Mitscherlich has remarked that crystals do not expand uniformly by heat, but that this dilatation is greater in one direction than in another; and that this difference is connected with their crystalline form. M. de Sénarmont has shown that conductibility for heat, which is equal in all directions for the crystals of the regular system, acquires in others a maximum or a minimum value, according to directions parallel to the crystallographic axes; so that the isothermic surfaces, which are spheres in the former case, are, in the other, ellipsoids elongated or flattened in the same direction. The optical axes do not altogether coincide with the principal axes of conductibility for heat; but this appears to be due merely to slight differences in the rate of progression, or the refrangibility of the luminous and calorific rays. Wiedemann, by employing a fine point through which he made electricity arrive upon a surface that he had powdered with licopodium or red lead, succeeded in determining, by means of the form assumed by this light powder, the conductibility of crystals in different directions. On a surface of glass, the powder which disperses itself around the points, in consequence of electric repulsion, forms a circular figure traversed by radii. When a plate of gypsum is used instead of glass the figure is found to be elliptical, and the great axis of the ellipse forms a right angle with the principal crystallographic axis, which proves that the electricity distributes itself more easily in a direction perpendicular to the axis than in any other. M. Wiedemann comes to the conclusion that crystals which possess a better conductibility in the direction of the principal axis, all belong to the class of negative crystals: while those which have a better conductibility in the direction perpendicular to the axis are positive, which indicates that the direction of best conductibility for electricity is also that according to which light is propagated relatively with greater velocity. Tyndale has shown, that if gutta percha which has been rendered fibrous in manufacture is cut so that the fibres are in the direction of this greatest length, or in a direction perpendicular to this greatest length, and placed under the influence of a magnet, they direct themselves equatorially. Ivory cut in the same direction manifests the same conditions, though both these substances are diamagnetic. The fibrous structure, and the planes of cleavage, thus determine the magnetic condition of a substance. The special properties presented by crystals, in regard to the action exercised upon them by magnets, is due to a particular mode of grouping their particles. This is also the cause of unequal dilatability, and of unequal conductibility for heat and for electricity. How curiously, therefore, does molecular structure determine the relation of a body to any of the forms of physical force!