ON FLAMSTEED'S STARS " DRSERVEDi BUT NOT EXISTING." 



73 



The last star is Br. 680, wliercfore its i)osition was taken from Miidler's Bradley, while the 

 others are from I'iazzi's Catalogue. If we add to the rigjht ascensions the constant 4'' 39™ 2P, and 

 snbtraet the declinations from 51° 30' 38", we shonld get the clock time of transit and the aj)- 

 parent zenit distance {per Uncas diagonales) that would have been observed by Flamsteed, viz: 



The zenit distance, within the probable error, is that of Piazzi 228, and by the assumption of 

 an error of 3™ in the time we can make also the right ascension agree with that star. This 

 seems to be, indeed, the most ])lansible explanation. There remains, however, one difficulty that 

 speaks against this identification. Flamsteed noted the star he observed, of Cth magnitude, while 

 Pi.azzi 228, is according to the Dm. 7"'. 5, and according to Piazzi and Bessel (Wj 4'' 1048) even 

 only 8™. There is here in the neighborhood the bright star Br. 080, which must have been in 

 the field almost together with the spurious star, Flamsteed's telescope having a field of about 1^°. 

 It would be singular, indeed, if Flamsteed should not have observed this star at all. But by two 

 plausil)le changes the record can be reconciled with it. 



First, as to the time of transit. Suppose the observer (Flamsteed) called out " twenty-five," 

 meaning 2.") seconds, but the amanuensis wrote 20™ 5^, not inquiring, of course, further for the 

 minute, which was 19. Hence, out of 9^ 19'° 25« became 9'^ 20™ 5". The zenit distance can be cor- 

 rected by the assum])tiou of 34° 55' instead of 35° 45', a mistake in writing that is easily made and 

 of which there are other examples in Flamsteed's manuscripts. Moreover, this observation was 

 hastened by the next star following soon after. It remains to be seen how the check reading ^jer 

 strias cochleae would have been. Leaving the fraction the same, but changing (upon the same 

 ground as stated before iu discussing the data for 90 and 101 Tauri) the number of entire revolu- 

 tions, if we replace 810.48 by 791.48, the difference between the two modes of reading comes nearer 

 to that derived from the other stars. For we Iiave — 



while the mean difference from the other stars (s. last column in second table above) is— IS" [or— 

 22" respectively]. 



The final position for 1700.0, thus corrected, becomes: 



The error still remaining in declination appears rather large ; but it will be remembered that 

 Baily found in the manuscript the sign of uncertainty attached to the zenit distance. 

 S. Mis, 154 10 



