B.— CHEMISTRY 39 



four pairs of dots, by which the four pairs of electrons of the octet are 

 indicated, at the four angles of a tetrahedron concentric with the atom 

 under consideration. The diagram will then indicate that another atom 

 linked by a given pair of electrons will be situated on the axis through the 

 corresponding angle of the tetrahedron. 



Fig. I. 



There is little doubt that atoms are deformable, and we have clear 

 evidence that valencies can be deflected. We can therefore only expect 

 the tetrahedron to be a regular tetrahedron when the central atom is 

 linked to four other atoms and these are all alike. The tetrahedral octet 

 will therefore only indicate the general character of the configuration of 

 the compounds represented with its aid ; it cannot be expected to predict 

 accurate values of intervalency angles. Subject to this limitation there 

 is, as far as I am aware, no established fact relating to the stereochemistry 

 of compounds formed in accordance with the octet rule which is at variance 

 with the indications of this tridimensional diagram. 



Where in such compounds we have to infer the presence of two- 

 electron links only, then we find invariably that a four-co-ordinate atom 

 has a tetrahedral configuration, a three-co-ordinate atom a pyramidal, 

 and a two-co-ordinate atom an angular configuration. This is illustrated 

 by the configurations of methane, ammonia and water. 



Fig. 2. 



The compounds of four-co-ordinate nickel, platinum and palladium, 

 for which there is much evidence for a planar configuration, evidently do 

 not come within the scope of this rule. The effective atomic numbers 

 of these metals in such compounds are each two short of the atomic 

 numbers of the succeeding inert gas. The three-co-ordinate compounds 

 of the elements of the third group of the periodic table likewise lie outside 



