OPTICAL ISOMERISM. Most of the cases not explained by the doctrine of the linking of atoms are presented by compounds identical in all their chemical and physical properties except the power of rotating the plane of polarized light. It is well known that this power is possessed by a number of crystalline substances in the solid state, and that every such substance can be ob tained in two 'enantiomorphons' crystalline forms, rotating the plane of polarized light through equal angles hut in opposite directions. But when melted or dissolved such substances usually lose that power completely—which shows that their optical property is peculiar, not to their nature, but to the particular crystalline form which they assume in the solid state. These are not the substances that stereo-chemistry has to deal with: such substances, inasmuch as they lose their difference of optical rotatory power with their crystalline form and are therefore identical in the liquid, gaseous, or dissolved state, cannot he considered as different chemical individuals, and hence cannot form the subject of a theory of chemical compounds. Cases of optical isomerism proper are presented by sub stances retaining or acquiring their difference of optical rotatory power in the liquid, gaseous, or dissolved state.
In 1S74 Le Del and Van 't Hoff discovered, in dependently and almost simultaneously, that all optically active compounds, and only optically ac tive compounds, contain one or snore `asym metric' carbon atoms, i.e. carbon atoms directly linked to four atoms or groups of atoms differ ent from one another. The following formulas, containing asymmetric carbon atoms (denoted by an italic C), represent optically active com pounds: Yet only one methylene chloride is possible. If the valencies of the carbon atom were assumed to be distributed, in the plane, not at equal angles, but, say, as in the symbol which contains no asymmetric carbon atom (i.e. no atom combined with four different atoms or groups), have in no case been found to represent optically active compounds. The relation be tween the optic rotatory power and the asym metric carbon atom must therefore form the cornerstone of any explanation of optical isomer ism that might be advanced. Altogether three
different explanations suggest themselves as pos sible: (1) the assumption that the four valencies of the carbon atom are unequal and that cases of isomerism may result from different distribu tions of the four atoms or groups combined with the carbon atom, among its different valeneies; (2) the assumption that the atoms or groups combined with the carbon atom arc in constant motion around it and that the molecules of opti cally different compounds may be made up of the same atoms or groups in different motions around the carbon atom; and (3) the assumption that the molecules of different compounds may be made up of the same atoms or groups, similar ly linked to the carbon atom, but differently ar ranged in space around it. The first of these assumptions compels us to recognize the possible existence of isomers that are positively known to be impossible. For instance, if the assump tion were correct, two different nitro-methanes would be possible: yet a systematic experimental investigation carried out by Henry showed that nitro-methane can exist in one form only. The second of the above assumptions is alto gether incapable of either purely theoretical development or systematic correlation with facts, and must therefore be abandoned. The third as sumption. viz. that the difference between opti cal isomers is due to differences in the configura tion of the atoms in their molecules, presents it self, therefore, as the only possible explanation. Furthermore, even this explanation must first of all discard the hypothesis of the arrangement of atoms in a plane. Indeed, were the plane graphical formulas, as we ordinarily write them, true to reality, then, for instance, two methylene chlorides would he possible, the configurations of whose atoms would be represented by the follow ing formulas: then the number of theoretically possible yet really non-existing isomers would be even greater.