and of the Altitude and Azimuth Instrument. 107 



bolh to the north and south, 240" of the circle are employed. 

 To extend it further coidd produce no good effect, for the un- 

 certainty of refraction would counteract the advantage arising 

 from a greater range. In using A, if an observation is liurt 

 by a passing cloud, or any otiier circumstance that may ren- 

 der it doubtful, it had better not be written down, even 

 should it agree with the general result; and, on the other 

 hand, if the observer is satisfied with an observation, it shoidd 

 not be omitted for being somewhat discordant. \^^ith R, if 

 any one of the steps of a series of repetition is not well per- 

 formed, the whole will be vitiated should the observer go on. 

 He may perhaps be able to save what lias gone before, by 

 reading off what has been done, and beginning a new set ; but 

 this cannot be done in all cases. I remember an instance, 

 whei'e, in a series of twenty double observations on pularis (that 

 had been made vvith extraordinary care), by the mistake of 

 the assistant, in the very last of them, by his turning the screw 

 of the indices instead of that of the level, the whole labour was 

 irretrievably lost. 



When an observer is so circumstanced as not to be able to 

 stiiy at a place more than a short time, R seems to have the 

 advantage ; for, in a few hours a series of repetitions may be 

 made on polaris to answer his purpose ; since the polar di- 

 stance of that star is now so well known, that little will be lost 

 by not having opportunity of observing it botli above and be- 

 low the ])ole. A is not, lioweA'er, without its resources in this 

 I'espect ; for in stars that have not more than 25" of polar di- 

 stance, there will be (juite time enough for reversing the posi- 

 tion of the uistrument, and obtaining coUimated results. I 

 nmst here remark that this can only be done in high latitudes, 

 because near the equator, the slow moving stars are too much 

 affected by variable refi-action to be i"elied on in nice matters. 

 This last remark must be placed very much to the discredit 

 of R, which in its nature is adapted to the slow moving stars 

 alone. 



Thus it has been shown that, respecting its most effective 

 operations, the repealing circle is local witli regard to both the 

 heavens and the earth. It must be granted that at every step, 

 even when a cjuick moving star is observed, if tlie time has 

 been noted, the wliole niay be reduced to the meridian : and 

 were it wanted, which it is not, I should claim the same con- 

 cession in favour of A. But let it be remembered, that at any 

 considei-able distance from the meridian, the time becomes so 

 important a datum, that it refjuircs an exactness, scarcely to 

 be found, except in a well regulated observatory. Therefore, 

 in oliserving stars near the t(|uator witli R, tlie observer has 

 liic choice of twodilJitulties ; In; nuiy either extend his obser- 



O 2 vations 



