of oblique solids, by introducing the idea of polarity in the axes of crystallization, Mr. Dana has successfully applied this molecular theory to crystallography, yet he goes no farther; and the most important and difficult steps in this branch of physical science still remain to be made, and many phenomena in crystallization, with the cause of which we are at present wholly unacquainted, still require to be explained by the theory. The author particularly refers to the important facts discovered by MM. Leblanc and Beudant, of the influence that solutions or mediums in which bodies crystallize have on the secondary forms which these bodies take; and states, that the present views of crystallography afford not even a glimpse of the least relation between such forms and the properties of the mediums. Why, he asks, does pure water appear, in general, to tend to simplify the forms, precisely as do certain mixtures, those of chlorite in axinite, quartz, felspar, &c., and why, on the contrary, do other mediums, acid or earthy, complicate them?
Impressed with the importance which must attach to the solution of such questions, M. Necker offers some ideas which long meditation on this important subject has suggested to him.
Adopting the ellipsoid as the form of the molecule, he remarks, that the more complicated the form of the crystal, the more the number of its faces increases, and the more, at the same time, does it approach to the ellipsoidal form of the molecule; and, on the contrary, the simpler the form becomes, the more does it recede from that with a curved surface. All crystalline forms may be considered as making a part of one or more series, which, in each system of crystallization, have for extreme terms, on the one side, the most simple solid of the system, or that which has the least number possible of faces, and on the other, the solid having the greatest number, namely a sphere or an ellipsoid. Although it is more convenient in the calculation of forms to start from the most simple polyhedral forms in order to arrive at the more complex, nothing proves that such has been the route which nature has followed. As long as we considered the integral molecules as polyhedral, it appeared natural to view them as grouping in polyhedrons; but when once we cease to admit polyhedral molecules, it then becomes most natural to suppose, that ellipsoidal molecules should have a tendency, more or less decided, to group in solids of the same form as themselves, when no extraneous circumstances interpose an obstacle to this tendency.
In order to give an idea of the kind of effect which would be produced on the form of the solid by these obstacles, such as the nature of the medium in which crystallization takes place, a hurried or tumultuous crystallization, &c., the author conceives that each molecule, as well as each solid formed by their union, has different axes of attraction, endued with different degrees of energy, and symmetrically disposed in groups, the weaker and the most numerous round the stronger, which are, at the same time, the smallest in number; all, in short, symmetrically arranged around the principal axes of crystallization, which are the most energetic of all. Thus we shall conceive that sort of polarity by which crystallization is distin-
