Why classify, and the first attempts

Only 31 elements were known in 1800 and 63 by 1865; today elements up to Z = 118 are known and named. Studying each one and its compounds separately is impractical, so chemists looked for a system that would organise known facts and predict new ones.

Dobereiner (German, by 1829) found groups of three similar elements, triads, in which the middle element's atomic weight is roughly the average of the other two and its properties lie in between: Li 7, Na 23, K 39; Ca 40, Sr 88, Ba 137; Cl 35.5, Br 80, I 127.

Check a triad: (7 + 39)/2 = 23 for Na, and (35.5 + 127)/2 = 81.25, close to Br's 80. The law of triads held for only a few elements and was set aside as coincidence.

de Chancourtois, a French geologist, arranged the known elements by increasing atomic weight on a cylinder in 1862 to show recurring properties; it drew little attention.

Newlands (English, 1865), Law of Octaves: listed by atomic weight, every eighth element resembles the first, like the eighth note of a musical scale. It worked only up to calcium, but the Royal Society later gave him the Davy Medal (1887).

Lothar Meyer (German) graphed physical properties such as atomic volume and the melting and boiling points against atomic weight. The pattern repeated, but its length grew along the list. His 1868 table resembled the modern one closely, yet appeared in print only after Mendeleev's.