516 
APPENDIX, 
[Errors in variation, 
Observations taken at different dips of the needle in the Northern mag* 
netic Hemisphere, from which are deduced the errors of variation at the 
Investigator’s binnacle when the ship’s head was at East or West. 
Time 
1801. 
Lat. N. 
Lon. W. 
Az. or 
Amp. 
Comps. 
Observers 
Ship’s Hd 
Variat. 
Observ. 
West. 
Variat. 
supp d 
true. 
Errors 
of 
observ". 
Errors at 
at S 
points. 
Dip 
N°. 
Remarks. 
July 
p. m. 20 
49 50 
49 48 
o / 
3 52 
4 2 
Az. 
Amp. 
No. 1,2 
No. 1 
T 
West. 
29 34 
29 30 
25 40 
O / 
3 54- 
3 50- 
J 3 52 
0 
72 
Off the Start. 
August 
a. m. 29 
5 38 
16 50 
Az. 
No. 1,2 
F 
S. E. by S. 
12 18 
13 15 
0 57 + 
1 43 
32 
Off the Afri- 
Sept, 
a. m 5 
2 21 
14 15 
— 
— 
T 
w. S. W 
14 54 
13 40 
1 14- 
1 20 
26 
can Coast. 
At dip 72°, error for 8 points 3° 52', or in -parts of the dip ,0537 
29, - - - 1 3H - - - ,0526 
Mean error for eight points in parts of the dip ,0531 
these are all the observations taken in the northern hemisphere, 
where there was any certainty of knowing the true variation ; for not being 
then aware of the difference which changing the direction of the ship's 
head produced, I did not seek to multiply the observations, nor to take 
them at opposite points.* In the southern hemisphere, the survey of Terra 
Australis required the variation to be frequently observed, and it will be 
seen in the following table what considerable differences arose on the 
direction of the ship’s head being changed ; it will also be remarked, that 
the errors were of an opposite nature to those in the northern hemisphere, 
and decreased with the dip of the needle. 
* The mean of two variations or bearings taken with the head in directly opposite 
points, will be the correct variation or bearing, however much the attraction in the ship 
may make them differ from each other. 
