1869: Mendeleev made the gaps in chemistry testable

Mendeleev arranged elements by atomic weight and chemical similarities, leaving room for unknown elements. Predictions developed in 1871 later found support in gallium and germanium. This was a powerful empirical classification, with errors and competitors—not a complete modern explanation of atoms.

A blank position can carry more information than a filled one. In 1869, Dmitri Mendeleev proposed a classification that connected known elements and indicated where unknown ones might belong. The lasting advance was a framework that could guide questions and be confronted with later discoveries.

A pattern, not just a list

A chemical element is defined by the number of protons in its atoms' nuclei. Nineteenth-century chemists did not yet have that modern explanation. They compared atomic weights, chemical reactions and the proportions in which elements formed compounds.

Mendeleev's 1869 German abstract, translated by Carmen Giunta, groups chemically analogous elements while considering increasing atomic weight. Its table leaves unknown positions, including entries near aluminum and silicon. The abstract predicts undiscovered analogues with atomic weights between 65 and 75.

Chemical resemblance could therefore challenge a simple numerical ordering. A disputed atomic weight or an empty position became a question to investigate. The table was an empirical proposal: recurring properties were observed and organized before their microscopic cause was understood.

From an empty position to a prediction

An analogue is an element expected to resemble another chemically. In the work developed in 1871, Mendeleev used names such as eka-aluminum and eka-silicon for proposed elements beyond aluminum and silicon in his classification.

The translated 1871 excerpt estimates an atomic weight of about 68 and a specific gravity of about 6 for eka-aluminum; it gives about 72 and 5.5 for eka-silicon. These are later, more detailed predictions. Attributing the entire set to the 1869 table would merge two stages of the work.

The reasoning compared neighbouring elements and their compounds. It was constrained inference from chemical relationships, not a modern electronic-structure calculation. The proposed substances still needed to be found, separated and characterized.

What a later discovery actually confirmed

In his 1877 account of gallium, Paul-Émile Lecoq de Boisbaudran dates his first indications to 27 August 1875. He reports two specific-gravity measurements, 5.90 and 5.97, with a mean of 5.935 at 23°C, relative to water at 23°C. Specific gravity is a dimensionless density ratio.

That result was close to the prediction of about 6. It did not confirm every proposed reaction or property. Lecoq says he initially did not know Mendeleev's description and identifies differences, including the unexpectedly low melting point. Finding gallium was experimental work, not simply filling a box on command.

Winkler's 1886 accounts describe the difficult separation of germanium from associated elements. Historical analysis links gallium 1875, scandium 1879 and germanium 1886 to the predicted analogues. Discovery date, publication date and later identification with a prediction are distinct events.

Why the table survived its mistakes

Mendeleev was not the only chemist developing classifications. Lothar Meyer and earlier classifiers also developed relationships among elements. Philip Stewart's historical study discusses this shared development and the problems posed by rare-earth elements and unsuccessful predictions.

The significance lies in combining an organizing pattern with predictions open to correction. Successful matches strengthened the framework, while mismatches demanded revised measurements, placement or interpretation. A few successes do not imply that every forecast was correct.

The modern table orders elements by atomic number: the number of protons in the nucleus. Electronic structure helps explain recurring chemical behaviour. Those later concepts should not be projected back as Mendeleev's original mechanism. The IUPAC table and documentation also distinguish atomic weights, isotope information and conventions for presenting the table.

Methods, sources and limits of this reconstruction

The 1869 abstract was read in Giunta's English translation; the later prediction passage was read in an English excerpt translated by Eduard Farber. These are access aids to primary historical writings, not independent experiments. The file labelled 1871 has conflicting bibliographic metadata: the detailed predictions belong to the work developed in 1871, while the German version is dated 1872 in Giunta's annotated historical bibliography. The original complete Russian and German volumes were not read here.

The gallium and germanium discovery excerpts were compared with the earlier predictions. The two gallium values and their mean are reported by Lecoq; no new chemical measurements or statistical tolerance were established. Experimental identification, separation and calibration would require a qualified laboratory. This article does not propose a home experiment or claim an independent replication.

The Science History Institute's 1871 table record supplies documentary context. The cover is an AI conceptual illustration of regular blank cards with one gap. Its geometry is not a historical facsimile, a complete periodic table or evidence of a particular working method.

Sources and editorial transparency

Article written and translated entirely by an AI system from the cited sources, with automated checks. These checks do not constitute human review, peer review or independent replication.