Small as the errors of this instrument, as just stated, may appear, contrivances have been suggested and ap plied, which reduce them to almost nothing. To give the horizontal circle this advantage, the top of the pedes tal should be made double, and the upper half move able in azimuth, carrying the instrument round with it, hut capable of being firmly secured in any given posi tion. By this contrivance, an angle in azimuth may be measured on as many different parts of the circle as may be thought necessary. This, in truth, gives to this in strument a property, but little inferior to the far-famed method of repetition. • A contrivance, which furnishes the upper circle with a similar advantage, was suggested by the ingenuity of our present astronomer-royal, as stated in his paper upon the declination of some of the fixed stars, (Philos. Trans. for 1806.) This is effected, by giving the reading microscopes a motion of about 60° concentric with the circle : for the sun and moon, Is hick arc continually changing their meridian altitudes, no such contrivance is necessary, except indeed for the former, at the points of his greatest declination ; but for the fixed stars, which, without it, would always have their places given by the same division, the revolving microscopes are of the greatest value. By changing the places of these as often as we think necessary, the errors of division may be reduced to a quantity too small to deserve notice. In all cases, the angle, through which the microscopes are shifted from their horizontal position, must be applied to the observations as an error of collimation.
We may observe, that this construction of the astro nomical circle seems peculiarly adapted to the conveni ence of a scientific gentleman when travelling. Its strong and compact figure renders it safe and commodious in carriage. All its verifications may be made in a short space of time by land objects ; and its being applicable to all the purposes of astronomy and geodxsy, renders it, ill our estimation, superior to the repeating circle, which is, for similar purposes, its only rival.
If the instrument turn freely in azimuth, and has like wise a divided azimuth circle, this circle should first be placed in a horizontal position, a verification not always absolutely necessary, (except azimuths arc to be observ ed) but which will generally be found convenient. It is performed, by bringing any one of the levels of the in strument into a parallel line with two of the feet screws. The bulb of the level being made to rest in a central po sition by the action of either of these screws, the instru ment is then to be moved half round in azimuth, and if the bulb of the level has changed its place, it is to be restored to its former position, half by the action of one of the feet screws, and the remaining part by the action of the screw, which changes the position of the points of suspension of the level itself. If, upon a second trial,
the level remains unchanged, then the level itself is ad justed, and the instrument is to be now brought into a position perpendicular to the first position. If the level now is no longer in its central position, the whole devia tion is to be corrected by the foot screw; a circumstance particularly to be attended to, for the level once adjust ed is always adjusted, whereas the position of the azi muth circle may be correct in one position, but erroneous in all the rest.
The axis of the vertical circle should next be made horizontal. This is known to be accomplished, when the hanging level remains stationary during a whole revolu tion of the vertical circle round its horizontal axis ; but, in mane instruments, the level is not capable of revolv ing with the circle, in which case it must be applied and reversed end to end, till the bulb remains in the same po sition at each application.
In large circles intended to he u.cd as transit instru ments, the exact verification of the horizontal axis is of great importance. If the lc eI be very sensible, it is usually furnished with a scale. In that case, it is better to read off the deviation cad' way, than to attempt to bring the level to a central position. This verification is performed in the following manner : Suppose the divi sions of the scale to represent seconds of space, and the central division, both or the cast and west end, to be marked zero, then apply the level, and take three or more readings as follows : always observing the transit of a star. On the centre, indeed, the observation must necessarily he so made, when zenith distances and right ascensions are determin ed with the same instrument. What we wish to point out, is the importance of paying the greatest attention to the horizontal wire. If the two wires are not exactly perpendicular to each other, one must be erroneously placed, and this had better happen to the vertical than to the horizontal wire.
The line of collimation should next be made to de scribe a great circle of the sphere. This, when the in strument admits of it, is done by reversing the vertical circle in its rs, and obsening if the same terrestrial ob ject be accurately bisected in each position ; or the in strument may be turned half round by the azimuth cir cle, which is an equivalent operation. In very large in struments between stone piers this cannot often be done. The observer, in that case, must be provided with two accurate meridian marks, one to the north, the other to the south, and when these arc once adjusted, there will be nanecessity to have recourse to this troublesome ope ration. We shall again have to recur to these verifica tions under the article TRANSIT INsTuUMENT.