In a single bubble, suffered to hang from the straw or tube, the appearance of colour is difficult to observe, for the quick rotation prevents any accurate observation, and all the colours seem to mix together; yet we can perceive that the colours begin at the orifice of the tube. The solution itself may, however, be blown into carefully, so that only one bubble shall appear. This remains white (colourless) if not much agitated; but if the solution is not too watery, circles appear round the perpendicular axis of the bubble; these being near each other, are commonly composed alternately of green and red. Lastly, several bubbles may be produced together by the same means; in this case the colours appear on the sides where two bubbles have pressed each other flat. 465. The bubbles of chocolate-froth may perhaps be even more conveniently observed than those of soap; though smaller, they remain longer. In these, owing to the heat, an impulse, a movement, is produced and sustained, which appears necessary to the development and succession of the appearances. 466. If the bubble is small, or shut in between others, coloured lines chase each other over the surface, resembling marbled paper; all the colours of the scale are seen to pass through each other; the pure, the augmented, the combined, all distinctly clear and beautiful. In small bubbles the appearance lasts for a considerable time. 467. If the bubble is larger, or if it becomes by degrees detached, owing to the bursting of others near, we perceive that this impulsion and attraction of the colours has, as it were, an end in view; for on the highest point of the bubble we see a small circle appear, which is yellow in the centre; the other remaining coloured lines move constantly round this with a vermicular action. 468. In a short time the circle enlarges and sinks downwards on all sides; in the centre the yellow remains; below and on the outside it becomes red, and soon blue; below this again appears a new circle of the same series of colours: if they approximate sufficiently, a green is produced by the union of the border-colours. 469. When I could count three such leading circles, the centre was colourless, and this space became by degrees larger as the circles sank lower, till at last the bubble burst. 470.