10 
Description of an improved 
[Jan. 
verniers to the zero ; and by adding the angles, so taken, to each other, and dividing 
them bv two, the correct angle will be obtained. . . 
There are two verniers on each index, which read from the centre division, ac- 
cording to the arc upon which the angle is taken, and the one may be read hue 
ward as a check upon the other. . , c „• „, 1 i I1T „„„ 
There is another and a greater advantage to be derived from the circular con- 
struction of the instrument, which is the power of repeatmg the observation, and 
thereby correcting the errors arising from the eccentricity ot the centeung or d 
V,f In g the instrument here described, there is a very considerable improvement for 
facilitating the operation of repeating the observations ; it principally consists of the 
two sets of figures, one upon the inner circle, and the other upon the (inter or divi- 
ded circle ; the. first set are for reading off the angle, which is taken with the index 
B ; and the second for reading off the number of the degrees, &c. which are to be 
divided by the number of repetitions. 
The method of observing with the instrument is as follows : — 
If it is to be used as a'sextant, the vernier A, which is attached to the inner 
circle, must be set to a primary division, (it will be the most correct to set it to 
720,) for which purpose there is, as a rough guide to the figures, aline upon the out- 
er chamfered edge of the vernier A, which corresponds with the zero of that ver- 
nier. Having adjusted this vernier correctly to 720, the angle must be taken by 
moving tlic index B forward, in the same manner as w ith the sextant, and the an- 
gle contained between the two objects will be seen upon the circle by the vernier 
belonging to that index, and the figures on the inner circle will be the guide for 
-reading off. The index error should, in this case, be found as with the sextant, by 
taking a small angle on each side the zero. If there should be time for a second 
observation, the same angle may betaken with the face of the instrument reversed 
by moving the index B, the contrary way to the former, or upon the off arc ; in this 
case the index error need not be attended to, as the half of the two angles will be 
the correct angle. 
The circle may also be used for taking a series of observations, by setting the 
vernier A to 720, and proceeding, in the first instance, by moving the index B for- 
ward, and then the inner circle, telescope, &c. and so on, each time reading off the 
angle, or noting the number of observations ; and dividing the amount of the 
degrees, &c. passed through by the number of the observations ; and if the number 
of the angles taken are equal, or nearly so, to the whole circle, it is evident that the 
errors of division and entering will have been taken into the account: — the index 
error must, in this case, be noticed. 
When the instrument is used as a repeating instrument, the vernier A should be 
very correctly adjusted to 720 ; and the index B moved forward until the tw o ob- 
jects, the angle of which is to be taken, are brought into contact ; read off, and 
note the amount, then leaving the index B fast, and unclamping the telescope, &c. 
move it through the space equal to twice the angle, or until the figures on the inner 
circle shew the amount of the first angle ; on the contrary side of their zero, bring 
the objects into contact, and it will here be evident that the space passed over will be 
equal to twice the angle. Note down two, and without disturbing the last adjust- 
ment, move the index B forward twice the angle, and bring the objects a^ain into 
contact ; then move the telescope, &o. forw ard as before, and note down tw'O in 
addition to the former. In this mode of observing, if the face of the instrument is 
continued always one way, it will be requisite to present the telescope, first to the 
left band object, and secondly, to the right ; but if the face is reversed alternately 
the telescope should be presented to the same object. Proceed with the observa- 
tions, and each time the telescope, &c. is moved forward, note two ; and when the 
vernier A has either approached or exceeded the whole circle, or 720, read off by 
that vernier, and divide the amount by the number ofthe angles taken, and the quo- 
tient will be the correct angle ; it will here be clearly evident, that all errors of di- 
vision, centering, and observation, will have been taken into the account, and, there- 
fore, the angle will have been correctly obtained. 
/ % 
Observations by Captain Sabine, from his work on the Pendulum. 
In the spring of 1821, Mr. Dollond, to whom practical astromony has so many 
obligations, was kmd enough to show me the design of a repeating reflecting circle, 
1 This cannot surely apply to the ordinary reflecting- circle.— E d. 
