R. Porson's conjectural readings at Rhesus 424, 442, 703, 863, 871, 883 are examined in order to contribute to the study of his philological method. It results that: 1) Porson's conjectures are not far removed from the paradosis; 2) Porson gave great wheight both to the manuscript and the indirect tradition; 3) he tried to justify his conjectures by giving examples of analogous corruptions in the same author or in others; 4) he quoted not only passages in support of his conjectures, but also of the transmitted text; 5) he sometimes tried to normalize the text both on the linguistic-syntactical and on the prosodic-metric level.