MAGNETIC INTENSITY. 509 
supposition would imply a disturbing cause vitiating the series as 
a measure of the magnetic intensity at the station. I have not ven- 
tured, therefore, to draw any conclusion from these observations, 
farther than to notice, as above, the limits within which, in either 
of the two first suppositions, the intensity would fall. 
A nearly similar reasoning applies to the observations at Keel- 
ing Islands ; of three series, one is decidedly so irregular, that no 
inference could be drawn from it; in the two other series the irre- 
gularities are neither so frequent, nor so large: my general impres- 
sion (in the uncertainty created by the irregularity of the first 
series), is, that the majority of the intervals are of twelve vibra- 
tions, and not of ten: if of twelve, the intensity would be about 
1,21 ; if of ten, about 0,85. 
The inconvenience of the rapid motion of the needle, occasioned, 
at one part of the voyage, the practice to be discontinued of 
observing every tenth vibration, and every twentieth was substi- 
tuted. This no doubt relieved the perplexity in which the observer 
occasionally found himself, in having to observe, and record, and 
be prepared again to observe, at every twenty seconds or less, and 
so far the change enabled him to observe better. But still, the 
disadvantage remains, in so quick moving a needle, that if a mis- 
take of two vibrations is made, the difference of time occasioned 
is not of so marked and decided a character as to be at all times 
at once distinguished. It is of much more importance that there 
should be no miscount of the vibrations, than that the times should 
be recorded correctly to the fraction of a second. It is only the 
earlier and later times that are finally influential ; but every unde- 
tected error in the number of vibrations falls with its whole weight 
upon the result. 
The occasional discrepancies in the results of the same, or of 
different, observers, or on the same, or on different, days, which 
are seen in the subjoined table, are not, I believe, traceable to the 
source I have been discussing, nor apparently to any other than an 
actual difference in the timeof the cylinder performing its vibration. 
A mean has been taken as the result at each station, except at St. 
Helena, where the discrepancy on the 11th and 13th of July was so 
considerable, that it has been thought more satisfactory to collect 
the observations of each day into separate results. 
The subjoined table comprises the result of each observation, 
