coincide in several particulars, for instance:—
The monochromes correspond both in number and in name.
Their positions in the scales correspond.
Their unit divisions are equal in number, and in dimensions.
Their colour positions correspond, when an artificial tintometrical spectrum is made by regularly overlapping monochromes.
It follows that when the two scales are superimposed as in Plate V., showing similar monochromes as lying in the same perpendicular, the same wave length numbers apply to both; concerning the dimensions between the monochromes, the spaces occupied by 72,716 wave lengths between the spectrum monochromes, also represent similar spaces in the tintometrical scales, and one-twentieth of this 3,635 represents the space of a single unit in each case.
In connexion with these co-related dimensions, some information is obtainable bearing on the limitation of a monochromatic vision for discriminating small colour differences. Under ordinary daylight conditions, the unit in the lighter shades of the tintometrical scale is divided into 100 fractional parts, each fraction therefore represents a space occupied by thirty-six wave lengths in the spectrum scale. This may be near the limit of dimension for monochromatic vision in such a gradually changing colour scale, as that of the spectrum, and may be some guide as to suitable slit areas in the synthetical building up of complex coloured light.
PLATE VI
SIX COLOUR CHARTS IN ONE OR ANOTHER OF WHICH ANY SIMPLE OR COMPLEX COLOUR FINDS A DEFINITE POSITION.
To face page 39.[Lovibond, Colour Theories.
In Plate VI. are shown the six tintometrical colour charts, as lying in their order on the tintometrical spectrum, illustrating that any measured colour factor lies in a perpendicular drawn through both spectra, and occupying the same wave length position, and may therefore be designated by that wave length number.
This explanation is not intended to convey that the colour energies do not really overlap beyond the boundaries of the dual combinations, but only that the vision is unable to distinguish as colour, such overlapping if it exists.