224 Mr. J. F. Heyes on the 



Now the development of the chemistry of the carbon 

 compounds has shown us that not merely is there chemical 

 contiguity of some sort between the atoms in a molecule 

 (e. g. H 2 = HHO or HOH), feut that there are special and 

 preferable contiguities between certain atoms or (for brevity's 

 sake, and sometimes provisionally for want of knowledge) 

 groups of atoms. These special associations, often termed 

 " atomic linkings," are seen to keep together through 

 many formative and transformative reactions ; and they 

 are indicated in our so-called constitutional formulae, where 

 these special associations are indicated by so-called " links " 

 or "bonds/"' The extraordinary progress in complex or- 

 ganic synthesis, the thousands of syntheses which have 

 resulted in so few years of labour in the great and, on the 

 whole, wonderfully systematic and connected text-books of 

 organic chemistry, compel us, I think, to retain the notion of 

 " links " or " bonds," or, as I prefer to call them, " valencies," 

 in writing these constitutional formulae. Not merely, then, 

 do we write divalent oxygen as -0- but also as 0= in 

 our constitutional formulae, according as we interpret, and 

 fairly consistently interpret, our reactions. Thus, in the 

 case of water, alcohol and ether, with their analogues, do we 

 find -0-, but in ketones and aldehyds =0, and in the 

 organic acids both aspects of divalent oxygen. Similarly the 

 nitrogen atom may be directly specially associated with three 



other atoms (as in ammonia, N ), or it may not ; but this 



does not necessarily interfere with its usually trivalent, and 

 perhaps occasionally pentavalent, character. 



It is in some such sense as this that I believe that oxygen 

 is occasionally, and probably in more cases than is usually 

 supposed, tetravalent. Pattison Muir* gives what is pro- 

 bably the most generally accepted definition : — " The atom of 

 oxygen can directly act on and be acted on by two other 

 atoms in a molecule ;" but adds, " in some molecules there 

 are only two atoms, one of which is oxygen." This of course 

 refers to the well-known gaseous puzzles, NO, CO, compared 

 with the standard molecule HOI. Perhaps it would be more 

 cautious to say, instead of " act on," "are directly chemically 

 associated with," since it is this statical aspect which our 

 formulae indicate. It is not therefore the case that a tetra- 

 valent element is necessarily associated with four monovalent 

 elements. Where this is the case, as in the hydrides, halides, 

 and methides, I should propose that the old word " tetratomic " 

 be used. It is here, in fact, that the idea of the inherent 

 * Muir, * Thermal Chemistry,' p. 67. 



