366 
ME. C. CHAMBEES ON THE SOLAE VAEIATIONS OE 
declination records have been printed in the several volumes of ‘ Bombay Magnetical 
Observations,’ form the fundamental data for the following discussion*. 
7. It ought to be mentioned that the sliding tube of the transit instrument with 
which the declinometer was read was rather loosely fitted in its socket, a fault which 
might cause sensible change in the collimation-error whenever the focusing adjustment 
was used ; now the instrument was once every month turned upon a divided scale fixed 
upon the north wall of the magnetometer-room and adjusted for distinct vision, after 
which the adjustment was restored for the collimator scale of the magnetometer. It is 
suspected that small but important irregularities in the observations, which will be pointed 
out later, are due chiefly to this fault. 
8. Throughout the periods January 1, 1859 to July 2, 1861, and November 13, 1861 
to December 31, 1865, the adjustments of the instrument were undisturbed ; but between 
July 3 and November 12, 1861 the adjustments were several times renewed, and the 
action of the instrument was at times so unsatisfactory that the observations of July 3 
to 8, August 7 to 12, and October 28 to November 12 are considered unfit for incor- 
poration with the general body. The following statement shows the several periods 
made use of, the observed and adopted values of the true meridian reading, and the 
values of the torsion-coefficient as experimentally determined : — 
Periods. 
True meridian reading. 
Torsion - 
coefficient. 
Observed. 
Adopted. 
January 1, 1859 to July 2, 1861 
28-3064 
28-379 
1-003009 
July 9, 1861 to August 6, 1861 
28-5334 
31153 
1-00039 
August 18, 1861 to October 27, 1861 
28-2515 
28-2515 
1-00088 
November 13, 1861 to December 31, 1865 
28-2733 
28-2733 
1-0028 
No reason is assigned in the records for the adoption for the first two periods of a 
different true-meridian reading to that which was observed ; but it may be presumed 
that in the latter case the intention was to produce that accordance, which doubtless 
existed in fact, between the values of declination in that and the preceding and following 
periods. Any periodical variations which the observations may, however, exhibit will 
be unaffected by the meridian reading, so long as none of the observations of any single 
period involved belong to more than one series, or so long as each series of observations 
is treated independently of the others ; and it is only when combining the different series 
together for the elimination of secular variation and for the evaluation of absolute decli- 
nation at a given epoch, that regard must be paid to the actual disconnexion of the 
several series. In the investigations of solar and lunar diurnal variations which follow 
or are in progress, no incomplete days, nor any comprising observations from different series 
have been made use of. 
* I should have preferred to operate directly upon the observed quantities (/), but was led to believe, at the 
commencement of the work, that part of the original registers had been destroyed ; I have since found that this 
was incorrect. 
