Brazil, the old Irish _Breasail_, was another name for land west of Ireland--where there is none short of America--on very many medieval maps, of which perhaps a dozen are older than the year 1400, the earliest yet found being that of Dalorto, 1325. Usually it appears as a nearly circular disc of land opposite Munster, at first altogether too near the Irish coast, as indeed the perfectly well-known Corvo was drawn much too near the coast of Spain, or as even in the sixteenth century, when Newfoundland had been repeatedly visited, that island was shifted by divers mapmakers eastward towards Ireland, almost to the conventional station of Brazil. Also, not long afterwards, the maps of Nicolay and Zaltieri adopted the reverse treatment of transferring Brazil to Newfoundland waters, as if recognizing past error and restoring its proper place. The name Brazil appears not to have been adopted by the Norsemen, but there is one fifteenth century map, perhaps of 1480, preserved in Milan, which shows this large disc-form "Brazil" just below Greenland ("Illa Verde"), in such relation that the mapmaker really must have known of Labrador under the former name and believed that it could be readily reached from that Norse colony. It seems altogether likely that "Brazil" was applied to the entire outjutting region of America surrounding the Gulf of St. Lawrence--that part of this continent which is by far the nearest Ireland. Besides the facts above stated, certain coincidences of real geography and of these old maps favor that belief, and they are quite unlikely to have been guessed or invented. Thus certain maps, beginning with 1375, while keeping the circular external outline of Ireland, reduce the land area to a mere ring, enclosing an expanse of water dotted with islands; and certain other maps show it still nearly circular externally, and solid, but divided into two parts by a curved channel nearly from north to south. The former exposition is possible enough to one more concerned with the nearly enclosed Gulf of St. Lawrence and its islands than with its two comparatively narrow outlets; the second was afterward repeated approximately by Gastoldi's map illustrating Ramusio when he was somehow moved to minimize the width of the Gulf, though well remembering the straits of Belle Isle and Cabot. There are some other coincidences, but it is unnecessary to dwell on them. Land west of Ireland must be either pure fancy or the very region in question, and it is hardly believable that fancy could guess so accurately as to two different interpretations of real though unusual geography and give them right latitude, with such an old Irish name (Brazil) as might naturally have been conferred in the early voyaging times. That an extensive region, chiefly mainland, should be represented as an island is no objection, as anyone will see by examining the maps which break up everything north of South America in the years next following the