It has hitherto been generally presumed that the division of the horizon into thirty-two points was a development of the late medieval period. Such a division, it has been said, was impossible in the pre-compass era. ‘It is questionable whether even so many as sixteen directions could have been picked out and followed at sea so long as Sun and star, however intimately known, were the only guides’, one eminent authority has declared; ‘Even the sailors in the north-western waters had only four names until a comparatively late date.’ Chaucer's reference in his Treatise on the Astrolabe to the thirty-two ‘partiez’ of the ‘orisonte’ has for long been quoted as the earliest evidence on the subject. The Konungs Skuggsjà, a thirteenth-century Norwegian work, however, refers to the Sun revolving through eight œttir; and the fourteenthcentury Icelandic Rímbegla talks of sixteen points or directions. An important discovery by the distinguished Danish archaeologist, Dr. C. L. Vebæk, in the summer of 1951, brings a new light to the whole problem and makes the earlier held view scarcely tenable. Vebæk was then working on the site of the Benedictine nunnery (mentioned by Îvar Bárdarson in the mid-fourteenth century) which stands on the site of a still older Norse homestead on the Siglufjörd, in southern Greenland. Buried in a heap of rubbish under the floor in one of the living-rooms, together with a number of broken tools of wood and iron (some of them with the owner's name inscribed on them in runes) was a remarkable fragment of carved oak which evidently once formed part of a bearingdial. This was a damaged oaken disk which, according to the archaeologists, dates back to about the year 1200.