COLUMBUS' DISCOVERY OF VARIATION. 197 



stand longitudinally in the direction of this meridian. It 

 then points to the geographical north pole, and is said to 

 have no variation. But if the north-pointing end lies to 

 the east or west of the meridian, then it is said to have 

 east or west variation. This it is absolutely necessary to 

 allow for in steering a ship, or in running a line in survey- 

 ing land, or in laying out a railway. The variation is not 

 the same at all points on the earth, nor is it constant at 

 any one point. Therefore, there is a variation of the varia- 

 tion, which is secular in that it occurs over very long 

 periods, besides being annual and even diurnal. Besides 

 these changes there are irregular variations or perturba- 

 tions, due to disturbances in the earth's magnetic field, 

 which need not here be considered. 



As I have already stated, the Chinese knew, certainly as 

 early as the nth century and probably before, that the 

 needle did not point to the true north and south, an<3 they 

 constructed their land compasses to allow for the angle of 

 discrepancy. But there appears to be no record showing 

 that any European ever recognized the variation of the 

 needle as a cosmical phenomenon before Columbus did 

 so on his memorable voyage. 1 



Now, briefly, what happened to Columbus was this. On 

 his first voyage to America, on the evening of September 



1 True, much has been written (See Libri: Hist, des Sci. Math, en Italic. 

 Paris, 1830, vol. ii., 71. Formaleoni: Saggio Sulla Nautica Antica del 

 Veneziaui. Venice, 1783, 51-2. Humboldt: Cosmos, v. Irving: Life of 

 Columbus) concerning an old chart made in 1436 by Andrea Blanco, and 

 now in the Library of St. Mark in Venice, upon which appears a figure 

 supposed to represent the points of the compass with a correction for 

 variation. This, however, Bertelli (Sulla Epistola di P. Peregrine. 

 Rome, 1868, niem. iii., 77.), has investigated and finds no suggestion of 

 variation present the correction simply being that necessary to apply to 

 the courses of a ship sailing on rhumbs of the compass (as N. E., N. W.) 

 to keep clear of the loxodromic curve, or endless spiral due to the cur- 

 vature angle between the earth and the meridian, which would never 

 lead to any determined point. I shall not take space here to repeat Ber- 

 telli's demonstration of this error; while the other anticipations upon 

 which he comments are merely inferences which he easily disproves. 



