and of the Altitude and Azimuth Instruments 13 



microscopes that read and subdivide the divisions of the circle, 

 but also supports the whole of the upper works. On opposite 

 sides of the cone, and distant from it about half tlie radius of 

 the plate, are erected two columns, of a height to support a 

 transit axis, so as to allow the telescope to pass the upper end 

 of the vertical axis, when it is pointed towards the zenith. The 

 transit axis is one-thlid longer than the distance between the 

 columns, upon which out-riggers are placed, having Y's or 

 angles on their extremities, that suj^port the axis : and each 

 of the angles is acted upon by an adjusting screw, not only for 

 making the transit axis horizontal, but also for placing the 

 centre of the circle of the same height with the horizontal mi- 

 croscopes. The horizontal axis is crossed, as m an ordinary 

 transit, by a telescope ; the length of wliich exceeds the dia- 

 meter of the circle by about one-third. The circle framed 

 upon the axis is double ; the two parts being placed at a di- 

 stance from each other to allow the telescope a lodging be- 

 tween them: and they are connected with each other by pil- 

 lars inserted perpendicularly between them. The front por- 

 tion of the circle (or that which bears the division) is of a less 

 radius than the other by about one-eighth part of an inch: the 

 longer radius of the portion behind is what is required for the 

 clamp and screw for slow motion to act upon, while the shorter 

 radius of the one in front keeps it clear of that apparatus when 

 the axis is reversed lor collimation. Many of these instru- 

 ments have been constructed with vernier readings ; but as I 

 consider those by the microscopic micrometer preferable, I 

 shall confine this description to the latter. My preference to 

 one of those excellent contrivances for minute subdivision is 

 mainly grounded on the circumstance that, in the emj^loyment 

 of the more ancient method, the indices rub against the divi- 

 sions which they subdivide ; whereas in the modern, which is 

 detached, the motion is free and unembarrassed. 



In the description of the repeating circle, the advantage of 

 three readings was stilted ; but that contrivance originated with 

 the instrument; I am now describing: and if it be a real im- 

 provement (wliich I believe no one will doubt), the repeating 

 circle owes the advantage solely to the latter. Three readings 

 are not only better than two, but also better than four: for, 

 widi four, when the objects are in the horizon, or near it, on 

 reversion the opjjosite indices only chanj^e places ; a circum- 

 stance clearly in favour of the odd numuer. But, in astro-* 

 nomy, where the upper circle is chicHy concerned, the same 

 a^lvantage does not occur: for, at the zenith, on reversion, the 

 telescope changes place in azimuth only: therefore, as the 

 indices have no change of place, more readings than two couM 



be 



