338 Colonel E. Sabine on the Magnetic Variation in 



obtained at the observatory itself, and at stations distant from 

 it a few miles in diiFerent directions. Amongst these may 

 be named the variations most carefully observed by Admiral Du 

 Petit Thouars in False Bay in April 1839 (29° 9'), and by Sir 

 James Clark Ross in Simon's Bay in April 1840 (29° 23'). The 

 general result is also confirmed by nearly cotemporaneous obser- 

 vations (1837 to 1845), made at sea in the vicinity of the Cape 

 by Admirals Vaillant and Du Petit Thouars, and Captains Sir 

 James Clark Ross, Crozier, Moore and Clark, which are detailed 

 in the memoir referred to, and are inserted in the map ; the 

 mean of all the sea observations, viz. 28° 51', being shown in the 

 map to correspond to the intersection of lat. —35° and long. 

 17° 30' E., which, with a proper allowance for latitude and lon- 

 gitude, accords almost precisely with 29° 07' at the observatory 

 in lat. -33° 56' and long. 18° 29' E. 



So far in regard to the accm-acy of the map in relation to the 

 period for which it was constructed, viz. 1840 ; but it may be 

 asked, may not the magnetic variation in the vicinity of the 

 Cape have altered since 1840 by the operation of secular change, 

 so as to make the determinations on which the map was based 

 inapplicable to the present epoch ? It is well known, that 

 generally in the Southern Atlantic the westerly variation in- 

 creases at a nearly uniform rate of secular change of about 8' in 

 the year; and it is also known that the same rate of secular 

 increase, or more exactly 7'"56, did obtain at the Cape of Good 

 Hope from the earliest observations of Davis and Keeling in 1605 

 and 1609, to those of Captain FitzRoy and of Sii* James Ross 

 in 1836 and 1840. (Cape Magnetic Observations, vol. i. p. Ix.) 

 Had this rate of secular change (which had lasted so long) con- 

 tinued to 1854 (when Captain Klein's observations were made), 

 the variation at the Cape would have been nearer 31° than 

 29° 07' ; that is to say, the westerly variation in my map would 

 have been in defect in 1854, instead of in excess as Dr. Buys Ballot 

 supposes. The Cape, however, was selected as a station for a 

 magnetic observatory,' amongst other reasons for this one in par- 

 ticular, that the form of the variation lines in its vicinity, and 

 their known progress from east to west in the temperate zone of 

 the southern hemisphere, gave reason to believe that the secular 

 increase of the variation which had taken place for so many years 

 was approaching the period of its termination, when it would be 

 succeeded by a nearly stationary variation continuing for several 

 years. The probability of this alteration in the secular change is 

 noticed in the memoir accompanying the map. (Phil. Trans. 1849, 

 p. 200.) That it has been realized, has been shown by the 

 records of the Cape Observatory. Between January 1841 and 

 August 1850, the westerly variation had increased from 29° 7' only 



