[389] Appendix xix. of the Report of the United States Coast Survey for 1880 (Washington, 1882) is a paper by Charles A. Schott of “Inquiry into the Variation of the Compass off the Bahama Islands, at the time of the Landfall of Columbus in 1492,” which is accompanied by a chart, showing by comparison the lines of non-variation respectively in 1492, 1600, 1700, 1800, and 1880, as far as they can be made out from available data. In this chart the line of 1492 runs through the Azores,—bending east as it proceeds northerly, and west in its southerly extension. The no-variation line in 1882 leaves the South American coast between the mouths of the Amazon and the Orinoco, and strikes the Carolina coast not far from Charleston. The Azores to-day are in the curve of 25° W. variation, which line leaves the west coast of Ireland, and after running through the Azores sweeps away to the St. Lawrence Gulf.
[390] Navarrete, Noticia del cosmografo Alonzo de Santa Cruz.
[391] Humboldt, Cosmos, Eng. tr., ii. 672; v. 59.
[392] Cosmos, v. 55.
[393] Cosmos, v. 59.
[394] Charts of the magnetic curves now made by the Coast Survey at Washington are capable of supplying, if other means fail, and particularly in connection with the dipping-needle, data of a ship’s longitude with but inconsiderable error. The inclination or dip was not measured till 1576; and Humboldt shows how under some conditions it can be used also to determine latitude.
In 1714 the English Government, following an example earlier set by other governments, offered a reward of £20,000 to any one who would determine longitude at sea within half a degree. It was ultimately given to Harrison, a watchmaker who made an improved marine chronometer. An additional £3,000 was given at the same time to the widow of Tobias Meyer, who had improved the lunar tables. It also instigated two ingenious mechanicians, who hit upon the same principle independently, and worked out its practical application,—the Philadelphian, Thomas Godfrey, in his “mariner’s bow” (Penn. Hist. Soc. Coll., i. 422); and the Englishman, Hadley, in his well-known quadrant.
It can hardly be claimed to-day, with all our modern appliances, that a ship’s longitude can be ascertained with anything more than approximate precision. The results from dead-reckoning are to be corrected in three ways. Observations on the moon will not avoid, except by accident, errors which may amount to seven or eight miles. The difficulties of making note of Jupiter’s satellites in their eclipse, under the most favorable conditions, will be sure to entail an error of a half, or even a whole, minute. This method, first tried effectively about 1700, was the earliest substantial progress which had been made; all the attempts of observation on the opposition of planets, the occultations of stars, the difference of altitude between the moon and Jupiter, and the changes in the moon’s declination, having failed of satisfactory results (Humboldt, Cosmos, Eng. tr., ii. 671). John Werner, of Nuremberg, as early as 1514, and Gemma Frisius, in 1545, had suggested the measure of the angle between the altitude of the moon and some other heavenly body; but it was not till 1615 that it received a trial at sea, through the assiduity of Baffin. The newer method of Jupiter’s satellites proved of great value in the hands of Delisle, the real founder of modern geographical science. By it he cut off three hundred leagues from the length of the Mediterranean Sea, and carried Paris two and a half degrees, and Constantinople ten degrees, farther west. Corrections for two centuries had been chiefly made in a similar removal of places. For instance, the longitude of Gibraltar had increased from 7° 50´ W., as Ptolemy handed it down, to 9° 30´ under Ruscelli, to 13° 30´ under Mercator, and to 14° 30´ under Ortelius. It is noticeable that Eratosthenes, who two hundred years and more before Christ was the librarian at Alexandria and chief of its geographical school, though he made the length of the Mediterranean six hundred geographical miles too long, did better than Ptolemy three centuries later, and better even than moderns had done up to 1668, when this sea was elongated by nearly a third beyond its proper length. Cf. Bunbury, History of Ancient Geography, i. 635; Gosselin, Géog. des Grecs, p. 42. Sanson was the last, in 1668, to make this great error.
The method for discovering longitude which modern experience has settled upon is the noting at noon, when the weather permits a view of the sun, of the difference of a chronometer set to a known meridian. This instrument, with all its modern perfection, is liable to an error of ten or fifteen seconds in crossing the Atlantic, which may be largely corrected by a mean, derived from the use of more than one chronometer. The first proposition to convey time as a means of deciding longitude dates back to Alonzo de Santa Cruz, who had no better time-keepers than sand and water clocks (Humboldt, Cosmos, Eng. tr., ii. 672).
On land, care and favorable circumstances may now place an object within six or eight yards of its absolute place in relation to the meridian. Since the laying of the Atlantic cable has made it possible to use for a test a current which circles the earth in three seconds, it is significant of minute accuracy, in fixing the difference of time between Washington and Greenwich, that in the three several attempts to apply the cable current, the difference between the results has been less than 7/100 of a second.