ASTRONOMY BOSS. 8 1 



tant, but hitherto insufficiently treated at many observatories, are 

 required to be known with substantial completeness and accuracy if 

 the full benefit of corresponding observations in the two hemispheres, 

 north and south, with the same instrument as contemplated, is to be 

 realized. 



Accordingly the pivots were reground and polished by the astro- 

 nomical staff of the observatory during the autumn of 1904 with 

 satisfactory results. On completion of this work the errors due to 

 any outstanding deviations from cylindrical form of the pivots were 

 thoroughly investigated by a special method found to be expeditious 

 and efficient. The result, with description of method employed, has 

 been published in the Astronomical Journal. The indications are 

 that during a revolution of the telescope about its horizontal axis 

 there is no deviation of the line of collimation from the normal plane 

 greater than one-hundredth of a second of time. 



Thorough examination of the circles last December disclosed the 

 fact that the graduations in some places had become seriously im- 

 paired during the necessarily repeated polishing for removal of tarnish 

 which the circles had experienced during 47 years of use. It became 

 doubtful whether the elaborate results for correction of graduation- 

 error obtained in 1 894-1 895 were still applicable. In short, it was 

 determined to procure regraduation of the circles by Warner & 

 Swasey, of Cleveland (Mr. Dudley Olcott, of Albany, having provided 

 means for the purpose), and then to determine in the most complete 

 and thorough manner the errors of the new graduations. 



During the summer of the present year this work has been carried 

 out, and has been extended to every multiple of 10' for both circles. 

 Moreover, the graduations, respectively distant 2', 4', 6', and 8' from 

 the 10' lines, have been investigated for systematic errors, with the 

 result of finding that this error, though rather large, is wonderfully 

 uniform throughout the circles. When the computations are com- 

 pleted the corrections for error of graduation, for each line on both 

 circles, should be known with a degree of accuracy amply sufficient 

 for the purpose. The ' ' circle-readings ' ' for the investigation proper 

 occupied a little less than fifteen weeks from June 6. Seven assist- 

 ants, with the director, were engaged upon the task. The observers 

 worked in three relays in one-hour shifts, two persons being 

 required in each relay, and the readings were kept up absolutely 

 continuously from 8 a. m. until 5 p. m. every week day; but more 

 commonly the hour of beginuing was 7 a. m. It is estimated that 

 about 450,000 micrometer measurements were recorded in this work. 

 The reductions have been proceeding contemporaneously with the 

 6 



