LONGITUDE. 



?M 



I. ir. III. IV, V. vr, vrr. 



51M3 34-.03 IT',09 00^00 17M9 81' 14 51'.32 



with a probable error for each value of dz 0".005. 



of 



seven wires to the middle wire:= — 0'.06 it 0.00. 



2. The computed value of the error of coUimation of the middle wire, nn deduced from a 



mean of twenty-five observations, made on the nights of October 27, 28, and 29, (see table 0,) 



is =, 



+ 0M6 =b O'.Ol ; hence the collimation of the moan of wires, = + OMO ± O'.Ol = (in 



arc)^-^^5zhO^M. 



This value for the collimation of mean of wires has been applied to all the observations of 

 the four lunations, in consideration of the improbability of any ])erceptible change in that 

 element of the instrumental errors, due to the superior mechanism of the instrument, 



3. The azimuthal deviation, however, has been applied to the observations of each night, as 

 then observed, on account of the unavoidable instability of the foundation of the instrument 

 stand, and shown by the changes of deviation exhibited in column 13 of tabic A. 



4. The ec[uatorial value of the divisions marked on the level used with this instrument has 

 been determined from observations made August 19, 1853, the result being =: l".0r»5 (in arc) 

 for one division (see table D.) 



The chronometer rate adopted was obtained by reducing to 0^ sidereal time the error for each 

 night by interpolation ; and although the rate therefrom deduced is very satisfactory and regular 

 for the period of any one lunation, it perceptibly continues to increase throughout the scries, 

 being for the 



First lunation^ in August 



Second lunation, in September 



Third lunation, in October 



Fourth lunation, in November 



+ 1M8 



+ 2'. 89 

 + 4'. 60 



In reducing the observations, therefore, the actual rate for the day of observation has been 

 "ised, as shown in column 18 (table A) of the tabulated computations. 



In applying the corrections due to errors of collimation, level, and azimuth, the tables 

 marked E have greatly facilitated the reduction — all which are combined in column 15 (table 

 A) of the tabulated computations, showing the ^^ corrected transits" of stars* 



The value for ^^ sidereal time of moon's semi-diameter passing meridian of station,'* shown 

 in column 19, (table A^) has been computed by interpolation, and applied to corrected transit 

 of the limb observed, giving the corrected transit of its centre ; which element has been used 

 in computing, by interpolation, the change of the moon's position. 



The positions of the moon, as observed with the transit circle at Greenwich observatory, have 

 served to give, by interpolation, its position for all the nights on which it was observed at the 

 mouth of the Eio Bravo, furnishing reliable and satisfactory results ; and these positions were 

 adopted in preference to those deduced from the altazimuth observations, which differ in most 

 cases from those actually observed by the transit circle. The positions adopted, as well as those 

 observed at Philadelphia High School and Eadcliffe observatories, during the period stated, are 

 given in table F, being the data used in computing the longitude. 



The several results for longitude, computed from the observations of each night, are exhibited 



in table Gr ; the mean of which is G' 28°" 30^48 =t 0^56, beiug deduced from forty observations 

 combined with the Greenwich observations, ten with Philadelphia High Sshool^ and four with 



