However, only a few years later Thomas Graham (1805-1869) proved that the three phosphoric acids are not isomeric. He used the proportion of 2 P to 5 O in the oxide which Berzelius had thought justified at least until "an example of the contrary could be sufficiently established."[21] Refining the techniques of Gay-Lussac (1816) and several other investigators, Graham characterized the three phosphoric acids as "a terphosphate, a biphosphate, and phosphate of water." Actually, this was the wrong terminology for what he meant and formulated as trihydrate, bihydrate, and monohydrate of phosphorus oxide. In his manner of writing the formulas, each dot over the symbol for the element was to indicate an atom of oxygen; thus, he wrote: ... :: .. ... . . H^{3} P H^{2} P and H P.[22] [Illustration: Figure 5.--OVEN FOR THE CALCINATION OF BONES, about 1870. "The operation is carried out in a rather high oven, such as shown.... The fresh bones are thrown in at the top of the oven, B. First, fuel in chamber F is lighted, and a certain quantity of bones is burnt on the grid D. When these bones are burning well, the oven is gradually filled with bones, and the combustion maintains itself without addition of other fuel. A circular gallery, C, surrounds the bottom of the oven and carries the products of combustion into the chimney, H. The calcined bones are taken out at the lower opening, G, by removing the bars of grid B." (Translation of the description from FIGUIER, _Merveilles de l'industrie_, volume 3, 1874, page 537.)] [Illustration: Figure 6.--AN ADVERTISEMENT with view of plant for manufacturing superphosphate about 1867. (From E. T. FREEDLEY, _Philadelphia and its Manufacturers in 1867_, page 288.)] Graham had come to this understanding of the phosphoric acids through his previous studies of "Alcoates, definite compounds of Salts and Alcohol analogous to the Hydrates" (1831). Liebig started from analogies he saw with certain organic acids when he formulated the phosphoric acids with a constant proportion of water (aq.) and varying proportions of "phosphoric acid" (P) as follows: 2 P 3 aq. phosphoric acid 3 P 3 aq. pyrophosphoric acid 6 P 3 aq. metaphosphoric acid. [Illustration: Figure 7.--FLORIDA HARD-ROCK PHOSPHATE MINING. (From Carroll D. Wright, _The Phosphate Industry of the United States_, sixth special report of the Commissioner of Labor, Government Printing Office, Washington, 1893, plate facing page 43.)] Salts are formed when a "basis," i.e., a metal oxide, replaces water. When potassium-acid sulfate is neutralized by sodium base, the acid-salt divides into Glauber's salt and potassium sulfate, which proves the acid-salt to be a mixture of the neutral salt with its acid. Sodium-acid phosphate behaves quite differently. After neutralization by a potassium "base" (hydroxide), the salt does not split up; a uniform sodium-potassium phosphate is obtained. Therefore, phosphoric acid is truly three-basic![23]