The Rutherfurd Photographs. 193 



or otherwise the plate may be so placed in the machine that a cir- 

 cle of declination through the central star shall be approximately 

 parallel to the guiding cylinder ; in this way we obtain rectangu- 

 lar coordinates which are nearby in the directions of right ascen- 

 sions and declinations, thus rendering their conversion into the 

 latter a comparatively easy matter. 



Two observers alternated in the measurements, one recording 

 while the other observed ; the details of each morning's work, 

 which usually lasted a little over two hours, are as follows : the 

 first observer reads the circle, runs and temperature, the second 

 reads on the central star thus : East image, scale, scale, east 

 image ; west image, scale, scale, west image. Continuing, the 

 second observer goes through the same operations for usually 

 three other stars, experience having shown that four stars could 

 be read conveniently without fatiguing the eye ; thus the ob- 

 servers alternate till twenty or twenty-five stars have been read 

 and then the temperature is recorded a second time. The morn- 

 ing's work is now half finished ; the same stars are then observed 

 in the reverse order, care being taken that each observer shall now 

 read those stars that he had not read in the first half; having 

 finally gotten back to the central star, temperature, runs and the 

 circle are read as at the beginning. This process of repeating all 

 the measurements in the reverse order, eliminates the effects of 

 any change in the machine or in the observers that is pro- 

 portionate to the time, for the mean of the two times of ob- 

 servation is nearly the same for all the stars. In the first half of 

 the morning's work the micrometer head is set at about 9. R 

 when pointed at a star ; but in the second half the reading is made 

 9. E 5 ; in this way periodic errors of the screw are nearly elimi- 

 nated, for both star and scale are read with two different parts of 

 the screw separated by half a turn. 



The measurements made in the first position of the plate, i. e., 

 with the stars having the greatest right ascensions farthest from 

 the cylinder are recorded as " x direct ;" on the next day the plate 

 is turned 90° in a counter-clockwise direction so that now the stars 

 having the greatest north polar distances are farthest from the cylin- 

 der. The measurements taken in this position are called u y direct," 

 while those taken in the two opposite positions, 180° and 270° 

 from the original position are called respectively " x reversed " 

 and " y reversed." As only twentj" or twentj^-five stars could be 



