i895. RESULTS OF ''CHALLENGER" EXPEDITION. ii 



2. The delineation for the first time of the contour Unes of 

 the great ocean basins. • 



3. The determination of oceanic temperatures, which subse- 

 quent observation has shown to be constant below the depth of 

 100 fathoms. 



4. The proof of a constant bottom temperature over large 

 areas in the ocean ; found to be due to such areas being separated 

 by submarine ridges from each other, and from the cold water of 

 the^ polar basins. 



5. The determination of the exact position of many islands 

 and dangers, of which the longitude, especially, had been pre- 

 viously very uncertain. 



6. The charting, as accurately as time permitted, of the 

 various unsurveyed parts of the world sighted, or touched at, 

 during the voyage. 



7. The determination of the ocean currents both on the 

 surface and at various depths. 



I. With respect to the accurate determination of the variation of 

 the magnetic needle on board ship, it is necessary to'point out that 

 before 1872 observations for variation taken by a vessel between any 

 two ports were generally dependent for correction (for the direction of 

 ship's head at the moment of observation) on the deviations obtained at 

 these ports at the beginning and end of each voyage, corrected, of 

 course, for the varying force of the magnetical elements. The usual 

 method of obtaining a table of deviations, i.e., of the errors of the 

 compass due to local attraction of the ship itself, being to land one 

 compass, and take reciprocal bearings on shore and on board, as the 

 vessel was swung round laboriously by anchors and hawsers, with her 

 head directed to each point of the compass in succession ; in some 

 cases, however, this had been done by bearings of the sun or of a 

 distant object. It is evident that the correctness of the result, and of 

 the variation curves, would be entirely dependent on there being no 

 magnetic disturbance on the land. 



Shortly after the " Challenger " left England a statement was 

 made that the Admiralty chart of variations was considerably in error 

 in the neighbourhood of the Bermudas, and instructions were given 

 that this statement was to be investigated. On the arrival of the 

 expedition at Bermuda observations were taken on shore, with the 

 result that the needle in different parts of the island showed an 

 extreme difference of 6'^ in the variation, viz., from 4° W. to 10° W. 

 — a somewhat surprising result in a group of islands of coral forma- 

 tion. The difficulty then was to find what the true error of the 

 needle was, and the following plan was adopted : — 



It is shov/n by the late Archibald Smith in the " Admiralty 

 Manual for Compasses " that if a table of deviations, or local errors 

 in a ship, be obtained by swinging the ship on a number of equi- 

 distant points, if the resulting easterly deviations be called + and the 



