November, 1913. 



KNOWLEDGE. 



423 



Chandler adopted an entirely new notation in his 

 three catalogues ; we give the reasons for the method 

 in his own words : — 



" The number of a star, upon a system of ordinal notation 

 designed to remedy the inconveniences attending the usual 

 current numbers . . . . if the numbers of any one list 

 are retained, the interpolated stars require a suffix letter, 

 resulting in a hybrid notation which is exceedingly objection- 

 able, and which sooner or later has to be reformed ; when the 

 whole process of degeneration, with its awkwardness and con- 

 fusion, begins anew. It seems certainly better to adopt a 

 system which attaches a permanent numeral to each star, and 

 which permits interpolation to a practically unlimited extent. 

 I would accordingly suggest that the numbers for variable 

 star catalogues be one tenth of the right ascension, 

 expressed in seconds of time, for the equinox 7900-0." 



This method limits us to 8640 variable stars in the 

 whole sky unless we adopt the " hybrid notation " 

 disparagingly referred to in the above quotation. 



The publication known as the Vierteljahrschrift 

 der Astronomischen Gesellschaft has contained, since 

 1870, a list of variable stars with ephemerides. The 

 work was edited or compiled formerly by Winnecke, 

 Schonfeld, and at present by Hartwig. Up to 1889 

 it contained stars to 2° south of the equator ; from 

 1889-1895 stars as far south as — 30° Dec. were 

 included; from 1895-1902 the catalogues were 

 extended to — 35° Dec, and after that variable stars 

 in the whole sky were included ; the catalogue of 

 1906 contained information for 709 stars similar to 

 that in Chandler's catalogues. The number in the 

 catalogue for 1913 is 962 stars north of — 23° Dec. 

 and 417 stars south of that. During the year 1912 

 148 new variable stars were discovered, 109 from 

 photographs and 39 visually; since 1900 about 

 80 per centum were found by ladies, chiefly from 

 Harvard and Arequipa photographs. The non- 

 continuance of the preparation for further editions of 

 Chandler's catalogue resulted in more attention 

 being given to variable star work at the Harvard 

 Observatory. For some years Professor W. M. 

 Reed had been forming a bibliography of variable 

 stars, on cards, giving the details for separate 

 stars, and in 1897 this extensive record had needed 

 15,000 cards ; then it was continued upon similar 

 lines by Miss Cannon, who added 20,000 cards ; and 

 the results of Professor Reed and Miss Cannon's 

 work formed volume XLVIII, No. Ill (published in 

 1903), and was called the Provisional Catalogue of 

 1,227 stars, and is usually considered to be the hrst 

 catalogue of variable stars independently formed at 

 the Harvard Observatory. A supplement was 

 published in The Harvard Circular, No. 77, and a 

 further supplement is in Harvard Annals, volume 

 LIII, No. VII. The continuation of this work at 

 Harvard with still greater vigour, aided by an 

 exhaustive and comparative examination of portions 

 of the Harvard extensive series of photographs, 

 revealed a large number of new variable stars, and a 

 new catalogue was required. The volume LV, 

 under the joint editors, Miss A. J. Cannon and 

 Professor E. C. Pickering, was published in 1907 to 

 1909 to replace the former first and provisional cata- 



logue, and it extends to 291 pages, entirely devoted 

 to variable star work. This second catalogue con- 

 tains 1,957 stars, with a supplementary list in Table 

 IX (page 272). On page 103 of volume LV begins 

 a list of 167 observers and observatories which have 

 contributed the observations included. 



The variable stars discovered in the globular 

 clusters are included in this number 1,957, but not 

 those 1,791 variable stars already discovered in the 

 Magellanic clouds. To December, 1906 only, there 

 were 3,748 variable stars to be dealt with : of 

 these, 2,909 have been discovered at the Harvard 

 Observatory, mainly from photographs taken there 

 and at Arequipa; 514 of these 2,909 stars were 

 found by S. I. Bailey in southern globular clusters, 

 221 more were detected by Mrs. Fleming from an 

 examination of third-type spectra (photographs taken 

 for the Henry Draper Memorial), and Miss H. S. 

 Leavitt has discovered 2,110 of the total 2,909 stars ; 

 they were mainly in the Magellanic clouds. 



The Harvard Observatory publications and work 

 upon variable stars really mark the introduction of 

 an entirely new system, and this brings us to the 

 third form of nomenclature. The great accession 

 to the number of variable stars in recent years, 

 largely owing to the extensive use to which the 

 photographic gelatine plate has been put, caused 

 Professor E. C. Pickering to form some easy, 

 permanent, and inexhaustible method of nomen- 

 clature for variable stars, and one that would at the 

 same time afford some other information than a 



mere name. 



The scheme which he first adopted in the Harvard 

 Provisional Catalogue was numerical in form, yet 

 not a mere number, but figures which also give the 

 position of the star in the sky with sufficient 

 accuracy to enable an observer to remember or 

 ascertain from an abbreviated or compact list — 

 without the necessity of consulting a large catalogue 

 — whether the star is suitably situated for observa- 

 tion. The form of designation which was adopted 

 for the Harvard Provisional Catalogue is retained 

 in the second catalogue of variable stars in Vol. LV. 

 The method has been found of great convenience in 

 actual practice. It is in constant and daily use at 

 that observatory, and the six figures, to be read 

 as three pairs, are quite readily retained in the 

 mind for all stars frequently observed, and the 

 accuracy indicated by the figures is sufficient to 

 enable the observer to set the 6-in. telescope and 

 identify the variable star field readily. 



The plan is to give the hours and minutes of the 

 star's R.A. for 1900, omitting all seconds (not the 

 nearest minute) and the degrees of declination, 

 omitting all minutes ; when there are several 

 variable stars with the same hours and minutes and 

 the same degree, as sometimes occurs in clusters, 

 then a small letter is added beginning with a, b, c, 

 and so on ; in no case yet, we think, has more than 

 a single letter been required. Though the method 

 is really independent of the constellations it is better 



