If we suppose a patch of two shades of red colour to be laid on a smooth surface, and alongside of this a patch of two shades of yellow colour, and if we suppose the whole mass of colour to be viewed from a distance such that one patch appears uniformly red and the other uniformly yellow, we shall have a rough illustration of the Berzelian compound atom. To the observer the whole mass of colour appears to consist of two distinct patches of contrasted colours; but let him approach nearer, and he perceives that what appeared to be a uniform surface of red or yellow really consists of two patches of unlike shades of red or of yellow. The whole mass of colour represents the compound atom; broadly it consists of two parts--the red colour represents one of the constituent atoms, the yellow colour represents the other constituent atom; but on closer examination the red atom, so to speak--and likewise the yellow atom--is found to consist of parts which are less unlike each other than the whole red atom is unlike the whole yellow atom.
We shall have to consider in more detail the reasoning whereby Berzelius arrived at this conception of every compound atom as a _dual_ structure (see pp. 209-212). At present I wish to notice this conception as lying at the root of most of the work which he did in extending and applying the Daltonian theory. I wish to insist on the fact that the atomic theory could not advance without methods being found for determining the number of elementary atoms in a compound atom, without clear conceptions being gained of every compound atom as a structure, and without at least attempts being made to learn the laws in accordance with which that structure was built. Before the atomic weight of oxygen could be determined it was necessary that the number of oxygen and of hydrogen atoms in the atom of water should be known; otherwise all that could be stated was, the atomic weight of oxygen is a simple multiple of 8. Berzelius did much to advance chemical science by the introduction and application of a few simple rules whereby he determined the number of elementary atoms in various compound atoms. But as the science advanced, and as more facts came to be known, the Berzelian rules were found to be too narrow and too arbitrary; chemists sought for some surer and more generally applicable method than that which Berzelius had introduced, and the imperious demand for this method at last forced them to recognize the importance of the great generalization of the Italian naturalist Avogadro, which they had possessed since the year 1811, but the meaning of which they had so long failed to understand.