2 LUNAR EFFECT 



of longitude being neglected. The observation nearest to the local mean solar 

 time of the moon's transit was marked with a zero, signifying 0"' of lunar time. 

 The time of the inferior transit was next obtained; and the observation nearest to 

 it in time was marked 12". The greatest difference in interval between the moon's 

 transit and the time of observation could in no instance exceed half an hour. In 

 the bi-hourly series, the observations nearest the moon's transit, or to either hour 

 angle, one hour before or one hour after the transit was marked. The mean of 

 a number of differences for the same hours thus gave a result corresponding 

 nearly enough with the hour. The number of observations intermediate between 

 those marked 1 '- and 12 h - were marked with the corresponding hour angle by 

 interpolation, care being taken to note the nearest full hour against each observation 

 in the bi-hourly series. The hourly series begins with October, 1843. In the case 

 of thirteen observations within twelve lunar hours, the one nearest midway between 

 the two consecutive lunar hours was omitted. 



In the month of March, 1842, which is selected as an example of the details of 

 working the bi-hourly series, the number of observations available is 298, of which 

 148 correspond to western and 150 to eastern hour angles. In the abstract which 

 follows + indicates a deviation of the north end of the magnet to the west, and 

 — a deviation to the east of the respective normal position for the hour. The 

 hourly normals are given in the first part of the discussion. No difference exceeds 

 eight divisions, this being the limit in number indicated by the criterion. 



