The word takes many forms on maps and in manuscripts: as Brasil, Bersil, Brazir, O’Brazil, O’Brassil, Breasail. As a personal name it has been common in Ireland from ancient days. The “Brazil fierce” of Campbell’s “O’Connor’s Child” may be recalled by the few who have not wholly forgotten that beautiful old-fashioned poem. Going farther back, we find Breasail mentioned as a pagan demigod in Hardiman’s “History of Galway”[72] which quotes from one of the Four Masters, who collated in the sixteenth century a mass of very ancient material indeed. Also St. Brecan, who shared the Aran Islands with St. Enda about A.D. 480 or 500, had Bresal for his original name when he flourished as the son of the first Christian king of Thormond. The name, however spelled, is said to have been built up from two Gaelic syllables “breas” and “ail,” each highly commendatory in implication and carrying that note of admiration alike to man or island. Quite in consonance therewith the fifteenth-century map of Fra Mauro in 1459[73] not only delineated and named this Atlantic Berzil but appended the inscription “Queste isole de Hibernia son dite fortunate,” ranking it as one of the “Fortunate Islands.” [Illustration: FIG. 4--Section of the Dalorto map of 1325 showing Brazil, Daculi, and other legendary islands. (After Magnaghi’s photographic facsimile.)] ANOTHER SUGGESTED DERIVATION On the whole, this seems the more likely channel of derivation of the name; or, if there were two such channels, then the more important one. For there is another suggested derivation, of which much has rightly been made and which we must by no means neglect. Red dyewood bore the name “brazil” in the early Middle Ages, a word derived, Humboldt believed,[74] by translation from the Arabic _bakkam_ of like meaning, on record in the ninth century. He notes that Brazir, one form of the name, as we have seen, recalls the French _braise_, the Portuguese _braza_ and _braseiro_, the Spanish _brasero_, the Italian _braciere_, all having to do with fire, which is normally more or less red like the dye. He does not know any tongue of medieval Asia which could supply _brasilli_ or the like for dyewood. He suggests also the possibility of the word’s being a borrowed place name, like indigo or jalap, commemorating the region of origin, but cannot identify any such place. His treatment of the topic leaves a feeling of uncertainty, with a preference for some sort of transformation from “bakkam” which would yield “brazil” probably by a figure of speech.