The vignette describes lobar pneumonia in an older man, and the laboratory clue of gram-positive cocci on blood agar narrows the field to the alpha-haemolytic streptococci. The clinically dominant pathogen in this setting is Streptococcus pneumoniae, so the question is really asking how to prove that a green-haemolytic gram-positive coccus is pneumococcus rather than a viridans streptococcus. Two features set pneumococcus apart and both rely on its powerful autolytic enzyme. Bile salts such as sodium deoxycholate activate this autolysin, dissolving the colony, which is the bile solubility test. The same autolysin makes the organism sensitive to optochin (ethylhydrocupreine). Viridans streptococci lack this behaviour and are both bile insoluble and optochin resistant. Working through the other options, the catalase test only separates the staphylococci (positive) from streptococci (negative) and is therefore useless for telling two streptococci apart. Bacitracin sensitivity is the marker for group A beta-haemolytic Streptococcus pyogenes. The oxidase test is reserved for gram-negative rods and cocci like Pseudomonas and Neisseria. Hence the single most useful identifying test for this gram-positive coccus is bile solubility. \[\boxed{\text{Bile solubility}}\]