OCR
52 LIGHT AND WATER strokes on their further sides." The characteristic of moving water is indeed to insist upon the vertical and ignore the horizontal lines of the image, though very regular ripples in the foreground often seem to contradict this principle by showing horizontal lines that do not exist in the object at all. The reflexion of the sunset sky in Plate X XI affords a striking instance of an apparent exception of this kind, the cloud forms being broken up horizontally by the extremely gentle motion of the water. Plate XXII, in which the ripples are fairly regular and travelling obliquely, shows the characteristic “ markings " formed on their farther sides by the reflexion of the dark wooded bank. But it will, no doubt, have been noticed that these horizontal strokes often take the form of elongated rings. In Plate XXIV the reflexion of the distant and nearly horizontal bank is seen to consist of a number of very flat rings. When the spectator’s position is not too low, so that the whole surface of a wave is in view, as when looking from Q, Fig. 14, there are generally ¢wo points on each wave at which a suitably placed luminous point, such as P, can be seen reflected. Thus, on the farther side of the first wave, P appears by reflexion at I and H, and on the second wave,at G and F. Beyond X, the two points are on the near side of each wave, viz., D and C on the fourth, and B and A on the fifth wave.2 Now if, in ‘ Compare also the rippled surfaces in Plates XII and XIII, page 44, and in Plates XVII and XVIII, page 50. " Ina sine curve (page 31) the two points of each pair would be proportionally farther apart than in the circular curve of the diagram.