6 



AMPLITUDE OF THE SOLAR-DIURNAL VARIATION 



In general during the year 1841 the readings are more changeable than during 

 the following years. 



The rearrangement of the instruments, and consequent shifting of the index of 

 the scale, alluded to in the preceding notes, interrupted the observations between 

 January 1 and January 12. 



The normal for October, 14 h , was obtained by comparing with the half monthly 

 means and taking the mean of the two results as in a similar case for the month 

 of September of the previous year. 



1 The suspension threads of the declinometer gave way on the 9th of August, and again on the 1 Oth 

 of January, 1844, but after readjusting the instruments, the magnet returned almost exactly to its former 

 reading a mean of the two changes gave as a correction +18.1 divisions, which was accordingly added 

 to all the readings of the year after August 9th, 21 hours. 



3 The correction to refer the mean of the nine last months to the mean of all the months is derived 

 from the readings of the preceding year, as being more uniform in character than those for the year 

 following. 



