I9I4.] IN CATALOGUING POSITIONS OF STARS. 305 



were to be surveyed with both instruments and the same comparison 

 stars were used in the two cases, our doublet would give the more 

 accurate results. If there were at hand in some special case com- 

 parison stars that are much superior in number and in accuracy to 

 those that are generally available, then the astrographic plates would 

 give the better positions. Needless to say, a doublet like ours with 

 twice its aperture and twice its focal length would give still better 

 results, but this would necessitate plates about 35 centimeters square 

 and a measuring engine large enough to accommodate them. 



Our objective would also be well adapted for compiling zone 

 catalogues of faint stars similar to those of the Astronomische 

 Gesellschaft. The latter now extend from 80° north declination to 

 18° south, a work that has required the cooperation of sixteen ob- 

 servatories during a considerable number of years. Observations 

 are in progress that will extend this catalogue farther south, but no 

 provision has yet been made for the southernmost zones. When 

 these come to be actively considered, the claims for the doublet 

 should be carefully weighed. The original plans for the Gesell- 

 schaft Catalogue included the repetition of the observations after 

 the lapse of about half a century. This time is now approaching in 

 the case of some of the northern zones, though others (notably the 

 one between 70° and 75°) are of more recent origin. The repeti- 

 tion of all the northern zones, if carried out photographically by 

 means of a doublet, would be a task well within the powers of a 

 single observatory. IMoreover there would be a very considerable 

 gain in accuracy. From the prefaces to the various zones of the 

 Gesellschaft Catalogue we learn that the probable error of one 

 observation in either right ascension or declination is on the aver- 

 age well over o".5o. This was derived from observations made 

 with the same telescope on different nights, and does not include 

 certain errors that would be brought out by comparing observations 

 made at different observatories. Our plates yield for the probable 

 error of one observation, o".i8. This was obtained by comparing 

 overlapping plates, but it does not include errors due to inaccuracies 

 in the positions of comparison stars. 



Although the tests we have made with our doublet may be ac- 



PROC. AMER. PHIL. SOC, LIII, 2IS, T, PRINTED DEC. II, 1914. 



