Ordnance Survey Astronomical Observations. 359 



perusal of their record of the past, would seem to suggest, — 

 if we rightly understand the numerous circumstances therein 

 detailed, — the propriety of a few remarks for the future, 

 something as follows : — 

 First, — Of the instrument. 



The one in question, though a great advance on all previous 

 zenith sectors, still was a zenith sector, and was affected by 

 some of their natural errors, — errors which have driven them 

 out of all astronomical observatories, and have occasioned in 

 this example, the necessity of applying a very ugly and rather 

 arbitrary correction to all the results, depending on the ex- 

 pansion of the arc by being squeezed out under the vertical 

 pressure of the pivot screw at the top of the instrument; which 

 screw at each station was turned different quantities to make 

 more or less squeezing, according to the judgment or memory 

 of the observer. The horizontal axis of the telescope, more- 

 over, was very short, and being unbalanced, tended to wear 

 unequally ; while the levels, being all on one side of the ver- 

 tical axis, would all be similarly affected by centrifugal mo- 

 tion in turning the instrument round, and might all, there- 

 fore, shew too great or too small a reading, by the amount of 

 the retardation of the bubble by friction. 



We should be inclined, indeed, in the present day, to re- 

 form sectors and levels altogether, by having a transit circle, 

 and obtaining the zero point with the collimating eyepiece. 



A single observation at each meridian transit, would then 

 give a complete result, and with much greater ease and pro- 

 bable accuracy, than when two observations have to be taken, 

 and a large instrument reversed and reset all in a few seconds, 

 with the stars, too, already beginning to describe downward 

 paths, by reason of their distance from the meridian. 



The prime-vertical transit instrument, has become a 

 favourite for these purposes on the Continent ; but although 

 it may give very concordant and apparently accurate results, 

 yet in the shape in which it is there manufactured, it must 

 be liable to so much flexure, and in a vital direction to the 

 integrity of the observations, that their absolute accuracy 

 may be always doubted. 



In the form of the ordinary transit, i. e., with a double axis, 



