120 
LIEUT. -COLONEL SABINE ON TERRESTRIAL MAGNETISM. 
which the correction is 0-0093) was sustained by No. 7 - at Mazatlan in November 
1839, and has been noticed and the circumstances connected with it stated in No. II. 
of these Contributions*. The corrections of Nos. 5. 7 - 9. 11. 12. and 13, derived from 
the intercomparison, commencing with the observations at Panama in March 1839, 
and ending with those at Woolwich in December 1842, are as follows : they are all 
additive at the stations named and at all antecedent stations, and render the whole 
series with each needle comparable throughout with each other, and with the obser- 
vations at Woolwich in December 1842. 
No. 5. 0 - 0143 from Seychelles; 0-0027 (additional) from San Bias. 
No. 7- O'Ol 15 from Seychelles ; 0-0013 (additional) from Amboyna ; and 0-0093 
(additional) from Mazatlan. 
No. 9. 0-0018 from Point de Galle ; 0-0021 (additional) from Singapore, October 
1840; 0-0038 (additional) from Tahiti. 
No. 11. 0-0035 from Jobie Island. 
No. 12. 0"0020 from the Cape of Good Hope; 0"0012 (additional) from Point de 
Galle; 0-0030 (additional) from Macao, September 1841. 
No. 13. 0-0034 from Amboyna; 0-0017 (additional) from Tahiti. 
We have next to examine how far the corrections thus derived correspond with the 
change in the times of vibration of Nos. 7- and 9, shown by direct observation at 
Woolwich at the two dates of August 1839 and December 1842. 
No. 7. No. 9. 
s s 
In August 1839 the times of vibration observed were .... 767’ 0 627*6 
The corrections assigned by intercomparison are equivalent to . + 19’8 + 5'6 
Observed times of vibration in August 1839, corrected . . . 786'8 633'2 
Observed times of vibration in December 1842 789" 1 635‘7 
Differences 2*3 2*5 
The differences are in the same sense in both needles, and are such as would cor- 
respond to a small loss of magnetism still undetected by the intercomparison. It is 
possible that they may be partially or wholly due to this cause, without prejudice to 
the effectiveness of the method itself, because, in its present application, the inter- 
coinparison has not been pursued to the correction of very small differences; it has 
however been carried out to such extent, that even the very moderate differences in- 
dicated by the above figures would not have escaped detection had they occurred at 
any single station or period. They correspond respectively to *006 and "007 parts of 
the whole horizontal force; and were there no liability to observation errors, — and were 
the horizontal force at Woolwich at all times a constant quantity, — we might be justi- 
fied in meeting them with a special correction. But they are exceeded in amount 
by the fluctuations of the force itself, which is subject both to periodical variations, 
dependent on the hour and season, and to other fluctuations of irregular occurrence, 
* Philosophical Transactions, 1841, p. 15. 
