Royal Astronomical Society. 297 



depend, were thin vertical arches attached to the bottom plate only 

 by small thumb-screws, and these arches sustained the force of the 

 micrometer-screw sideways. Three arrangements more entirely 

 opposed to the just principles of construction of micrometers can 

 scarcely be conceived. The wire-plate ought not to be strained 

 sideways by a projecting weight, and ought not to carry any weiglit 

 which will increase the friction of its movement ; the action of the 

 screw ought to be in the plane of motion of the wire-plate ; and the 

 supporting point, or appui, of the screw ought to be perfectly firm. 



The following changes were therefore made without delay. The 

 long screw was laid aside, and in its place was used a micrometer- 

 screw or stalk attached to the end of the wire-plate in the usual" 

 way ; and the micrometer was made in every respect like an ordi- 

 nary micrometer, the micrometer-head carrying the concave- screw 

 which embraces the stalk and draws it towards the micrometer- 

 head, the action of the screw being in the plane of the wire-plate 

 and directed longitudinally across the middle of the wire, and the 

 drawing action of the screw being resisted by springs at the oppo- 

 site end of the wire-plate. The motion of the micrometer was 

 limited in this construction to about one-tenth of the whole breadth 

 of the end-plate of the telescope ; but this was far more than was 

 required to measure the zenith distance of y Draconis ; it was, 

 however, made subservient to the measure of nearly the whole 

 breadth in the following way : — On the fixed plate of the micro- 

 meter ten crosses of wires were fastened, as nearly as possible at 

 equal distances, and on the moveable plate were fixed eleven wires 

 at nearly the same distances. By means of the crosses on the fixed 

 plate and the micrometer-movement, the interval from each wire on 

 the moveable plate to the next wire was ascertained, in terms of the 

 micrometer-screw, with great accuracy, and therefore the aggregate 

 or distance of extreme wires was very well known. The intervals 

 were also known in seconds of arc, by observations of the transits 

 of stars when the instrument was turned to a position distant 90° 

 in azimuth from its usual position. Thus the value of the screw, 

 and the elements for making an observation in any part of the whole 

 range available, were completely obtained. With this construction 

 of micrometer, the appui of the micrometer-head was almost close to 

 the bottom plate, and was perfectly firm. The eyepiece was carried 

 by another sliding- plate, moving in grooves unconnected with the 

 micrometer ; and it was moved by a separate rack and pinion. It is 

 presumed that in this form the grand micrometer was perfectly 

 trustworthy. 



In the use of this instrument it is evidently necessary to observe 

 the same star successively in reversed positions of the instrument, the 

 micrometer-head being on the north side in one observation and on 

 the south side in the other observation ; and the values of the 

 grand-micrometer reading must be corrected in each observation by 

 quantities depending on the two micrometer-readings of the plumb- 

 line, in order to obtain for the two observations the angular distances 

 of the star from definite (though imaginary) lines in space, equally 



