Finding Out What Something Is

Five parts on the question every other track assumed had been answered: given an unlabelled sample, how does anybody know what it is? It opens on the flame test, usually taught as a list of colours to memorise and actually a consequence of the shell model — heat lifts an electron a fixed step, it falls back, and the energy it returns is the colour you see, which is why the same lines appear in a street lamp, a firework and a star. The second sorts the standard gas tests by what they do rather than pairing each gas with a result, and notices that the commonest outcome is the splint simply going out, which rules out two candidates at once. The third separates a mixture without pushing or filtering anything, by setting two attractions in competition and letting each substance settle where its own balance puts it. The fourth is a procedure whose order is load-bearing: a carbonate gives a white precipitate with two different reagents and would be confidently misidentified, so the acid goes in first. And the last admits that none of this is how a real laboratory works, and explains why the chemistry is identical and the reading is not.