The Bakerian Lecture .
On the elementary Particles of certain Crystals .
By William Hyde Wollaston , M.D. Sec. R.S. Read November 26 , 1812 .
[ Phil. Trans. 1813 , p. 51 .
] In this lecture the author undertakes to explain a difficulty that has occurred in crystallography , respecting the primitive molecule of those bodies that assume the octohedral and tetrahedral forms , when broken in the direction of their natural fractures .
The substance that he selects as most convenient for experiment is fluor spar , which may very readily be divided into any number of acute rhomboids , having the angles of their surfaces 600 and 120 ? .
These might be regarded for all the practical purposes of crystallography as the integral molecule , and from thence all the other modifications of these solids might very simply be derived .
But it is observed , that each acute rhomboid thus obtained may be again split in a new direction at right angles to its axis , so that a tetrahedron may be detached from each extremity , leaving from between them a regular octohedron .
Consequently this rhomboid cannot be considered as the primitive , and we are left in doubt not only which to prefer of the two lastnamed solids , but even whether either of these can be primitive ; since no possible arrangement of tetrahedra alone , or of octohedra alone , will fill any space without leaving vacuities .
The author having observed that both these forms would result from the arrangement which spheres would most naturally assume by mutual attraction , proposes a theory founded on that observation , which he thinks is not liable to objection , and endeavours to extend this hypothesis , by showing that with some modifications a corresponding theory may be applied to other forms well known to crystallographers .
With regard to the triangular arrangement of balls in a plane , and their tetrahedral grouping in solidity also , he finds that lie has been anticipated by that universal genius Dr. Hooke : but he observes , that Dr. Hooke 's ideas upon this , as upon many other subjects , are but imperfectly developed ; and that he takes no notice of the octohedral group , formed by placing four balls in a square , with one above and one beneath them .
Accordingly , Dr. Hooke could know nothing of that which forms the principal novelty of the present observation , namely , that when a mass of spheres has been formed wholly according to the triangiular or tetrahedral arrangement , then certain sections of the same mass present also the square arrangement , and acco rsingly certain portions of it exhibit the octohedral group .
The author remarks , in support of this theory , that a large proportion of those substances which assume the octohedral form , are considered by chemists as simple bodies , and are therefore more likely to have the simple form of spheres than such as consist of more than one element .
Since the supposition of spherical particles appeared to him to afford so satisfactory an explanation of an acknowledged difficulty in crystallography , he was led to consider what other forms would result from the union of solids most nearly allied to spheres ; and he observed that obtuse rhomboids , like those of carbonate of lime and other substances , would be formed by the union of oblate spheroids , as indeed Huyghens had long since observed ; and that by the union of oblong spheroids , the natural result would be triangular and hexangular prisms , as are found in beryl and phosphate of lime .
But the most singular arrangement noticed , is that which affords an explanation of the origin of cubes in crystallography .
These , he supposes , may consist of spherical particles , of two different kinds , regularly intermixed in equal numbers ( in conformity to the most recent views of binary combination in chemistry ) ; for these , he observes , will not tend , as before , to the octohedral arrangement , but will be perfectly in equilibrio when every group of eight balls composes a cube , according to the most obvious course of alternation of the two elements .
For in that case all similar balls will be equidistant from each other , and will also be equally distant from all adjacent balls of the opposite denomination .
In a note are subjoined some observations on a theory of crystallization proposed by M. Prechtl , who imagines that a mass of soft spheres may all be compressed into tetrahedra , which is demonstrably impossible .
That by another degree of softness or of attraction , spheres , each surrounded by five others , may be compressed into triangular prisms , without regard to the different degree of compression that must take place in the direction of the axis ; that other spheres again less compressible than before , and consequently surrounded by as many as six others , may be formed into cubes , which indeed is admitted to be a very possible supposition .
It is observed , however , that M. Prechtl denied that ; a sphere can be surrounded by more than six , although , in fact , the most probable supposition is , that each soft sphere would be surrounded by twice that number , and would form a mass of regular dodecahedra .

