drilles of rhombs, often with the most charming effect. By the direction of these lines the observer divines the course of the current, and their inclination furnishes him with an exact measure of its velocity.
In the case of cascades, the greater variety of the phenomena forces us to go more into detail. We begin with an inclined vessel, glass, pot, or pail, in process of emptying. How does the surface
of the outrunning liquid look? It would be safe to wager ten to one that any person at the first instance would represent it as in Fig. 5, v, by parallel parabolic lines. We not rarely find artistic productions in this style dating from the time when they composed landscapes from fancy in their studios. Without being too severe on these errors, which are still not far away from us, we will try to do better, and to correct the faults of the figure, one at a time.
We lay aside, for the moment, the usual ribbon-like form, which is false, and examine first the question of the vertical lines. They are formed by the series, A, a, a, of points by which the same molecule of water passes (or is supposed to pass) successively. Is there any proof of their material existence? No. It must be admitted that in certain cases, like that of a thin sheet falling from