An examination of accounts of sixteenth-century voyages shows that the estimations of latitude then arrived at by astronomical means were subject to widely varying error. The purpose of the present study has been to discover by practical experiments at sea, with replicas of instruments in use at the time, to what extent the size and irregularity of the errors can be attributed to the characteristics of the instruments and to what extent they arise from other causes. Observations were made from a 9-ton yacht under conditions not wholly dissimilar to those obtaining in some of the very small ships whose voyages are recorded.
The evidence for latitude errors is derived from published studies and from the writer's own analyses of three voyages made in northern waters by ships of the Muscovy Company: those of Stephen Borough of 1556 and 1557 and that of Arthur Pet and Charles Jackman of 1580. These were chosen because the masters in that Company's service received the best nagivational advice of the time3 and from the early 1550s onwards regularly recorded both latitude and compass variation. Furthermore, the relatively low altitude of the Sun in the higher latitudes was conducive to greater accuracy with a cross-staff.