The first of the principal adjustments is that of setting the line of sight at right angles to the cross-axis, when it necessarily describes a great circle. A distinct object must be selected not far from the hori zon, and bisected by the middle wire, using the azimuth-screw. The axis is then carefully lifted out of its Y's, and returned end for end, or rerersed, and the object viewed again. It is now to be bLiected as before, half by moving the azimuth-screw, and half by the screw at / and its antagonist, each of which draws the plate on which the wires are fixed. The operation must be repeated until no difference can bo seen in the bisection, whichever be the position of the axis. When there is a micrometer, the operation is somewhat easier. The micro meter wire is brought first on one side and then on the other of the centre wire, eo as just to shut out the light between them, and the mean of these two readings is the reading of the zero point, that which corresponds to the exact superposition of the wires. The distinct object is bisected by the micrometer wire; and we will suppose that the screw has to be turned three revolutions for this purpose from its zero position. Now reverse the axis, and suppose the micrometer has only to be moved two revolutions from its zero to bisection. The half-way, or two revolutions and a half, is the distance which should be shown. Carry the micrometer wire two revolutions and a half back from its last position, and that ie the position which the centre wire should occupy. If the joint thicknesses of the two wires (previously determined) equal ten parts, carry the micrometer five parts still far ther back, and bring up the middle wire to touch it by the screws at 1 as before. The error of collimation is now corrected, as will be found by repeating the observation. The collimating screws have generally a capatan-head, which is awkward, and indeed dangerous so near the eye ; and we think the square head, which is turned by a key, is more manageable. The screws should be drawn tight, but not forced. We have here followed the usual rule of directing the observer to select a well-defined object near the horizon. If the pivots of the axis are equal, the mark may be considerably elevated or depressed without introducing any error ; and an object which is not very distant may be seen sharply defined when the aperture of the object-glass is diminished. In fixed observatories there is usually either a meridian mark north and south at a considerable distance, or two near marks, which are made distinct by interposing lenses. At Greenwich and Cambridge a south meridian mark is combined with a supplementary transit in the north slit ; at Oxford two near marks are made visible by lenses fixed in the observatory slits. By using north and south marks at the same time, you are warned of any alteration in the Y's during the reversal, which, when the instrument is heavy, is always to be feared. Such shifting would cause the error of collimation to appear different according as it is taken from one or the other mark. In standard observatories the error of collimation is not actually corrected unless it is large, but measured by the micrometer, as we have described, and the effect thus allowed for by calculation.
When the illuminated end is west, and the telescope pointing south, let the middle wire appear to be 3.263 Revolutions to the right-hand of the collimating mark : as the telescope reverses, this means that the transit points 3263 It to the left of the mark, or to the east. Now reverse the telescope, the illuminated end is east, and suppose the middle wire still to appear to the right of the mark, but only 3.187 It. These observations are best made on a calm cloudy day ; often, after rain, the mark, if distant, is seen sharply and steadily. If there were no error of collimation, the micrometer should give the same quantity at both observations, or (3.263 R. + 3.187 It ), that is, 3'225 It. The error in collimation, therefore, is R., and the instrument points to the east that quantity when the illuminated end is west. The scale of the micrometer screw is known. Let one revolution cor respond to the space an equatorial star would move over in 3• of time ; then 0'038 It. is the space which an equatorial star would move over in 0"114, which is set down as the error to be employed in calculation.
Now the effect of an error of collimation is to make the instrument describe a parallel to a great circle, and distant from it by the collima tion error. If the instrument points to the east of the south, it also points the same quantity to- the east of the north, and a star above pole will pass the apportnt before the true meridian. The correction, therefore, to Ise applied to the observed transit is and when the instrument is reversed will be —. The time which a star takes to pass between the two circles, or the correction, is equal to the error of collimation or, in the supposition we have made, is— cos. declin. of star * + 0.114 sec. declin. Ill. end West. — 0914 see. declin. Ill. end east.
The sign changes for stars sub polo.
Recently a method has been devised for recording transits, founded upon the principles of electromagnetism. The sidereal clock is made to break the electric circuit at regular intervals of one second, and by a process similar to that employed in Morse's telegraph, the effect is impressed on a recording apparatus in connection with the clock. The observer is also enabled to break the circuit at any instant between two successive beats of the clock, and to record the fact in its proper place upon the registering paper. Hence in determining the transit of a star, the observer breaks the circuit at the instant of the star's passing each successive wire, and the results are imprinted in their proper place upon the recording apparatus. The distance between one of such recorded results and the nearest second as imprinted by the regular break-circuit apparatus of the clock, will indicate the fractional part of a second corresponding to the instant of the star's passage of the wire. In this method the eye and the sense of touch are called into operation instead of the eye and ear, as in the usual method. It originated in the United States of America, where it has been practised since the year 1849. It has been subsequently adopted at the Royal Observatory, Greenwich, and also at Altona, on the Continent. The differences depending on personal equation are almost annihilated by this method, which is well adapted for recording the transits of a great number of stars within a short compass of time.
The next adjustment is to make the axis horizontal. If the poles of motion are in the horizon, the great circle which the instrument, freed from collimation, describes, must pass through the zenith. Put on the level, and bring the bubble into the middle. Now rock it a little to and from the observer, and see whether the bubble still remains in the same place. If, in pushing the level from you, the bubble runs towards the left hand, this shows that the level tube itself is set askew upon its support, and that the left end, being nearest the observer, is elevated by that motion above the right end. Screw the small screw seen at m, and release its antagonist (these are pushing screws) until a considerable rocking motion scarcely moves the bubble at all. There is generally a cross level, to show when the principal level is upright ; and this - should be brought to have its bubble in the middle, when the principal level has been adjusted as above. The fork in the present example serves the same purpose when the level itself has been carefully adjusted. To level the axis, bring the bubble of the level to the same reading at each end (the numeration of the division is supposed to begin from the centre) by the elevating screw at the left hand Y. Reverse the level and bring the ends again to the same reading, half by the elevating screw of the v, and half by the two screws seen at a, which raise or depress the level tube in its supports. On returning the level to its first position, the bubble should still be in the centre ; but if not, it must be there, half by the 't elevating screw and half by its peculiar screws ; and the operation must be repeated till this is effected, that is, if the observer cannot or will not calculate the effect of a small error, which may easily be measured. If he can (and there are few observers at present who cannot), the process is pretty much as follows ; the graduation being supposed to be units each equal to 15" : Illuminated end West. Telescope South. Altitude 45°. Observer North. Level. East end of bubble, 4.1d. West end, 52.84.