1892.] on the Surface-Film of Water, &c. 547 



that the bits of cork are attracted one to another and crowd together 

 in one place. Let us inquire why the floating bits of cork are thus 

 attracted towards one another. If any solid capable of being wetted 

 by water is partly immersed in water, the liquid rises round it in an 

 ascending capillary curve. If the solid is not wetted by water, the 

 curve will turn downwards. We may get ascending or descending 

 capillary curves in other ways. If, for instance, I were to lay a 

 sheet of paper upon water, and turn its edges up at certain places, we 

 should get marked ascending curves at these points. The raising of 

 some parts of the surface causes other parts to sink, and may bring 

 about descending curves, or make previously formed descending 

 curves more marked. We shall find it helpful in our experiments 

 to notice one very simple plan of producing a descending capillary 

 curve round the edge of a vessel. If we take a glass of water, and 

 fill it until the water is level with the brim, we naturally 

 speak of the glass as full; but if we are careful to avoid rude 

 shaking, we may still add a considerable quantity of water without 

 spilling any. The glass will then become what we may call over- 

 full, and its surface will be bounded by a descending capillary 

 curve. Now, it is of immediate importance to us to observe that 

 lihe capillary curves, whether ascending or descending, attract one 

 another, and that unlike curves repel one another. The theoretical 

 explanation of this point is not difBcult, but it must not detain us 

 here. To place the fact itself beyond dispute, we will try a little 

 experiment. A circular dish of water is now placed in the field of 

 the lantern, and we will introduce into it a small disc of wood. 

 Both the disc and the side of the vessel are wetted by water, and 

 an ascending capillary curve rises round each. The result is 

 that the two bodies attract one another. Every time the disc is 

 moved away it is powerfully drawn towards the side of the vessel. 

 With a little syringe we will add water to the dish in sufficient 

 quantity to raise the level above the edge of the vessel. You will 

 observe that the wooden disc is now repelled by the edge of the 

 vessel, and floats free in the centre. By sucking up a little water, it 

 becomes attracted once more, and so we may go on, causing it to be 

 attracted or repelled, according as we add or subtract a small quantity 

 of water. But what has all this to do.with the duckweed ? In order 

 to explain the behaviour of duckweed, I must ask you to examine a 

 careful representation of its form. This common plant has not, to 

 my knowledge, been faithfully represented in any botanical book. 

 You will see that the leaf is of an irregular oval shape, broader at 

 one end than at the other, aud that the narrow end is pointed. A 

 raised ridge extends along the length of the leaf, from the point to 

 the middle of the opposite or rounded border. Duckweed almost 

 invariably propagates itself by budding. New leaves are pushed out 

 symmetrically on each side of the point. They grow bigger and 

 bigger, and gradually free themselves. The point upon each leaf 

 marks the place where it was last attached to the parent leaf. Some- 



