Mr. M. M. Pattison Muir on Chemical Classification. 197 



is necessary to fix our attention upon the chemical properties no 

 less than upon the physical properties of groups of' compounds. 

 Assuming the hypothesis of valency, and given a certain 

 number of elementary atoms and the valencies of these atoms, 

 we deduce the possible existence of a group of compounds. 

 Have we any chemical means for recognizing the existence of 

 these compounds ? From a study of the chemical processes 

 involved in the formation, in the reactions, and in the decom- 

 positions of a group of bodies, can we gain any insight into 

 the structure of these bodies ? 



Now, granting that there is some kind of structure, we may 

 safely assume that difference in structure is correlated with 

 difference in chemical behaviour. From a study of chemical 

 behaviour we should then arrive at a knowledge of the struc- 

 ture of the bodies under consideration. Let us take an ex- 

 ample. Analysis and vapour-density determinations point to 

 the formula C2 Hg as expressing the empirical composition 

 and molecular weight of common alcohol. From the theory 

 of valency (the valencies of the constituent elements being C'^, 

 H', and 0^^) we deduce the possibility of the existence of two 

 compounds, each having the empirical formula C2 Hg 0: for one 

 of these we have the constitutional formula CH3 — CH2 — — H, 

 for the other the formula CH3 — — CH3. Which constitutional 

 formula is to be assigned to alcohol? When alcohol is acted 

 upon by sodium or potassium, one atom of hydrogen is ex- 

 changed for one of metal ; the remaining five atoms of hy- 

 drogen cannot be so exchanged : it is also possible to substitute 

 a monovalent group for one of the hydrogen atoms in alcohol. 

 From the constitutional formula CH3 — CH2 — — H, we 

 should infer that one of the hydrogen atoms would be ready 

 to undergo reactions which would not affect the remaining 

 five. The formula CH3 — — CH3 points to no such difference 

 between the atoms of hydrogen. Further, the first formula 

 represents the sixth atom of hydrogen as alone in direct com- 

 bination with oxygen: those substances which readily replace 

 one atom of hydrogen in alcohol are themselves ready to 

 combine with oxygen. Other reactions are known in which 

 a monovalent group or element replaces the whole of the 

 oxygen and simultaneously one atom of hydrogen in alcohol : 

 thus, 



C2 He + HC1= C2 H5 CI + H2 0, 



C2HeO + HN03=C2H5N03 + H2 0. 



The possibility of such reactions is evident if we adopt for 

 alcohol the constitutional formula CH3 — CH2 — — H. But 

 the product of the first reaction formulated above, C2 H5 CI, 



