IN the July issue of Science Progress, Mr. N. H. de V. Heathcote brings together the data at present available for fixing the date of the discovery of the deviation of the compass from true north. The figure in the instructions of the Nautical Chart of Bianco of 1436 which has been taken as evidence that a correction for a deviation of 18° west of true north was allowed for, the author concludes has nothing to do with deviation. The first definite record of deviation he considers to be that made by Columbus in September 1492, during his first voyage across the Atlantic. In his record of his return in 1496 from his voyage to India, Columbus mentions Flemish compasses which read 11¼º west of north when the Genoese compasses read north. Mr. Heathcote points out that pocket sundials were in use in Germany about the middle of the fifteenth century which were set in the meridian by a compass; in the latter an allowance for a deviation of 6° east of true north was made, while in Etzlaub's road map of Germany of 1492 instructions are given for orienting it correctly by compass, an allowance for deviation of 11¼º being made. He concludes that pocket sundials with an arrow set 11¼º east of north were familiar objects in Germany before the time of Columbus.