with both its diagonals. I presume also that it has been shewn, by the agreement in the incidences of the plane P on the edge b c of the fracture, fig. 9. with the plane P on the edge b c of the crystal, fig. 11. as well as of the plane P with the plane a of each of those figures; that the plane P of the former figure, is really the result of a cleavage parallel with the plane P of the latter figure; and also, by fig. 10. that a mechanical division is equally practicable parallel with each of the four planes P, composing the pyramid of the crystal fig. 11. Let therefore all these cleavages be represented on fig. 15. and it will be seen that the result is a mechanical division of it into tetrahedrons.
It has already been said that in the first instance the Abbé Haüy, was induced to believe the cube to be the primitive form of the oxyd of tin, but that he was afterwards led to adopt the flattened octahedron composed of the two pyramids of fig, 15. joined base to base. In this latter opinion, there seems to me, from the evidence now offered, no room for doubting his correctness. For whatever has been said tending to shew a connexion between the fractures that have been described, and a crystal delineated by fig. 15. relates, with equal aptitude, to one having either a longer. or a shorter prism, and equally well to one having no prism at all: for it will be seen by fig. 17. that the form of fig. 15. is merely the result of a decrement on the edges of an octohedron formed by the meeting of its two pyramids base to base; which octohedron is given by itself fig. 18. as the primitive form of the oxyd of tin. But it has not hitherto been seen unmodified; nor has any crystal been noticed approaching it more nearly than that delineated by fig. 21. Pl. 16.
- On the subject of the integrant molecule I do not feel competent to say more, than that it has been already shewn, by a combination of all its known cleavages, that the primitive crystal is mechanically divisible into tetrahedrons; but as these tetrahedrons will necessarily be irregular or rather unequal in their form, it may not be satisfactory