THE SPACE RELATIONS OF ATOMS. 497 



little to be said against such an arrangement, and for 

 the carbon atoms at least it is rendered probable by the fact 

 that it obtains in the case of the allied hexamethylene 

 (see below). On the physical side there is the fact that no 

 optically active substituted benzene has been discovered, 

 although derivatives lacking symmetry have been specially 

 examined by Le Bel to this end. All these substances 

 then, except such as possess an asymmetric atom, may be 

 classed as probably two-dimensional. 



There is another class of bodies for which a plane 

 arrangement is accepted as the cause of inactivity, namely 

 the substituted ammonias, NR'R"R"'. A priori these 

 might possess either a two or a three-dimensional configura- 

 tion, but as they are uniformly inactive, and as the complex 

 nature of the radicals attached to nitrogen in the compounds 

 tested for optical activity by Kraft, Behrend, Ladenburg 

 and others, makes it improbable that their inactivity was 

 due to internal instability, these substances also must rank 

 among the exceptions of the second class. 



(3) Three-dimensional molecules containing more than 



one ARR R" group. Besides meso-tartaric acid already 



i • 1 1 i_ cHOHC fi H, 



mentioned, we may instance hydrobenzoin ^ T - r .~ T j r , TT 



CnOHC 6 H 5 



CHOHC0 2 H 

 and trioxyglutaric acid CHOH 



CHOHC0 2 H 

 (the asymmetric atoms are italicised). 



Both these substances are inactive owing to internal com- 

 pensation. The last case is complicated by the existence 

 of an isomerism akin to the cis-trans isomerism already 

 mentioned, so that here we have two substances which 

 cannot be rendered active, corresponding to the formulae 



When there are four similar asymmetric groups we 



