substance into two enantiomorphously related antipodes exhibiting 

 opposite rotatory power, the question as to the special stereometrical 

 configuration must in each case be brought to the fore. If the con- 

 ception of Van 't Hoff and Le Bel as to the direction of the four 

 carbon-valencies be adopted, and keeping that in mind, the stereo- 

 metrical "model" of the molecule be constructed, we have only 

 to make sure that the configuration thus obtained has no axis 

 of the second order, nor any symmetry-plane, nor a centre of 

 symmetry l ). If this be done, the compound will be such as to be 

 resolved into enantiomorphous isomerides. But if one of the symme- 

 try-properties of the second order mentioned be found in the model, 

 the substance will not be resolvable into two enantiomorphously 

 related isomerides, not even if there be eventually several asym- 

 metric carbon-atoms present in it. A few instances may be 

 exemplified here to make this clear. 



21. The oldest known example of a compound which cannot 

 be resolved into enantiomorphous isomerides although it possesses 

 several asymmetric carbon-atoms, is meso-tartaric acid: 



COOH 



*CHOH 

 *CHOH 



COOH 



Evidently the one half of the molecule is here the mirror-image 

 of the other half; or what is the same thing: the stereometrical 

 configuration of atoms has here a plane of symmetry, and therefore 

 it cannot differ from its mirror- image. Formerly this was explained 

 by saying that the rotatory effect of the one half of the molecule 

 was counterbalanced by that of the other half, and such molecules 

 were therefore said to be inactive and non-resolvable by internal 

 compensation. The explanation given here is however better, because 

 it does not introduce any superfluous hypothesis, but elucidates 



l) The "plane of indirect symmetry", as mentioned by Ladenburg, and by 

 Groth in the discussion of the impossibility to resolve certain diketo-piperazines 

 into antipodes; corresponds to no other operation than inversion. Such molecules 

 cannot be separated in enantiomorphous modifications, because they have an 

 inversion-centre, which is equivalent to a binary-axis of the second order; they 

 are therefore congruent with their mirror-images. It is quite superfluous to 

 introduce here any new name. 



