152 
LIEUT. -COLONEL SABINE ON TERRESTRIAL MAGNETISM. 
From (5.) and (6.), changing- the signs of (6.) and summing, we have 
47-892 a = -4553 b + -8093 ; a = + -0263. 
We have also the equations at east and west ; 
East .... 2-61/ a = + -0645 
West .... -2-617 a =- -0753 ; 
whence 
« = pi = + -° 26 ^ 
or including the observations at east and west in the general sum, we have 
a = + -0264. 
After the arrival of the Expedition at Hobarton, and before it sailed to the Antarctic 
Circle, a similar series of observations was made in the Erebus, on the 29th October 
1840, and again repeated on her return to Hobarton the following autumn, viz. on 
the 29th June 1841. The south end of the needle being now the one which dipped 
below the horizon {& being — 70° 40'), the deviation of the compass was found to take 
place in the contrary direction to that which had been observed at Gillingham, the 
disturbance being towards the west as the ship’s head went round from north by east 
to south, and towards the east as her head passed from south through west to north. 
The line of no deviation was not found to correspond accurately with the north 
and south points of the compass on either of the occasions at Hobarton, but in 1840 
coincided more nearly with the north by west and south by east, and in 1841 with 
the north by east and south by west. We may perhaps ascribe with probability 
irregularities of this nature to slight modifications in the distribution of the iron at 
different periods, which we cannot but view as of not unlikely occurrence ; for ex- 
ample, such as might be occasioned by the ship being secured at different times by 
the starboard or the larboard chain cable. In looking through the observations of 
the Erebus, it is evident that there was no systematic or constant deviation of the 
plane of the ship’s attraction from that of her principal section ; but that the points 
of no disturbance were sometimes a little on the one side, and sometimes a little on 
the other, of the north and south points. It appears, therefore, not improper to 
class these irregularities with those others of accidental occurrence which occasion 
small discordances in partial results, and are usually ranged under the general 
technical head of errors of observation. 
If, further, we compare generally the deviations in 1840 with those of April 1841, 
the latter appear systematically rather the more considerable in amount. Viewed as 
a single fact, this circumstance might be regarded simply as indicating that some 
change had taken place in the interim in the arrangement and distribution of the 
ship’s iron, and an easy and natural explanation might appear to be afforded. It is 
however one of several facts which have presented themselves in the course of a care- 
ful examination of the observations of the first two years of Captain Ross’s expedition, 
which seem to point to the possibility of a somewhat different cause, viz. that when 
