Page:A Voyage to Terra Australis Volume 2.djvu/539

On ship-board.] It appears that differences, probably similar to those in the Investigator, were also observed on board La Recherche, one of the ships with which the French admiral D'Entrecasteaux went in search of the unfortunate La Pérouse. Monsieur Beautemps-Beaupré, the able surveyor to the expedition, found so much uncertainty in compass bearings that he abandoned, as far as was possible, the use of them; substituting the sun's azimuth and angular distance from some one point, and measuring the angles from that point to other objects. He says of the compass, "We found by a great number of observations, but principally by the differences between the bearings of points set with each other from opposite directions, that no confidence could be had in bearings taken with the compass from the deck of a large vessel, nearer than to 3°, even under the most favourable circumstances. For instance, it has often happened that from one position, as C, the cape A has been set in a line with cape B; and afterwards, from another position, D, cape B has been set with A; and that we have found considerable differences in the results of the two observations. We also remarked, that the compass showed differences of several degrees in variations at sea, though observed with the greatest care and within the space of a few minutes." (Voyage de D'Entrecasteaux, par M. de Rossel. Vol. I. p. 600. A Paris, 1808.)

I do not find any other distinct mention of differences found in the variation, from changing the direction of the ship's head or the place of the compass; but it appears from the following extract, that the Investigator was not singular in having a variation of 4° greater than the truth in the English Channel. Captain Vancouver, in his passage towards Madeira, says (Vol. I. p. 6.), "The error in reckoning, amounting almost to a degree (of longitude), seemed most likely to have been occasioned by our not having made sufficient allowance for the variation of the compass on our first sailing; as, instead of allowing from 22° to 25°, which was what we esteemed the variation, our observations for ascertaining this fact, when the ship was sufficiently steady, shewed the variation to be 28° and 29½° westwardly."