456 SCIENCE PROGRESS. 



the short line joining them is almost a point. The fourth 

 class is composed of more complicated molecules, as CH 3 C1, 

 C 2 H 4 , C 2 H e O, C 2 H 5 C1, to which, therefore, the form of flat- 

 tened sphenoids is attributed. 



The stereochemistry of elements other than carbon has 

 made little progress, except in the case of nitrogen. 

 Speculations on the space relations of nitrogen compounds 

 have long been rife, but it is only of recent years that they 

 have acquired practical interest. In the case of trivalent nitro- 

 gen, as in the case of carbon, it was the discovery of isomers 

 unforeseen by the structural formulae which led to the new 

 doctrines. These mysterious nitrogen isomers were found par- 



R' \ 

 ticularly among the oximes, R „ C = NOH, and here it is 



constantly observed that when R' and R" are not identical, an 

 isomerism arises which is lacking when identity is established. 

 Thus aldoximes exhibit this isomerism, but it is lacking in 



tj ^>CNOH ; again, it is noticed in nineteen non-sym- 

 metrical ketoximes but is lacking in the di-phenyl com- 

 pound. When the number of = CNOH groups in the 

 molecule increases, the number of isomers also increases, as 

 in the case of benzil dioxime and phenyl glyoxime. (Some- 

 times, however, only two isomers can be isolated.) Later, 



the phenyl hydrazones, y\cNNHC 6 H 5 and y> CNN 



(C 6 H 5 ) 2 , were found to present similar isomerism, provided 

 always that the groups X, Y, were dissimilar. Observing 

 that this dissimilarity was essential to the occurrence of the 

 isomerism, Hantzsch and Werner were led to explain it by 



RN NR 



the formulas II II 



XCY XCY 



In one isomer the radical R attached to the nitrogen 



atom is attracted towards X, in the other towards Y, the 



nitrogen valences being diverted by these attractions from 



the plane which, according to reasoning previously given, 



they ordinarily occupy ; and there is always a difference 



