: by 
ON THE MAGNETIC INTENSITY OF THE EARTH. 37 J 
themselves indicating otherwise, the usual course is to distri- 
bute a loss equally through the interval in which it is known 
to have occurred. I have therefore pursued this course in 
regard to the loss sustained by No. 7; and in the case of No. 8 
I have allowed a double proportion in each of the first fourteen 
months. The observations furnish two tests of the propriety 
of this distribution: the general agreement of the results of 
the two needles with each other at the different stations is one ; 
the other is the agreement of the force thus calculated at Rio 
in 1817 and 1821. In both the accordance is satisfactory. 
On computing the intensity at the Cape of Good Hope and 
the Isle of France by No. 9, using for that purpose its time of 
vibration at Paris in 1817, the results appeared to agree ex- 
tremely well with those of Nos.7and 8. It is hence inferred, that 
until the accident at the Isle of France, No. 9 had undergone 
no change of magnetism, and I have therefore brought into the 
account all the results obtained with it before that occurrence. 
As the effect of changes of temperature on these particular 
needles does not appear to have been ascertained experiment- 
ally, no corrections are applied on account of temperature ; 
but, as I have before remarked, such corrections are of minor 
importance in so extensive a series as the present. The table 
in page 40 exhibits the computed results, and appears to need 
no other explanation, except that the column entitled ‘Time of 
vibration as a dipping-needle at Paris” exhibits the times of 
vibration corrected for loss of magnetism. 
