THE STEREOCHEMISTRY OF CARBON The most symmetrical arrangement possible for the valencies of a quadrivalent element is that in which their relative directions correspond with those of the four axes of a regular tetrahedron (fig. 2).
This view of the spatial distribution of the carbon valencies was adopted by F. A. Kekule as early as 1867, and he had models constructed to represent the carbon atom, somewhat of the form shown in fig. 3, consisting of a sphere with four rods attached in the directions of the axes of a circumscribed regular tetrahedron. It was, however, a few years later (1873) that the attention of chemists was first directed seriously to the need for considering the spatial configuration of molecules. In that year J. Wislicenus showed that the lactic acid of sour milk and the sarcolactic acid of muscle juice (which differed in the hydration and solubility of their salts and in their optical properties, the latter being optically active) had the same chemical structure, and hence concluded that their difference could only be due to the different position of their atoms in space. In the following year J. H. van't Hoff published a pamphlet entitled "A Proposal respecting the Exten sion of Chemical Structural Formulae into Space," in which he showed that the "tetrahedral" theory of the arrangement of the carbon valencies supplied an in terpretation of the molecular dissymmetry of organic com pounds which was in accord ance with the experimental facts. Similar views were put
forward almost simultaneously by Le Bel.
On van't Hoff's theory the molecule of a carbon compound of the type Cab c c has the configuration I., as shown in fig. 4, and this has a plane of symmetry, namely, the plane passing through the edge a b and the middle point of c c cutting the tetrahedron in half as indicated by the dotted line; the molecule is therefore superposable upon its mirror-image, and a compound of this type should exist in one modification only. Replacement in this compound of one of the two radicals c by a fourth radical d will give, however, two enantiomorphous configurations, II. and III., according as the one or the other of these two radicals is replaced, for this substitution destroys the plane of symmetry and leaves a completely asymmetric molecule. A compound of the type Ca b c d containing a carbon atom linked to four different radicals should therefore be molecularly dissymmetric and should exist, according to the principles established by Pasteur, in two antimeric optically active forms. A carbon atom thus linked is termed an asymmetric carbon atom.
This conclusion accorded well with the observed distribution of optical activity in organic compounds. All optically active com