Matter has a decomposing and an elective power with regard to both radiant light and heat; most coloured bodies, such as flowers, green leaves, dyed cloth, &c., though seen by reflection, owe their colour to absorption. The light by which they are seen is reflected, but it is not in reflection that the selection of the rays is made which causes the objects to appear coloured. When light falls upon red cloth, a small portion is reflected at the outer surfaces of the fibres, and this portion, if it could be observed alone, would be found to be colourless. The greater portion of the light penetrates into the fibres, when it immediately begins to suffer absorption on the part of the colouring matter. On arriving at the second surface of the fibre, a portion is reflected and a portion passes on, to be afterwards reflected from, or absorbed by, fibres lying more deeply. At each reflection the various kinds of light are reflected in as nearly as possible the same proportion, but in passing across the fibres while going and returning they suffer very unequal absorption on the part of the colouring matter; so that in the aggregation of the light perceived the different components of white light are present in proportions widely different from those they bear to each other in white light itself, and the result is a vivid colouring. In certain substances however, as gold and copper, the different components of white light are reflected with different degrees of intensity, and the light becomes coloured by these reflections. Gold is yellow by reflection; red cloth is red by absorption. In the same sense, physically speaking, in which the red cloth is red, gold is not yellow but blue or green; such is in fact the colour of gold by transmission through gold leaf, and therefore gold is greenish blue by absorption. In this case we see that while the substance copiously reflects and intensely absorbs rays of all kinds, it more copiously reflects the less refrangible rays with respect to which it is more intensely opaque. In general absorption and radiation are independent of colour. There is a vast diversity in the property which substances possess with regard to the transmission of radiant light and heat; glass, for instance, transmits light abundantly, but is impervious to heat from non-luminous sources; while other substances, which are altogether opaque to light, transmit heat copiously, as the bisulphide of carbon, which of all liquids is the most diathermic, while water in all its forms is almost impervious to heat.