The idea of portraying the accuracy of a navigational system by means of contours is very old and was first introduced to readers of this Journal by Jessell and Trow. Later writers, Sameshima, Namikawa and Hasegawa, have published accuracy contours for Loran chains. This note is a further contribution in this direction and its aim is to investigate what sort of accuracy can be achieved by marrying a position line from one chain with a position line from another.

N. Sameshima has studied the accuracy contours arising from a hyperbolic system where position lines are obtained from two pairs of stations, each pair having one station, the master station, in common. Y. Namikawa and K. Hasegawa, in a later article, constructed accuracy contours for a number of Loran chains, mainly in the Pacific, but also in the Atlantic areas. This article discusses the more difficult case where fixes are obtained through using pairs with no common station, and indicates what improvement may be expected through introducing a fourth station in a number of selected instances.