STAR POSITIONS AM) CAr.ACTK CO-ORDI N ATKS. 47 



brig-hter stars then most studied ; to-dav it is known that stars 

 are travelling- in comniunities, and that the number of these com- 

 munities is not large, that the>', perhaps, can be counted on the 

 fingers of one hand, but it would not matter if there were i.ooo 

 communities — it is much easier to deal with i .000 communities 

 than with 13,500,000 individuals; and it is also known that the 

 number of large or erratic i)roper motion stars is very re- 

 strained. Hence, the real vahie of the Cartc-dit-Cicl consists in 

 its photographs or their enlargements; in fact, in the Cartc-du- 

 Cid and not in the places of stars derived from it. F.xamination 

 of these Cartes taken at different dates bv means of superposition 

 and projection in an enlarging lantern or l)v a blink-apparatus 

 (as made by Messrs. Zeiss) will at a glance discriminate those 

 stars which are moving " out of communitA- '" ; measures of these 

 and of a fciv of those moving " in communit\- " will give all the 

 information as to these motions whicli is rc(|uired. r>y this 

 means, instead of measuring on an average 300 stars per square 

 degree, about eight or nine will suffice. The saving from this 

 point of view alone is forty-fold. 



A superficial criticism is that the galactic plane is not rigidly 

 marked on the sky. Rut is the equator? The former is defined 

 l>y the galactic latitudes of the .stars at the chosen epoch, and at 

 later epochs by such latitudes as will reduce the totalitv of proper 

 motions f freed from the effect of motion of the solar system) to 

 a minimum. To this plane must be referred the equator and 

 equinox of the Earth, and not vice-versa. In the future.' .star 

 places will be determined solely by photographv. but to-dav we 

 are in a transition i^eriod and one of rivalrv. P.ut where the 

 ultimate victory will be is not uncertain. If we look over Boss's 

 Preliminary Star Catalogue, we see that with hundreds of 

 observations spread over a century, we m^\ hope to know such 

 a well observed star's place as determined with meridian instru- 

 ment with an accuracy of about -j- 0.05" at the centre of gravity 

 of the observations or of o.i" forty vears afterwards, f^rofessor 

 Kapteyn. in measuring photographs of the Hyades Group of Stars 

 fGroningen Publications, \o. 14). finds an accuracv as great 

 from one set of plates, whilst ^Nlr. F. Slocurn. in measuring the 

 parallax of Nova Gemini (2), finds a probable error of -f" 0.007" 

 from observations spread over less than one vear. Tt might be 

 objected that the photographic measures are differential 

 measures, which are always more accurate than absolute 

 measures, but it is probable that these differential photographic 

 measures can be continued around the skv in belts and the trian- 

 gulations closed with all possible accuracv. At the present time 

 Professor H. PT Turner is trying various methods of finding the 

 absolute places of stars by photography at the University Obser- 

 vatory at Oxford. The results so far obtained are encouraging, 

 but as might be expected with any new method, various difificul- 

 ties arise and have to be combated. It is. however, quite 

 doubtful if the probable error of a well-observed meridian star 



