B.— CHEMISTRY 41 



Had the confidence then existed in the octet theory and its stereo- 

 chemical impHcation that we have since acquired, the pyramidal configura- 

 tion of the three-co-ordinate sulphur in these compounds could have 

 been most easily predicted. 



A comparison of the sulphoxides with the N-ethers of the oximes is 

 possibly not without interest, since it shows that the presence of a 



R Ri 



■0 y^'^^ 



R R 



co-ordinate (or semipolar) link is not in itself sufficient to produce a 

 pyramidal configuration. In the oxime ether the nitrogen atom and 

 the three atoms directly attached to it are evidently co-planar. 



Very considerable interest attaches to the possibility of the existence 

 of compounds of tri-co-valent carbon of pyramidal configuration. The 

 triaryl alkyl compounds of the alkali-metals in all probability contain 



R 



the anion R : C : If this is so, this ion must certainly be pyramidal, and 

 R 



hence molecularly dissymmetric when the three aryl radicals are different, 

 but it is not possible at present to predict how stable the antimeric con- 

 figurations would be. However, questions of considerably greater 

 theoretical impoitance centre about another type of compounds in which 

 there seems reason to infer the presence of a pyramidal tri-co-valent carbon 

 atom of marked configurational stability. 



When Hantzsch discovered the aci-fonn of phenylnitromethane he 

 appeared to have obtained exceedingly strong evidence of the correctness 

 of Michael's formula for the sodium salts of the nitroparafiins. There 

 seemed little doubt that the aci-iorm of a secondary nitroparaflin had the 

 constitution which in modern symbols we should represent as : 



R OH 



R O 



In a compound of this structure the > C = N < group and the four 

 atoms directly attached to it would evidently lie in one plane. The 

 remarkable discovery by R. Kuhn and Albrecht that optically active 

 methylethylnitromethane retained its activity when converted into its 

 sodium salt showed therefore that the Michael formula could not 

 possibly be right. 



With Mr. H. Cole I have recently obtained confirmation of Kuhn 

 and Albrecht 's discovery. We have been able to prove that the ac/-form 

 of phenylcyanonitromethane is molecularly dissymmetric, since we find 

 that the compound can be resolved with alkaloids in the ordinary way. 

 When the brucine salt is converted into the sodium salt a strongly 

 optically active solution is obtained. The activity of the sodium salt 



c 2 



