22 
LIEUT.-COLONEL SABINE ON TERRESTRIAL MAGNETISM. 
Table VI. (Continued.) 
Station. 
Needle. 
Corrected Time. 
Horizontal Intensity. 
Number of 
Needles. 
Days. 
Observations. 
Clarion Island. . . 
5 
478-7 
962 
962 
1 
1 
2 
5 
475-4 
9751 
7 
541-0 
983 
8 
472-5 
981 
Martins Island < 
9 
438-1 
981 
^980 
7 
5 
17 
11 
474-6 
981 
12 
403-6 
984 
13 
394-8 
978 J 
5 
480-9 
9531 
7 
548-9 
955 
8 
479-4 
953 
Bow Island. . . . < 
9 
443-9 
955 
>957 
7 
8 
55 
11 
479-5 
961 
12 
408-2 
962 
13 
399-1 
957^ 
In this Table each needle has been given an equal influence on the mean result, with- 
out reference to the number of observations made with it. Where the observations 
do not afford certain and independent evidence of the unchanged state of each of the 
needles in respect to magnetism, weights assigned from other considerations must 
necessarily be arbitrary and uncertain. For example, at the last station in the Table, 
Bow Island, thirty-six observations were made with No. 5, and not more than three 
or four with each of the other six needles. But we have already seen, on intercom- 
parison, reason to suspect that No. 5. may have sustained a slight loss of magnetism 
at the station preceding Bow Island, and it is the only needle in which any change 
of the kind is indicated subsequently to the general comparison at San Bias. Whilst, 
therefore, on the one hand, we might not be justified, without more clear and decided 
evidence, in altogether setting aside the result with No. 5, so on the other hand we 
should not obtain the most probable final deduction, by giving to that result a weight, 
in comparison with that of each of the other needles, proportioned to the number of 
observations, and resting on the probable error of a single observation, — apart from 
changes of magnetism in the needle itself. 
We will now revert to the stations visited in the first voyage which were not sub- 
sequently revisited, and at which the values of the horizontal intensity may be de- 
rived by means of Nos. 5, 6, or 8. Table VII. contains an abstract of the obser- 
vations with these needles, in all of which the times of vibration were obtained in 
arcs commencing with 40°. The column entitled “ Corrected Times,” shows the 
mean time of vibration reduced to a standard temperature of 60°. The arithmetical 
mean of the times at Panama in March 1837, October 1839, and March 1839, has 
been taken as the approximate time of vibration at Panama throughout the interval ; 
and the ratio of the horizontal intensity at the other stations has been computed 
accordingly, as shown in Table VIII. In this Table, as in Table VI., each needle 
