You can see roughly the ~rate~ at which the water goes off from the leaves by completely filling with water an apparatus like that in fig. 19. As the leafy branch (which is firmly fastened in the cork with no air leakage) uses up the water, it must be drawn along the narrow tube, which is graduated so as to show the quantity lost. [Illustration: Fig. 19. Experiment to measure the amount of water given off by leaves in a given time. At first the tube is full of water, which is drawn back to points 1, 2, etc., as the leaves use it.] From these experiments we find that even although we do not actually see it coming off, yet ~the leaves of the plant give off a great deal of water in the form of vapour~. By this process large quantities of water are drawn through the plant, and the salts in weak solution in it are kept and used by the plant as they are needed for building up its structure. Now you may think that the loss is simply the result of evaporation from the leaves, because the surface of the leaves is great, and they would therefore naturally lose a considerable amount of water by evaporation. But this view is only partly correct, because the giving off of water by leaves or “transpiration,” as it is called, is regulated by a number of little pores in the skin of the leaf, which can open and close. You can see the importance of these pores as water regulators in plants which have them only on one side of the leaf, because practically all the water escapes from the side on which they are situated. [Illustration: Fig. 20. Leaf A greased on the ~lower~ side, leaf B on the ~upper~ side, and C not at all. B withers as fast as C.]