MAGNETIC DECLINATION AT BOMBAY. 
373 
part of the day ; such curves are the easterly for Kew, St. Helena, Toronto, and the 
Cape of Good Hope, and the westerly for Hobarton, St. Helena, Pekin, and Nertschinsk. 
The correspondence as to shape does not, however, hold as to time, either absolute or 
local, for the hours of maxima are not confined to any particular portion of the day. 
The Bombay astronomical hours of little or no westerly disturbance are from 6 to 17, 
the considerable and regular variation from hour to hour occurring during the remaining 
hours. The principal features of the easterly disturbance- diurnal variation curve for 
Bombay are like those of the westerly curve, with this distinction, that there is a consi- 
derable amount of easterly disturbance at all hours of the day, the least hourly ratio 
being 062 ; the easterly curve is also less remarkably regular than the westerly. There 
is great similarity in the declination-disturbance diurnal variations of St. Helena and 
Bombay, and the principal inflections of the curves occur at the same local hours ; but 
the night hours at St. Helena are, unlike the same at Bombay, almost entirely free 
from easterly disturbance. In the disturbance- annual variations there is, as at Toronto, 
a tendency, though not very decided, towards maxima at the equinoxes and minima at 
the solstices. Having regard to the note at foot of Table VIII., it is seen that the 
decennial variation of aggregate disturbance is distinctly indicated, as far as seven years 
are capable of showing it, both in the easterly, westerly, and combined disturbances, the 
minimum occurring in 1864 and the maximum about the end of 1859. The aggregate 
of disturbances in the latter year exceeds that in the former in the proportion of 2-5 to 1. 
The disturbance- diurnal variation curve shows a preponderance of easterly disturbance 
at all hours except 20 h and 21 h ; from 5 h to 19 h the preponderance is great, but is less 
considerable during the remaining hours : on the whole the proportion of aggregate 
easterly disturbance to westerly is as 1*6 to 1. 
13. The regular Solar-diurnal Variation . — After the separation of disturbed obser- 
vations differing from the final normals by P'4, the averages were taken of the monthly 
final normals of each hour — in the first place for every month, and in the second place 
for each of the fourteen half years ; and by means of these averages were calculated the 
diurnal variations * for each month (on an average of seven years), for the whole year, 
and for each individual half year from 1859 to 1865, the one half year comprising the 
months April to September, and the other the months January to March and October 
to December. The diurnal variations are shown in the following Tables. 
* By diurnal or annual variations is always to be understood the series of hourly or monthly values expressed 
by the excess of each above the mean value for the whole day or for the whole year respectively. 
