DECLINATION. 
nt revolves freely in 
azimath, then having obferved feveral. ars, in the manner 
d, the inftrument fhould be turned half round in 
ezimuth, fd the ean line oes bife 
on no account be handled m 
for this aoa 
firft een fhould now be obferved again in precifely ied 
fame man but the microfcopes will not indicate the fam 
quantities. Thi in the fir silepa they gave zenith diftaaces, 
th now give altitu » 1f one mifcrofcope is made 
to oe the zenith diftance, and ‘the other the altitude, they 
will be now reverfed, and the-one which gave oa aevg 
will now give altitude When the ftars have been 8 ob- 
of it. For in 
ing the above cbfervations, he fhould begin by ar 
the index error, or error of collimation, from each ftar fepa- 
nae 3 and the difference of the quantities thus found fhou Id, 
if the infttument and obfervations are good, be contained in 
nee narrow limits. 
alf the extreme difference will be very nearly the greateft 
error to which the obfervations st be liable. 
Upon this fubje&t, vid. Ph. Tra 
Rh illuftrate this, we fhall take re example fome obfervae 
ns made with an aftronomical circle of two feet diamet eter, 
confrcted by Mr. Troughton, and which moved freely in in. 
uth. 
Altitude. . 
6th Sept. Rigel, Pof. E. 50° ax" aoe & 
ath: E. © 21 40 
Mean, 30 21 38.1 
Sth Pof.W. 30 20 22.5 
oth W. 30 20 23.5 
Mean, 30 20 23. 
Difference between E.and W.. o 1 15.12 
Half, or error of Cal ll. O © 37.5 
True alti tude, 30 23 00.5 
Sirius, W. 22 20 40 
. £E 22 20 
o 1 21 
Error of Col. o © 40%,5 
True Altitude- 22 20 20.5 
2 Sagitte. W. 
: oO F 21.3 
Error of Coll. © 0 40.6 
True altitude 8 57 26.4 
x Ww. 31° 18’ 54" 
E. 31 20 15 «7 
Oo 4§ 24 
Error of Coll, © © 42 
True altitude 31 19 36.7 
« Cap. WwW. 25 38 417 
i. 29 40 00.6 
o 1 18.9 
‘Error of Coll. cae) 39-45 
True altitude 25 39 21.15 
9 Ww. 18 9 39.25 
EE. i 10 54. 5 
oO F 14.25 
Error of Coll. © © 37.12 
True altitude © 10 16.39 
et 
a 
21 13 53 
2115 7.7 
Oo I 14.7 
© 0 37-35 
y Ww. 
E. 
Error of Coll. 
True altitude 21 14 30.35 
3 Ww. 2 45 39.75 
E. 0 460 55. 5 
115 7} 5 
Error of Coll. ~ © 0 3787 
21 46 17.6 
Enror of Coll. Number of Obf, 
Rigel 37".§ 6 
Sirius 40. 5 . 3 
2? Sagitte . 40. 6 38 
x 42. 2k 
w« Cap, 39-45 3. 
- 6 . 37-12 . 2 
y . . 37:35 2 
Py : , 37.87 . Z- 
Mean Error of Collimation 39. 0 
From this it fhould appear that 3” was probably the 
reatett jis committed m-the determination of the alti- 
tudes of tthe above ftars, including errors of eve d 
Had fe cil wate been continued, the difcordance would 
have been lefs, as the errors of obfervation would have been 
eb 
pies lt done awey, and the errors of divifion only would 
have aed, 
Ccz. “The 
