LIEUT.-GENEEAL SABINE ON TEKEESTEIAL MAGNETISM. 455 



90" at west, 180° at south, and 270° at east, the true magnetic direction was, everywhere 

 in the southern hemisphere, less on the points from 0° to 180° and more on the points 

 from 180° to 360°, than the amount actually shown by the compass-card. It thus hap- 

 pened that, as a general practice in the ' Erebus' and ' Terror' whilst in the southern hemi- 

 sphere, the deviations were recorded as negative, or — , on the western side of the com- 

 pass-card, and positive, or -f, on its eastern side, the signs so employed having no direct 

 relation whatever to the distinct question whether the Declination itself were easterly or 

 westerly. In these Contributions, and in conformity with general usage in treatises on 

 Terrestrial Magnetism, the Declination is counted east when the north end of the 

 magnet declines from the Geographical North towards the east, and west when the 

 declination declines towards the west; and as both east and west declinations are found 

 in different parts of the southern magnetic hemisphere, east declinations being charac- 

 teristic by the — sign, and west declinations by the -\- sign, the effect of the deviations 

 having a — sign, was to augment the apparent or observed declination on the eastern 

 points, and diminish it on the western points, in those parts of the hemisphere where 

 the declination itself was east ; and, vice versa, to diminish the apparent or observed 

 declination on the eastern points and augment it on the western points in those parts 

 of the hemisphere where the declination itself was west. 



Corrections applied to the Observations of the Declination /or the Ship's Attraction. 



1. In the '■Erebus' — In the subjoined Table (No. I.), columns 2 and 3 exhibit the 

 deviations observed in the ' Erebus' on the points specified in column 1, at Port Louis in 

 the Falkland Islands, on August 19, 1842, recorded in the Philosophical Transactions 

 for 1844, page 88, and at Simon's Bay at the Cape of Good Hope on the 20th of April, 

 1843, which are now printed for the first time. As the observations at Port Louis were 

 at the commencement of the third year's survey, and those at Simon's Bay at its close, 

 and as the dip at the two stations was very nearly the same in amount, a mean of the 

 deviations of the declination at the commencement and close of the year's survey has 

 been adopted, and placed in column 4, as the foundation of the calculated deviations to 

 be ascribed to intermediate times and localities. With these values of the deviation on 

 the several points, the constants B, C, D, and E in the equations by which the deviations 

 in dips of other amounts may be computed have been obtained, employing for that 

 purpose the method described in the Philosophical Transactions for 1846, Art. XVIII., 

 pages 350-352. The constants thus derived are as follows: — 



B=-2°32'*; C=-0°08'; D=+0°22'; E=-t-05'. 



* B supplies the weU-known coeflScient a so much used in the eaxlier Numbers of these Contributions 

 (B=atanfl). Comparing only those values of a which were obtained after the arrival of the 'Erebus' in the 

 Southern Hemisphere, the mean of the observations at Hobarton in 1840 and 1841 gave a=-0272 (Philoso- 

 phical Transactions, 1843, Art. X. p. 154) ; those at the Falkland Islands in August 1842, -0292 (Philoso- 

 phical Transactions, 1844, Art. VII. p. 88); and by the observations now discussed a=-0331; the increased 

 value being doubtless due to the magnetism acquired and temporarily retained in the high southern dips to 

 which the ' Erebus' had been subject whilst in the Antaictic seas. 



3s2 



