of the magnetic Needle. i 93 
the Investigator seem to bespeak that it is so; they arc as 
follows. 
Time. 
Latitude. 
Longitude. 
Number 
of com- 
passes 
used. 
Number of 
sets of obser- 
vations taken. 
Place of the 
compass. 
Suppos- 
ed true 
varia- 
tion. 
Observed 
variation. 
Ship’s 
head. 
Observer. 
1801. 
21, PM 
Start Point 
in sight . 
two 
5 azimuths 
binnacle 
— 
29 34 w 
W 
Mr. Thistle 
— 
to the NE 
one 
1 amplitude 
— 
— 
29 30 
— 
22, PM 
49 0 10' N 
^ 2 ^' W 
two 
4 azimuths 
r upon the 
— 
24 1 2 
WNW 
. 
— , AM 
48 15 
6 45 
one 
1 amplitude 
1 booms in 1 
j the middle f 
— 
24 49 
WSW 
— 
28, PM 
38 1 
14 20 
five 
10 azimuths 
l of the ship 1 
— 
20 57 
SW 
— — 
— 
— 
— 
1 1 — 
binnacle 
— 
z 5 34 
— 
Commander 
31, PM 
Porto Santo in sight 
two 
4 — - 
— 
22 45 
— 
Mr. Thistle 
_ 
to the NW 
— 
4 — 
booms 
19 51 
— 
— 
. 24, AM 
10 20 
22 15 
one 
2 — 
binnacle 
— 
12 45 
SE b.S 
Jommander 
29, AM 
5 40 
16 30 
two 
4 ~ 
— 
— 
12 18 
— 
Lt. Flinders 
5, AM 
2 15 
14 00 
3 ~ 
— * 
*4 54 
WSW 
Mr. Thistle 
These observations, particularly those of July 28, seem to 
be decisive in showing that the variation is more westerly 
when taken upon the binnacle of a ship whose head is west- 
ward in north latitude, than when observed in the center of 
the ship, which is a strong confirmation of the suppositions 
before given ; but the observations on the change of the ship's 
head are too few to be satisfactory. Almost every sea officer 
can tell whether he has observed the variation of the compass 
to be greater when going down the English Channel than 
when coming up it : and indeed it would be very easy for a 
ship lying in harbour to ascertain the point beyond controversy. 
Should this point be well established, I think it would follow, 
that from a high south latitude where the differences are great 
on one side, they are most likely to decrease gradually to the 
equator, and to increase in the same way to a high north lati- 
tude, where they are great on the other side ; thus the smaller 
mdcccv. C c 
