or THE MAGNETIC DECLINATION. 
509 
One glance at this Table informs us of the great difference between the several curves 
for the two years. The yearly curve for 1860 shows a maximum easterly declination at 
23 h and a minimum at 6 h , with a secondary maximum at ll h and a range of 32'-39 ; 
whilst the same curve for 1861 shows a maximum at the moon’s lower transit, and a 
rather indistinctly expressed minimum at 3 h , the greatest range being 38"'45. Differ- 
ences, similar to those just pointed out, will be found on examining the various curves, 
but it may suffice for the present to single out one of the most striking anomalies, if 
we are allowed to speak, of such, our knowledge on this point being still very indistinct. 
I refer to the winter half-yearly curve when the moon’s declination is indiscriminately 
north and south. In 1860 we notice in this case a distinct maximum at 22 h and a 
minimum at 8 h , secondary extremes occurring at ll h and 13 h . The greatest range is 
28"-23. But on examining the same curve for 1861 we find a vast difference; a maxi- 
mum deviation towards the east at the moon’s inferior passage and at 3 b , whereas it is 
scarcely possible to speak of secondary extremes, all other undulations appearing as acci- 
dental irregularities when compared with the main oscillation, with a range amounting 
to 43' 7, 82. It can hardly escape our observation that both curves show a clear sweep, 
but in a very different sense ; and we understand now the reason of the irregularity of 
the mean winter curve for the five years, which scarcely deserves the name of a curve. 
This fact frequently induced those more especially engaged in these pursuits to doubt 
the existence of a lunar-diurnal variation-curve during the absence of the sun from the 
hemisphere of observation, a notion which, after the above explanations, may safely be 
pronounced to be erroneous. We feel, however, also inclined to recognize in the facts 
above pointed out the excuse for such a conclusion. 
The summer semiannual curve for 1861 exhibits a secondary maximum and minimum, 
respectively, at 23 h and 20 h , the primary extremes occurring at the moon’s lower passage 
and at 6 h . As all curves for this year exhibit similar anomalies, when compared with 
what we adopted for the rule, the idea suggested itself that some error might have 
crept into the discussion ; but this was soon proved to be an erroneous supposition by a 
perfectly independent and fresh discussion which gave in the main points results iden- 
tical with those arrived at on the first occasion. 
In course of the year 1861, the instruments hitherto in use at the Flagstaff Obser- 
vatory, were replaced by new ones just received from Munich, and in the month of May 
the necessary adjustments were so far advanced as to allow of the registration of the 
new instruments being commenced with the beginning of the month of June. It need 
scarcely to be mentioned that the greatest care was taken to ensure uniformity of regi- 
stration in every respect, and it is not likely that the new arrangements would have 
influenced the observation in any undue manner. Not satisfied, however, with such 
guarantees for an exact observation and discussion, I resolved to examine the anomalies 
of this year still further. For this purpose the lunar-diurnal variation for the period from 
May 1860 to April 1861 was derived with special regard to the moon’s declination, as 
likewise for the same months in 1858 and 1859, and in comparing the respective results 
