(i.) The first method is illustrated by Pasteur's separation of the antimeric sodium ammonium tartrates. Whilst it is thus of great historical interest, it has been of comparatively little prac tical value in the subsequent development of stereochemistry on account of its exceedingly limited applicability, for when an equi molecular mixture of the d- and 1-forms of a molecularly dis symmetric substance crystallizes, it almost always happens that uniform crystals, built up of pairs of enantiomorphous molecules, are formed. Such crystals, being analogously constituted to those of racemic acid, are termed racemic crystals, and the association of d- and /-molecules of which they are formed is termed a racemic compound, though there is no evidence that the molecules con stituting each pair are linked by forces different from the cohesive forces binding together the molecules in a crystal.
(ii.) The second method of resolution may be illustrated by considering the salt of an inactive molecularly dissymmetric acid with an active base. If the symbols + and — are used to denote enantiomorphous configurations, the antimeric forms of the acid may be represented as (+A) and ( —A), and the base, e.g., a laevorotatory alkaloid, as ( —B). Combination of the acid and the base will then yield the two salts (+A) (—B) and (—A) (—B). It is clear then that two compounds which can be thus repre sented are not antimeric ; they are said to be diastereoisomeric. Similarly an inactive dissymmetric base (+B), ( —B) will give rise with a dextrorotatory acid (+A) to the diastereoisomeric salts (+A) (+B), (+A) ( —B). Diastereoisomeric compounds,
being different and not antimeric, must differ to a greater or less degree in all their properties, and a pair of diastereoisomeric salts may in fact differ very considerably in solubility so that they can be easily separated by fractional crystallization. Each then yields on decomposition one of the pure antimeric forms of the acid or base as the case may be. This is the most important of the methods of resolution, and most of the very large number of optically active substances that have been prepared artificially have been obtained by its means.
(iii.) The biochemical method depends on the different effects of living organisms on antimeric compounds. Thus moulds grown in a solution of ammonium racemate destroy the d-tartrate and leave the /-tartrate ; yeasts added to solutions of inactive glucose ferment the naturally occurring d-glucose and leave the /-antimer ; similarly, rabbits fed with the inactive forms of common a-amino acids assimilate or destroy the active modification occurring in nature and excrete the other. Since digestive chemical changes are mainly effected through enzymes, the biochemical method of reso lution evidently depends on the well-known high specificity of enzyme action. Enzymes act very unequally on antimers, and must therefore themselves be dissymmetric; Emil Fischer used the simile that the enzyme fits the substrate as the key fits the lock.