quences are actually realized in natural phenomena, and such satisfaction we can have in the present case. Consider what must be the form which a mass of liquid molecules isolated in space would necessarily take. Remember that these molecules are moving with perfect freedom within the body, but that the extent of the motion of each molecule is limited by the attraction of the mass of the liquid. Remember also that, according to the well-known principles of mechanics, this attraction may be regarded as proceeding from a single point, called the centre of gravity. Remember, further, that the molecules have all the same moving power, and you will see that the extreme limits of their excursions to and fro through the liquid mass must be on all sides at the same distance from the central point. Hence the bounding surface will be that whose points are all equally distant from the centre. I need not tell you that such a surface is a sphere, nor that a mass of liquid in space always assumes a spherical form. The rain-drops have taught every one this truth. Still, a less familiar illustration may help to enforce it. I have therefore prepared a mixture of alcohol-and-water, of the same specific gravity as olive-oil, and in it I have suspended a few drops of the oil. By placing the liquid in a cell, between parallel plates of glass, I can readily project an image of the drops on the screen, and I wish you to notice how perfectly spherical they are. And I would have you, moreover, by the aid of your imagination, look within this external form, and picture to yourselves the molecules of oil moving to and fro through the drops, but always slackening their motion where they approach the surface, and on every side coming to rest and turning back at the same distance from the centre of motion.

MOLECULAR STRUCTURE OF SOLIDS.

51

Neither liquids nor gases present the least trace of structure. They cannot even support their own weight, much less sustain any longitudinal or shearing stress. A solid, on the other hand, has both tenacity and structure, and resists, with greater or less energy, any force tending to alter its form, as well as change its volume. The tenacity and peculiar forms of elasticity which solids exhibit are characteristics which are familiar to every one, but the evidences of structure are not so conspicuous. The structure of solids is most frequently manifested by their crystalline form, and this form is one of the most marked features of the solid state. But although, under definite conditions, most substances assume a fixed geometrical form, yet, to ordinary experience, these forms are the exceptions, and not the rule. I will therefore make the crystallization of solid bodies the subject of a few experimental illustrations.

For the first experiment, I have prepared a concentrated solution of ammonic chloride (sal-ammoniac), and with this I will now smear the surface of a small glass plate. Placing this before our lantern, and using a lens of short focus, so as to form a greatly-enlarged image on the screen, let us watch the separation of the solid salt as the solution evaporates. . . . Notice that, first, small particles appear, and then from these nuclei the crystals shoot out and ramify in all directions, soon covering the plate with a beautiful net-work of the filaments of the salt. We cannot here, it is true, distinguish any definite geometrical form; but it can be shown that these very filaments are aggregates of such forms, and their structure is made evident by a fact, to which I would especially call your attention—that, as the crystalline shoots ramify over the plate, the sprays keep always at right angles to the stem, or else branch

at an angle of 45°, which is the half of a right angle (Fig. 8).

For a further illustration of the process of crystallization I have prepared a solution in alcohol of a solid

substance called urea, with which we will experiment in precisely the same way as before. . . . The process of crystallization, which is here so beautifully exhibited, is one of the most striking phenomena in the whole range of experimental science. It is, of course, not so wonderful as the development of a plant or an animal from its germ, but then organic growth is slow and gradual, while here beautiful, symmetrical forms shape themselves in an instant out of this liquid mass, revealing to us an architectural power in what we call lifeless matter, whose existence and controlling influence but few of us have probably realized. The general order of the phenomena in this experiment is the same as in the last; but notice how different the details. We do not see here that tendency to ramify at a definite angle, but the crystals shoot out in straight lines, and cover the plate with bundles of crystalline fibres, which meet or in

CRYSTALLINE STRUCTURE OF ICE.

53

tersect each other irregularly as the accidental directions of the several shoots may determine (Fig. 9). As before, we cannot recognize the separate crystals; indeed, large isolated crystals, such as you may see in collections of minerals, cannot be formed thus rapidly. They are of slow growth, and only found where the conditions have fa vored their development. But all the mineral substances, of which the rocks of our globe consist, have a crystalline structure, and are aggregates of minute crystals like the arborescent forms whose growth you have witnessed.

The external form is but one of the indications of crystalline structure, and by various means this structure may frequently be made manifest when the body appears wholly amorphous. Nothing could appear externally more devoid of structure than a block of transparent ice. Yet it has a most beautiful symmetrical structure, which can easily be made evident by a very simple experiment, originally devised, I believe, by Prof. Tyndall. For this purpose I have prepared a plate of ice about an inch in thickness, whose polished surfaces are parallel to the original plane of freezing. I will now place this plate in front of the condenser of my lantern, and, placing before it a lens, we will form on the curtain an image of the ice-plate, some twenty times as large as the plate itself. The rays of heat which accompany the light-rays of our lantern soon begin to melt the ice; but, in melting it, they also dissect it, and reveal its structure. . . . Notice those symmetrical six-pointed stars which are appearing on the wall (Fig. 10). Prof. Tyndall calls them, very appropriately, ice-flowers, for, as the flower shows forth the structure of the plant, so these hexagonal forms disclose the six-sided structure of ice. You can hardly fail to notice the similarity of these forms to those of the snow flake. The six petals

of the ice-flowers on our screen make with each other an angle of 60°, and, if you examine, with a magnifier, flakes of fresh-fallen snow (Fig. 11), or the arborescent

FIG. 10.-Ice-Flowers.

forms which crystallize on the window-panes in frosty weather, you will find that, in all cases, the crystalline shoots ramify at this angle, which is as constant a character of the solid condition of water as is the right angle of sal-ammoniac.

There are other solids whose crystalline structure, like that of ice, becomes evident during melting; but a far more efficient means of discovering the structure of solids, when transparent, is furnished by polarized light.

It would be impossible for me, without devoting a great deal of time to the subject, either to explain the nature of what the physicists call polarized light, or to give any clear idea of the manner in which it brings out the structure of the solid. I can only show you a few experiments, which will make evident that such is the fact. We have now thrown on the screen a luminous disk, which is illuminated by polarized light. To the unaided eye it does not appear differently from