OF THE UNITIiD STATES. 325 



the circle is such as to make the use of the screws at the 

 legs for this purpose actually olycctiunable, as it brin<;;s the 

 instrument out of its vertical and occasions always a tedious 

 adjustiuent of the level ; besides, tliat when a number of ob- 

 servations are to be made consecutively, it would be neces- 

 sary to level the whole instrument again after every one, 

 while otherwise the circle, once well adjusted, will remain 

 so for as many observations as may i)e made in a whole 

 night, if proper care has been taken as to the solidity of the 

 stand, which, in the field, must always be placed on three 

 plugs reaching deej) into tlie ground. 



It is evident, that the measurement of a series of angles 

 in the plane of two objects may be conducted, in respect 

 to the successive motions of tlic front and back telescope, 

 exactly as lias l)een described above, and the results taken 

 in the same manner. 



As two opposite readings correct the eccentricity, if any, 

 it is evident that the indiscriminate mean of any two opposite 

 readings sliould always be equal ; but the diilcrent sinkings 

 of the readings, and the diflferent influence of the weight of 

 the telescope and other moveable paits of tlie instrument, 

 may introduce diflercnces. as is well known and sufficiently 

 discussed; therefore the indiscriminate mean of all four read- 

 ings is also here to be preferred, and will ultimately be found 

 to give a better result than a discrimination would give. The 

 Day-Book examples will show, in their places, how I pro- 

 ceeded, in this respect, to scrutinise the results of my obser- 

 vations. 



It is well known that this instrument is calculated to cor- 

 rect all its own eriors l)y the ell'ect of its repeating projierty 

 and construction, particularly in the vertical angles. Still 

 it is well known that (for instance in determinations of lati- 

 tude) it is proper to take the indiscriminate mean of an efpial 

 numl)er of results of observations froin the north and the 

 south side of t!ic xenith. 



For the same reason, if ever anv doul)t should exist as 

 to the want of stability in the parts ol the instrument, which 

 VOL. II. — r 2 



