# How did chemists figure out atomic weights, when hydrogen, etc., were first discovered?

Henry Cavendish isolated hydrogen in 1766, but I want to know if he knew the gas was actually $H_2$ instead of just a plain $H$ gas.

This would make a major difference in the atomic weights, which would have implications for the periodic table once people like Mendeleev noticed that atomic weights were usually integer multiples of hydrogen's.

If he knew it was $H_2$ gas, how did he know? If not, then when and how did we figure out that hydrogen gas is actually two $H$ atoms in one molecule?

Depending on the answer I may ask the same thing for Oxygen and Nitrogen.

• I believe Dalton's theory on atoms was only advanced many years later. – Gerald Edgar Feb 5 '16 at 16:56
• This question is misguided, implicitly assuming that Cavendish should have known stuff that wasn't known for a century after he discovered hydrogen. The first concepts by Lavoisier (20+ years after Cavendish) and Dalton (40+ years after Cavendish) were rather primitive. Even Mendeleev's table had issues, being based on atomic weight rather than atomic number. Fixing that would take another 40+ years. Note well: I am not derogating Lavoisier, Dalton, or Mendeleev. They were visionary geniuses. – David Hammen Feb 6 '16 at 22:43

The question would not have made sense in 1766, and for a while after. At the time chemistry was still transitioning from the dominance of alchemic ideas, notation and terminology. Lavoisier's 1789 list of elements for example still bears the hallmarks of Aristotle's four elements with multiple "earths" and "airs", and includes caloric fluid (heat) and "dephlogisticated air". Hydrogen is called "base of inflammable air". Some modern ideas about chemical composition start to appear in the work of Cullen, who drew some rudimentary reaction diagrams in 1757, and his student Black even drew schematic circles (later "molecules") split into constitutive parts (later "atoms"). Swedish chemist Bergman systematized these ideas in his influential 1775 chemistry textbook, which included 64 diagrams of chemical reactions. However, the theory was not yet sophisticated enough to assign (or even make sense of) numbers to constitutive parts, and distinguish say $H_2O$ from $HO$.
The atomic theory in a form resembling classical chemistry was only proposed by Dalton in 1803. In particular, he made the distinction between "simple atoms" (elements), "compound atoms" (molecules), and "complex atoms" (complex molecules), which would make the question meaningful. Dalton took the atomic weight of hydrogen as unity and determined the ratios of elements in some compounds based on it, e.g. the ratio for nitrous anhydride was 2 to 3 corresponding to modern $N_2O_3$. But his symbol for water $\bigodot\!\!\bigcirc$, with the circles representing hydrogen and oxygen respectively, indicates that he did not always have such information.
What enabled determination of correct chemical formulas was Avogadro's law formulated in 1811: equal volumes of gases contain equal numbers of molecules under standard conditions. This meant that ratios of molecular masses could be calculated by measuring gas samples. Moreover, Avogadro first clearly stated that even element molecules can be composed of several atoms, allowing for the possibility of $H_2$ not contemplated by Dalton. Expanding on the work of Avogadro Gaudin published many familiar chemical formulas in 1833, including the ones for $H_2$, $N_2$, $O_2$ and $H_2O$. The notation was also essentially modern, Berzelius introduced it to replce Dalton's in 1813-1814.