Discusses the problem of "ghost" Burns manuscripts, recorded in the past but not located by recent editors; reports that BuR 955, a manuscript of Burns's song "The Chevalier's Lament," last reported at auction in 1930, can now be reclassified as a forgery by Alexander Howland "Antique" Smith (1859-1913)), based on a torn-out page from the 1930 auction catalogue (in the Roy Collection, University of South Carolina) annotated by J. DeLancey Ferguson (1888-1966); and discusses conflicting evidence for the date at which Burns composed the second stanza of the song.