Mr. J. A. BrouQ un Terrestrial Magnetism. 



87 



1731, and diminution thence to four in 1746. The decennial 

 period is certainly not marked in that part of the last century. 



The mean diurnal variation of magnetic declination is, how- 

 ever, a maximum, both by Cassini's and Gilpin's observations, in 

 1787. It is perhaps worth noting the results I have been able 

 to find on this question ; they are as follow : — 



1759* 

 1771-73 



1778-80 

 1784 

 1785 



1786 1 

 1787J 



1788 j 



1789 

 1790 

 1791 



10-7 



7-2 



9-2 



96 



10-8 



13-9 



14-5 



151 



150 



13-7 



14-4 



120 



11-9 



11-3 



1792 

 1793 

 1794 

 1795 

 1796 

 1797 

 1798 

 1799 

 1800 

 1801 

 1802 

 1803 

 1804 



1829 

 1830 

 1831 

 * * 

 1834 

 1835 

 1836 

 1837 

 1838 

 1839 

 1840 

 1841 

 1842 

 1843 

 1844 



13-7 

 12-4 

 11-8 



8-2 



100 



12-9 



123 



122 



110 



9-5 



8-5 



7-6 



7-6 



7-4 



There are evidences in this list of a maximum diurnal oscilla- 

 tion in 1787, 1829 and 1837 : 1847 or 1848 was also an epoch 

 of maximum. The mean period is therefore about 10 years ; 

 the interval from 1829 till 1847 gives about 9 years; but the 

 interval 1787 to 1847, or 60 years, may either give six periods 

 of 10 years, or seven periods of 8-6 years. 



It would appear that the maximum is not always exhibited ; 

 thus it docs not appear at all in the period after 1787 ; on the 

 contrary, there appears a tendency to a maximum in 1803. 

 Again, the maximum in 1829 gives an interval of less than 8 

 years to the maximum in 1836-37 ; the next interval is, however, 

 upwards of 10 years, as the following will probably be found to 

 be. If we were to assume the interval as 885 years, the period 

 of a revolution of the lunar apsides, we should have had the 

 maximum in the following years, counting back from 1848, 

 1839,1830-32,1821-24,1812-15,1803-07,1794-99,1786-90. 

 The exact period is therefore not quite certainf- 



* 1759, Canton, London. 1/71-7^. Van Svvinflen, Francker. 17/8-80, 

 Cotte, Montmorency. 1784-88, Cassini, Paris. 1786-1804, Gilpin, 

 London. 1810-11, Bowdich, Salem, U.S. 1817-20, Beaufoy, London. 

 1820-.'34, Arago, Paris. 18.'35-44, Gauss, Gottingen, and Lamont, Munich. 

 These ranges are not quite comjjarable, as in a few cases the range is that 

 of the mean oscillation, and they should vary slightly at the different places. 

 The ranges from Gilpin's observations are derived from two to twelve 

 months' observations in each year ; but where the observations were not 

 made in all the months of the year, a correction has been applied to 

 obtain the range for the year. 



t It is not a little curious, that the observations of M. Coulvier Gravier 



