574 REPORT— 1902. 



What is symbolised by position formulae, and indeed by the formula altogether, 

 are the chemical activities and abilities of the substance and its derivatives, and 

 their analogies with those of other substances. When not in interaction, a sub- 

 stance has no constitution and no formula. It is certainly not on any experimental 

 grounds that it can be regarded as some spatial arrangement of unlike parts. To 

 take the simplest case ; if we start with sodium hydroxide and symbolise its mole- 

 cule by some mark, such as X to begin with, the interaction of the substance with 

 an acid leads us to replace the X by two symbols and a connecting mark. One of 

 these will be Na for the sodium radical ; let the second be Z for the other radical, 

 and let a dot or stroke be placed between the symbols to mark them as those of a 

 pair of radicals in interaction. In other interactions, such as that with melted 

 potassium acetate, we find need for a new pair of symbols, one being H for the 

 hydrogen radical, while the other may be Q. But it is easy to decompose two 

 molecules of sodium hydroxide in one operation into molecules of sodium, hydro- 

 gen, and oxygen, from which fact we learn that Z is replaceable by the double 

 symbol 0-H, and Q by 0-Na. Thus, Na-Z and H-Q become equally Na-O-H, 

 which records the ultimate radicals of sodium hydroxide, together with all iis 

 interactions, immediate and remote. But it does this with no more implication of 

 spatially placed and tied parts than is made by expressing the measured flow of 

 time by a straight line, or than is to be found in f- seconds of time, or in c' as the 

 third power of a number, unless we specifically condition this symbol as stereo- 

 metric. A formula is not to be read — on experimental grounds, I mean — ^as a 

 symbol of parts juxtaposed and joined on, and should be regarded as an intricate 

 but legible monogram telling the chemical nature of the substance. Every symbol 

 in it is to call to mind a phase of the chemical activity of the substance or of its 

 derivatives, a phase that may be for the time as the substance itself to the iuAesti- 

 gator, just as a pigment substance becomes only a red or a white to the painter. 

 For example, salt is often uothing more than its chlorine phase to the chemist 

 when he wants only a soluble chloride ; whether it is of potassium, sodium, or 

 ammonium, then matters not to him. 



The double linking of the carbons in ethylene is a symbolised expression ot 

 facts, without reference to hypothesis. The two carbon radicals of ethane or of 

 alcohol behave together just as does the single carbon of methane or the nitrogen 

 of ammonia in being, but with a valency of six, continued to other compounds 

 devoid of all the other radicals of the ethane or alcohol — that is, of the hydroxyl 

 and the hydrogens. The quadrivalency of each carbon is made up by the inter- 

 action necessary to difssociate or to bring together the two methyls, which counts 

 as a unit of valency to each carbon. Ethyl hydrogen sulphate decomposes into 

 sulphuric acid and ethylene, the hydrogensulphate radical with a hydrogen radical 

 becomes the acid as the one product, while the methylene radicals again pair otf 

 as the two methyls had done when ethane was formed, thus producing the non- 

 saturated substance, ethylene. Since there is a perchlorethylene, the .second link- 

 ing mark falls between the two carbons ; and when ethjdene passes back to an 

 ethane compound two units of valency are displayed by it without the carbons 

 becoming dissociated. 



Position formulfe of isomerides, such as those of propyl alcohol and acetone, 

 present no difficulty, because they are interpreted as the expressions of unlike 

 double decompositions. It is not luifrequently the case that no constitutional or 

 structural formula can be given to a substance which shall express all the pairs of 

 radicals possible in its interactions, of which the best-studied example is that of 

 ethyl acetoacetate. This state of things, known as tautomery, admits of no other 

 interpretation than that there are really two substances existent, of which one only 

 is known, the other or so-called ' pseudoform ' requiring the assumption of its 

 existence as a transition-substance only. The notion of the shifting hydrogen 

 radical is but the hypothetical way of viewing the intervention of the intramole- 

 cular change by which the substance becomes its ' pseudoform.' 



The cyclic formida of benzene expresses the fact that, unlike a fatty hydro- 

 carbon, benzene shows but one pair of interaction radicals, hydrogen and phenyl. 

 The ' ortho-,' * meta-/ and * para-' positions in benzene derivatives are only eSpres- 



