the Constitution of Benzene. 431 



(fig. 8 ) is practically identical with that adopted by von 

 Baeyer in the paper mentioned, though he does not put it 

 into such a definite form. In this formula all the hydrogen 

 atoms are on the same side of the ring plane, while in the for- 

 mula (fig. 9) the hydrogen atoms are alternately on different 

 sides of this plane. Von Baeyer gives sufficient ground for 

 the adoption of the first formula, and it is also clear that this 

 formula better accounts for the di- and tetra-additive com- 

 pounds of benzene. The other formula (fig. 9), however, 

 suggests a reason for the well-known association para and 

 ortho, and the isolation of the meta disubstitution deriva- 

 tives, from the fact that the atoms in the ortho and para 

 positions are on different sides of the ring-plane, and those 

 in the meta position on the same side, it will be noticed 

 that centres of the carbon atoms (fig. 9) occupy the angles of 

 an octahedron (see Thomsen, Ber. xix. p. 2944) resembling 

 Thomsen's octahedral formula in the fact that the diagonal 

 bonds (being formed of the six affinities directed towards 

 the centre of the system) are different in nature from the 

 peripheral, which connect only individual carbon atoms, but 

 differing from his conception in the fact that these diagonal 

 bonds and not the peripheral are broken to give the hexa- 

 m ethylene nucleus. 



It is perhaps most probable that of the two formulas the 

 one which represents all the hydrogen atoms on the same side 

 of the ring-plane is the one most in accordance with facts. 



Xow it will be noticed that in these formulas all the 

 carbon atoms are asymmetric, and if replaced respectively by 

 their images give rise to the formulas of geometrical isomers. 

 The bodies so obtained are in fact the eight theoretically 

 possible isomeric benzenes, previously mentioned, obtained 

 by successive replacements of one or more of the six carbon 

 atoms by their images, but we have no evidence as to the 

 existence of any such isomers. 



There is an objection which attaches itself to these for- 

 mulas, and it is one which appears also, from the statement of 

 Hermann {Ber. xxi. p. 1958), to attach itself to every benzene 

 formula except when all the 12 atoms lie in one plane. It is 

 an objection also to the old prism formula, though I am not 

 aware that it has been brought forward before. 



The objection is this, that disubstituted derivatives of 

 benzene which contain two different substituting groups in 

 the ortho and meta positions, are each capable of representa- 

 tion respectively by two formulas, of which one is the non- 

 superposable image of the other, leading us to the prediction 

 of an isomerism which from analogy we should expect to be 

 characterized bv rotatorv power, while the properties of the 



2G2 



