The regenerative power of the organism, by means of which it may replace a lost part, or by means of which a piece may become a new whole, is also something not present in machines. In using a machine for comparison we should not leave out of sight the fact that machines are themselves the work of organisms, and have been made for some purpose useful to the organism. They may perform the same purpose for which we would use our own hands, for they differ from parts of the body mainly in that they are made of different compounds having different properties, as the above comparisons have shown. But the regulations of the machine have been added to it by man on account of their usefulness to himself, and are not properties of the material of which the machine itself is composed. This shows, I think, the inappropriateness of making any comparison between these two entirely different things. If, then, we find the comparison between machines and organisms unprofitable, can we find any other things in inorganic nature that can be better compared with the phenomenon of adaptation of the organism? The following phenomena have been made the subject of comparison from time to time. The bendings, which are gradually made by rivers often lead to a meeting of the loops, so that a direct, new communication is established, and the course of the river is straightened out. The water takes, therefore, a more direct course to the sea. It cannot be said, however, to be of any advantage to the river to straighten its course. Again, a glacier moulds itself to its bed, and gradually moves around obstacles to a lower level, but this adaptation of the glacier to the form of its surroundings cannot be said to be of advantage to the glacier. On the contrary, the glacier reaches so much the sooner a lower level where it is melted. The unusual case of a solid being lighter than the liquid from which it forms, as seen in the case of ice, has been looked upon as a useful arrangement, since were the reverse the case all rivers and ponds would become solid in winter in cold climates, and the polar regions would become one solid block of ice. But no one will suppose for a moment that there is any relation between the anomalous condition of the lightness of ice, and its relation to the winter freezing of streams, ponds, etc. It has even been suggested that this property of ice was given to it in order that the animals living in the water might not be killed, which would be the case if the ice sank to the bottom, but such a method of interpreting physical phenomena would scarcely commend itself to a physicist. The formation of a covering of oxide over the surface of a piece of iron delays the further process of oxidation, but who will imagine that this property of iron has been acquired in order to prevent the iron from being destroyed by oxygen?