298 Royal Astronomical Society, 



inclined to the vertical, one on the north side, the other on the south 

 side ; and if this double operation be effected at one transit of the 

 star, the result for the star's zenith distance is obtained without any 

 computation of the star's corrections, and without any reliance on 

 the permanence of the state of the instrument for more than a few 

 minutes ; but if the observation with micrometer-head north is made 

 at the transit of one day, and that with micrometer-head south at 

 the transit of another day, it is necessary to compute the change in 

 the star's place (which can be done with undoubted accuracy), and 

 also to be assured that no change has taken place in the relative 

 position of the various parts of the instrument. For the details of 

 the calculations of every kind applicable to these cases, it is best to 

 refer to the introductions to the volumes of Greenwich Observations 

 from 1837. 



For a considerable time, the instrument, not being furnished with 

 two wires on the micrometer-plate at a distance nearly equal to 

 double the star's zenith distance, could not be used for the double 

 observation at a single transit (as the short duration in the field of 

 view did not leave time enough for the numerous turns of the mi- 

 crometer, with change of the observer's position, &c.), and it was 

 therefore reversed after the completion of each day's observation. 

 The results of the observations were not satisfactory, and the Astro- 

 nomer Royal detei'mined on so fitting up the instrument that it 

 might be used for the double observation at a single transit. 



The arrangements obviously necessary were, to fix on the micro- 

 meter-plate two wires at distances nearly equal to double the star's 

 zenith distance, and to determine their interval accuiately ; and to 

 provide stops for the reversion of the instrument, for the movement 

 of the eyepiece, and for the movement of a part of the illuminating 

 apparatus, which it was necessary to shift at each observation. But 

 another change, the necessity for which seems to have been gene- 

 rally overlooked, then suggested itself. It is unsafe to consider any 

 wire as absolutely straight : when it has once received a bend, even 

 no more than the bend of passing round a reel, it retains a portion 

 of that bend although stretched even to the breaking-point ; and 

 some of those curved parts of the wire may be under the micro- 

 scopes by which the position of the plumb-line is checked. Now 

 this is unimportant, provided that the bend is always turned in the 

 same direction relatively to the instrument ; for then its only effect 

 is to alter by a constant quantity the correction to the value of the 

 grand-micrometer reading, which constant disappears on taking the 

 difference of the two corrected grand-micrometer readings, upon 

 which, in fact, the determined zenith distance depends. But if the 

 position of the bend changes relatively to the instrument, the result 

 is affected with error. Now it is extremely probable that the posi- 

 tion of the bend will change ; that is, that the plumb-line will turn 

 relatively to the instrument at every Reversion, but more particularly 

 so at rapid reversion ; and some method must be adopted to prevent 

 this turning. 



For this purpose the Astronomer Royal adopted the following 



