Permanence of the Rutherfurd Photographic Plates. 213 



repeat all the comparisons, taking the stars in the inverse order. 



" (d) Repeat (b) and (a). 



" The series of measures (a), (b), (c), and (d) will constitute a 

 morning's work ; and this can be carried on till all the stars have 

 been measured in one coordinate. The microscopes connected 

 with the position circle should be read each day, before beginning 

 and after ending work. The same should be done with the ther- 

 mometer. 



" To measure the other coordinate, the position circle must be 

 turned through a right angle, as nearly as may be, and all the 

 stars must be measured in the same way." 



It will be seen that throughout the above instructions the prin- 

 ciple of repetition in the reverse order has been as closely adhered 

 to as before. We have not been content with measurement made 

 according to the above plan, however, but have again repeated all 

 the measures in exactly the same way with the position circle 

 turned through an angle of i8o° from each of its former positions. 

 This doubled the labor, but it has the advantage of eliminating 

 any personality in bisecting the images. For if the observer has 

 a habit of always putting the thread of the microscope too far to 

 the right in making a bisection, then when the star-image is re- 

 versed i8q° the same personality will affect the resulting coordi- 

 nate with the opposite sign. Consequently the mean of the two 

 measures will be free from any such personality, if it remain con- 

 stant. Needless to say, as every star was measured twice at least 

 each day, it has been possible to have the measure of each star on 

 each day divided equally between the two observers. Conse- 

 quently, the final position for each day is made to depend on two 

 observers instead of one. In making the measures with the plate 

 reversed i8o°, operations i and 6 in the observer's instructions 

 were interchanged for the more complete elimination of the peri- 

 odic screw-errors. 



We shall give an example of the method employed for record- 

 ing and computing the comparisons of the star and scale. The 

 star measured is Anon. 34, which was one of the standards. Of 

 course the blank forms for recording these measures were arranged 

 so as to bring all the measures of any given star so far as possi- 

 ble together on one sheet of paper. But all the measures of the 

 present research were made in the manner shown below, so that 

 the series of operations, of which this is a specimen, is to be re- 



Annals N. Y. Acad. Sci., IX, October, 1896.— 15. 



