Mr. M. M. Pattison Mnir on Chemical Classification, 205 



and we find that a second class of these bodies may be best 

 regarded as containing a new group — OH — OH. For in- 

 stancCj we have primary butyl ic alcohol with the constitu- 

 tional formula OH3— OH^— CH2— OH2— OH ; but we have 

 also secondary butvlic alcohol with the constitutional formula 

 CH3— CH2— CH— CH3. If the latter alcohol really contain 



I 

 OH 



the group — CH — OH, we are compelled to regard the posi- 

 tion of this group as different from that assigned to the group 

 — CH2 — OH in the primary alcohol. Chemists speak of the 

 OH as placed " at the end of the chain " in one case, and in 

 the '' centre of the chain " in the other. 



Such ideas as these, founded on the theory of ^^alency, de- 

 veloped by that of atom-linking, and, I think, in keeping with 

 the reactions of the bodies to be formulated, compel us to 

 believe that the action of each atom in a molecule is exerted, 

 to a certain extent, upon all the other atoms ; that just as in the 

 planetary system we have a regular order of bodies, each oc- 

 cupying its own position and performing its own proper 

 movements, yet each influencing and perturbing every other, 

 so in a chemical molecule we have a group of atoms regularly 

 arranged, each performing fixed and definite vibrations, each 

 occupying a definite relative position, yet each exerting some 

 influence upon every one and upon all of the remaining atoms*. 



Such views as these are an outcome of the theories of va- 

 lency and of atom-linking : they have led to many discoveries 

 in chemistry in the past, and probably will be of yet greater 

 importance in the future. In using them, however, it is very 

 necessary to remember the small amount of knowledge which 

 we really possess concerning the connexion existing between 

 structure and chemical properties. Having gained the general 

 idea of such a connexion, and finding that from what we know, 

 or think we know, we can explain a considerable number of 

 facts, we are ever ready to apply the explanations found to 

 suit a few cases to every case, and so we come to imagine 

 that our knowledge is practically complete. 



For this reason, although I think that constitutional formulas 

 are one of the greatest aids which the chemist can possess him- 

 self of in his investigations, yet I also think that these formulte 

 must be carefully employed — that it is necessary for the chemist 

 ever to remind himself that he is dealing with an hypothesis only, 

 and Avith an hypothesis which is as yet but in the first stages 

 of its development. The indiscriminate use of such expres- 



* In the paper already cited I have endeavoured to trace the bearing 

 of this view of the molecule upon the general theory of isomerism. 



