364 
ME. C. CHAMBEES ON THE SOLAE YAEIATIONS OE 
the mahogany box, the latter has two small windows of flat glass suitably inserted in its 
curved side, and a lamp is placed on a stool to the northward of the masonry pillar which 
supports the declinometer, and is kept constantly burning. The joints and crevices, and. 
for greater security, the whole outside surface of the cylindrical box was covered with 
paper to keep out small spiders or other insects. 
3. The observations are made by a transit instrument (30 inches long and with pivots 
15 inches apart) which is fixed on an isolated pillar of masonry, at a distance of 27 feet to 
the southward of the declinometer pillar, and which in this position is capable of being 
used also for observing the transits of stars through an opening (provided with shutters) 
in the roof of the observatory. This instrument was adjusted in a line with the centres 
of the scale and lens attached to the magnet, and its basement was then firmly imbedded 
in masonry, and the telescope was focused so as to give a distinct view of the scale. A 
vertical spider-line is placed in the common principal focus of the object-glass and eye- 
piece, and the instrument was adjusted so that an object on the true meridian should be 
cut by this line. With these arrangements any change in the azimuth of the freely 
suspended magnet is indicated by a change of corresponding magnitude in the position 
of the image of the collimator scale, as seen in the field of view of the telescope ; and 
the actual position of the magnet at any moment may be found by noting the scale- 
division that is cut by the reading-wire of the telescope, if only we know the position 
corresponding to any one reading of the scale, and the change of position for a change 
of unity in the reading. The latter constant, the scale-coefficient, is easily determined 
by substituting an azimuth instrument in the place of the transit instrument, and observ- 
ing with it the horizontal angle included between any two widely different divisions of 
the scale when the magnet is at rest. Experiments recorded on pages x and xi of the 
Introduction to the Bombay Magnetical Observations for 1864 showed that the scale 
was very correctly placed in the principal focus of its lens, and that the scale-coefficient 
is 6'*806 ; but the actual value of the coefficient used in the reduction of the observa- 
tions was 6 ,- 841, that having been the result of a first determination made in 1865, 
when the scale and lens were apparently differently adjusted with respect to each other. 
No correction has, however, been applied to the reduced observations for the present 
purpose, since the laws of periodical variations are exhibited with equal clearness what- 
ever the unit with which they are expressed ; and it will only be necessary to bear in 
mind what is the real unit made use of if we should wish to compare the variations of 
this element of the terrestrial magnetic force at Bombay with those of either or both 
of the two others, or with the results of observations made elsewhere, so that all may be 
reduced to a common unit as a preliminary step*. 
4. The direction of the magnetic axis of the bar corresponding to a particular scale- 
reading was found in the following manner : — 
First, the level and collimation errors of the transit instrument were approximately 
destroyed, and the telescope was then pointed upon Polaris when on the true meridian, 
* It has not been deemed requisite to do this for any of the discussions that follow. 
