Permanence of the Rutherfurd Photographic Plates. 211 



various errors of the machine, we come now to the measures of the 

 plates. As we have already pointed out, the arrangements are 

 such that the stars on the -plate are compared successively with 

 the divisions on a metal scale. Evidently it is merely necessarj^ 

 to subtract the reading of the scale belonging to the central star 

 from that belonging to any other star, in order to get the coordi- 

 nate of the latter, referred to the central star as the origin. It 

 will be most convenient to use the central star as the origin, be- 

 cause the old measures were made in polar coordinates, i. e., posi- 

 tion angle and distance from the central star. 



The first difficulty that presents itself arises from the fact that 

 we are measuring glass plates with a scale made of German silver. 

 As the coefficients of expansion of glass and German silver are 

 very different, errors of appreciable magnitude might easily be 

 caused by variations of temperature during the observations. 

 For this reason, we have always repeated all our observations in 

 the reverse order, immediately after finishing a series of any kind. 

 The manner of doing this will be explained below. We have also 

 taken the greatest care in securing readings of the position circle 

 microscopes and the runs of the micrometer. Both were taken at 

 the beginning and end of each series of observations, experience 

 having shown that it was unnecessary to take them more frequentlJ^ 



All the observations were made by Mrs. Herman S. Davis, and 

 Mrs. Annie Maclear Jacoby. The following are the instructions 

 that were written out by me for the guidance of the observers. I 

 give first the instructions for making a complete comparison of a 

 star with the scale : 



" I, See that the eye-piece is central over the wires, and the 



reading of the micrometer about g.ooo. 



2. Read on the eastern impression of the star. 



3. Read on the next smaller number of the scale. 



4. Repeat 3. 



5. Repeat 2. 



R 



6. Change the reading of the micrometer to about 9.500. 



7. Repeat 2, 3, 4, and 5, using the western impression of the 

 star." 



These seven operations constitute a complete comparison of the 

 star with the scale, and when speaking of such comparisons, 

 complete comparisons are alwaj^s meant. 



