MERIDIONAL OBSERVATIONS. 239 



The mean value 3" 36 was adopted. Mr. Johnson had considerable trouble 

 with this striding level, as it was continually requiring to be adjusted. For 

 this reason the level was not applied quite as often as was desirable. It was 

 Mr. Johnson's habit to apply the level, in both positions of the transit 

 instrument, on the finish of his night's work. Although the piers were of 

 stone, the illuminating lamp seems to have disturbed the level error. The 

 "clock slow" obtained from stars with lamp E. is generally greater than 

 with lamp "W. Mr. Johnson always observed clock stars with both positions 

 of the instrument. 



The correction to the level error for inequality of the pivots was found by 

 repeated levellings to be 0"'42 (micrometer E.). 



The Error of Collimation of the fixed center wire of the transit was obtained 

 from observations of close circumpolar stars with reversed positions of the 

 transit axis. All such observations were transits over the movable wire, and 

 the error of collimation of the center fixed wire is obtained from them by 

 assuming that it coincided with the movable wire when the micrometer 

 reading was I0 r> 043. The individual determinations are given in Table II. 



The Error of Azimuth was obtained in the usual manner from the obser- 

 vations of close circumpolar stars combined with those of clock stars. The 

 errors are given in Table III. There is no mention made of the adjustments 

 having been altered between the observations of December 3 and December 4. 



TRANSITS of STARS observed at WAIMEA (Table IV.). 



The places of the stars Bradley 95, 402, 3147, 3187, Groombridge 642, 

 750, and Lalande (Fed.) 693 have been brought up from the Greenwich new 

 7-year and 9-year Catalogues for 1864 and 1872. The constants for com- 

 puting the apparent places were extrapolated for 1875 '0 from those contained 

 in the annual volumes of the Greenwich Observations. 



For complete transits, the mean observed time of transit by chronometer N 

 over the five wires was reduced to the center wire by applying the correction 



i5 Sin NPD additive when the micrometer-head was east. For imperfect 

 transits, each wire was reduced to the center wire by means of the intervals 

 given on page 238. "When only one side wire has been observed, the time 

 recorded is given at the foot of the page. The level error proper for the 

 position of the transit axis, as given in Table I., has been used, taking the 

 mean where it has been observed more than once. The value of the Colli- 



