considered essential; Graham, however, showed that there exist three series of salts—one set derived from P2O5,3H2O, one from P2O5,2H2O, and a third from P2O5,H2O. His way of stating the fact was that water could play the part of a base; for example, the ordinary phosphate of commerce possessed, according to him, the formula P2O5,2Na2O,H2O, two-thirds of the “water of constitution” being replaced by oxide of sodium. Liebig, then professor at Giessen (1803–1873), founded on these and on similar observations of his own the doctrine of poly-basic acids—acids in which one, two, three, or more atoms of hydrogen were replaceable by metals. Thus, instead of writing, as Graham did, P2O5,2Na2O,H2O, he wrote, PO4Na2H; and for orthophosphoric acid PO4H3. The group of atoms (PO4), therefore, existed throughout the whole series of orthophosphates, and could exist in combination with hydrogen, with hydrogen and metals, or with metals alone. Similarly the group (P2O7) was characteristic of pyrophosphates and (PO3) of metaphosphates, for P2O5,2H2O=(P2O7)H4; and P2O5,H2O=2(PO3)H. The first clear ideas of the structure of the molecule were, however, gained from the study of the compounds of carbon. It was difficult to apply the dualistic theory to them. For few of them are electrolytes, and therefore their products of electrolysis, being non-existent, could not be classified. Nevertheless, Gay-Lussac regarded alcohol, C2H6O, as a compound of C2H4, ethylene, and H2O, water; and oxalic acid (anhydrous), C2O3, as one of CO2 with CO. The discovery of “isomeric compounds,” _i.e._, of compounds which possess the same ultimate formula and yet differ entirely in their properties, forced upon chemists the necessity of attending to the structure of the molecule; for only by such a supposition could the difference between two isomeric bodies be explained. In 1823 Liebig discovered that silver fulminate and silver cyanate both possessed the empirical formula AgCNO; in 1825 this was followed by the discovery by Faraday that oil gas contains a hydrocarbon identical in composition with ethylene, C2H4, yet differing from it in properties; and in 1829 Wöhler, professor in Göttingen (1800–1882), discovered that urea, a constituent of urine, could be produced by heating ammonium cyanate, NH4CNO, a substance of the same formula. It therefore became clear that the identity of a compound must depend on some other cause than its ultimate composition.