When arranging the elements, Dmitri Mendeleev did not force every position to be filled. He left spaces for unknown elements and predicted some of their properties.
Later discoveries, including gallium and germanium, corresponded with those predictions. The modern periodic table is based on atomic number, differing from the early arrangement.
Early history
In 1817, German physicist Johann Wolfgang Döbereiner began one of the earliest attempts to classify the elements. In 1829, he found that he could form some of the elements into groups of three, with the members of each group having related properties. He termed these groups triads. Chlorine, bromine, and iodine formed a triad; as did calcium, strontium, and barium; lithium, sodium, and potassium; and sulfur, selenium, and tellurium. Various chemists continued his work and were able to identify more and more relationships between small groups of elements. However, they could not build one scheme that encompassed them all.
John Newlands published a letter in the Chemical News in February 1863 on the periodicity among the chemical elements. In 1864 Newlands published an article in the Chemical News showing that if the elements are arranged in the order of their atomic weights, those having consecutive numbers frequently either belong to the same group or occupy similar positions in different groups, and he pointed out that each eighth element starting from a given one is in this arrangement a kind of repetition of the first, like the eighth note of an octave in music (The Law of Octaves). However, Newlands's formulation only worked well for the main-group elements, and encountered serious problems with the others.
Mendeleev
The definitive breakthrough came from the Russian chemist Dmitri Mendeleev. Although other chemists (including Meyer) had found some other versions of the periodic system at about the same time, Mendeleev was the most dedicated to developing and defending his system, and it was his system that most affected the scientific community. On 17 February 1869 (1 March 1869 in the Gregorian calendar), Mendeleev began arranging the elements and comparing them by their atomic weights. He began with a few elements, and over the course of the day his system grew until it encompassed most of the known elements. After he found a consistent arrangement, his printed table appeared in May 1869 in the journal of the Russian Chemical Society.
When elements did not appear to fit in the system, he boldly predicted that either valencies or atomic weights had been measured incorrectly, or that there was a missing element yet to be discovered. In 1871, Mendeleev published a long article, including an updated form of his table, that made his predictions for unknown elements explicit. Mendeleev predicted the properties of three of these unknown elements in detail: then-missing heavier homologues of boron, aluminium, and silicon; he named them eka-boron, eka-aluminium, and eka-silicon ("eka" being Sanskrit for "one"). In 1875, the French chemist Paul-Émile Lecoq de Boisbaudran, working without knowledge of Mendeleev's prediction, discovered a new element in a sample of the mineral sphalerite, and named it gallium. He isolated the element and began determining its properties. Mendeleev, reading de Boisbaudran's publication, sent a letter claiming that gallium was his predicted eka-aluminium.
Mendeleev nevertheless had some trouble fitting the known lanthanides into his scheme, as they did not exhibit the periodic change in valencies that the other elements did. After much investigation, the Czech chemist Bohuslav Brauner suggested in 1902 that the lanthanides could all be placed together in one group on the periodic table. He named this the "asteroid hypothesis" as an astronomical analogy: just as there is an asteroid belt instead of a single planet between Mars and Jupiter, so the place below yttrium was thought to be occupied by all the lanthanides instead of just one element.