EARTH’S MAGNETISM AT PAVLOVSK AND POLA (1897-1903). 
303 
time of upper culmination of the moon at the epoch of new moon, and the a, b and 
C, 6 coefficients should both be given. 
The Lunar Diurnal Magnetic Variations at Pavlovsk and Pola. 
§ 7. In undertaking the present investigation it was necessary to limit, as far as 
possible, the quantity of observations dealt with, and this made it desirable to choose 
a period of years which were magnetically “ quiet,” in order to minimise the irregular 
deviations of the magnetic needle which have to be averaged out in determining the 
lunar diurnal variations. The period adopted was that from 1897 to 1903 inclusive, 
comprising seven years during the sunspot minimum immediately preceding the 
present one. 
The choice of observatories to be first dealt with fell on Pavlovsk and Pola ; both 
of these publish the hourly values of the magnetic elements, and their latitudes 
rendered them suitable for the main purpose of this work. The prior examination of 
these rather than of other observations was due to the greater novelty of results 
relating to the lunar magnetic variation in fairly high latitudes ; the phenomenon has 
been studied in detail only at tropical observatories hitherto. 
The following particulars concerning Pavlovsk and Pola are important for our 
purposes :— 
Pavlovsk. Pola. 
Latitude. 
Longitude .... 
Declination (1900) . 
Horizontal Force (1900) 
59° 41' 13" N. 
30 29 15 E. 
0° 37' E. 
16550y. 
44° 51' 49" N. 
13 50 46 E. 
9° 25' W. 
22210y. 
The correction to the epochs of the harmonic components of the variations, 
described in § 6, is therefore +2°‘l in the case of Pavlovsk, and + 1°'0 in the case of 
Pola. 
§ 8. The results from these two observatories are exhibited in Tables I. to YI. 
Tables Ia. and Ib. give a summary of the. semi-diurnal components of the lunar diurnal 
magnetic variation for the three elements and three seasons (as well as for the mean 
of the whole year), in the form of hourly inequalities. These are obtained from the 
mean of a number of whole lunations, so that any component of period other than 
half a lunar day is to be regarded as merely accidental. As may be readily seen, the 
results exhibit very clearly in most cases the true semi-diurnal character of the 
constant term in the lunar variation. 
The inequalities are given for twenty-five times, separated by intervals of one solar 
hour, there being twenty-five solar hours in a lunar day. The small correction to the 
epoch, mentioned in § 6, has not been applied, as this table is given only to illustrate 
the main feature of the lunar diurnal variation, i.e., its semi-diurnal character. It will 
be noticed that the lunar variation is inverted during the winter months in the case 
of declination, and partially so in the case of horizontal force. 
