70 
MR. A. W. RUCKER AKD UR. T. E. THORPE OK A MAGKETIC 
There are now five British magnetic observatories wmrking with instruments of 
substantially the same patterns, and we think it would be very desirable if an agree¬ 
ment could be come to as to the form in which the results of the observations are 
published. In our own case we have been anxious to check the numbers given by 
experiment in every way, as it is only thus that we are able to determine what 
reliance may be placed on observations at stations where time or w^eather prevented 
the repetition of one or more of the measurements. 
The following is an example of the method of tabulation adopted in the case of the 
Declinations :— 
S is the interval between the southing of the Sun and the mean of the times at 
which the “front” observations were made on the Sun’s azimuth. It is taken as 
positive if the observations were made after noon. 
Alt. signifies the altitude of the Sun above the horizon at the same time. 
p, is the cori’ection made by means of the “ back” observations to the geographical 
meridian determined from the front observations alone. This quantity serves to 
indicate the order of the error that may have been introduced by the omission of the 
back readings on some occasions. 
G.M. is the reading on the horizontal circle which corresponds to the geographical 
meridian. In most cases two determinations of the geographical and magnetic 
meridians were made at an interval of some hours. In this time the tripod may 
have shifted slightly, owing to unequal sinking of the ground, warping of the legs, &c. 
Hence, if no fixed mark was read, each geographical meridian must be regarded as 
corresponding only to the magnetic meridian which was observed nearly at the 
same time as itself. As it is often difficult to select a suitable fixed mark, we have, 
whether a fixed mark was observed or not, regarded each reading for the geographical 
meridian as corresiaonding to that for the magnetic meridian which was nearest to it 
in point of time. 
In the case of the magnetic meridian (the reading for which is indicated by M.M.) 
we tabulate the following data : 
(1) The G.M.T. at which the observation was made reckoned from midnight, 
(2) Half the difference between the readings when the magnet is erect and inverted 
is the angle between the magnetic and geometrical axes of the magnet. If then ft’om 
this quantity (^) we subtract its mean value as determined from all the observations 
made in the same year (^q), we obtain a measure of the error due to inaccuracy of 
reading or adjustment, and to imperfect compensation of the diurnal variation or of 
any disturbance which may have been occurring at the time. 
(3) oj is the error which would have been caused had the torsion of the thread not 
been allowed for. 
(4) V and A are the deviations of the element from its mean value due to diurnal 
variation and disturbance respectively. The first, as has been already explained, is 
calculated for local time from the Kew curves of diurnal variation, the latter is the 
