1. Preface by Edward C Pickering, Director of the Harvard College Observatory.
The study of variable stars has formed an important portion of the work at the Harvard Observatory. In recent years, increased attention has been given to this subject, owing to the large number of variables discovered by means of their spectra, by direct comparison of negatives of clusters, and by superimposing positives on negatives; also, from visual estimates of the brightness of large numbers of variable stars of long period. A bibliography of variable stars was begun by Professor W M Reed, in 1897, who wrote fifteen thousand cards. It was transferred to Miss Cannon in 1901. Since then, she has added twenty thousand cards, and from them prepared a Provisional Catalogue of Variable Stars, published in these Annals, 48, No. 3, and annual supplements published in Circular 77, and Annals, 53, No. 7. A Second Catalogue of Variable Stars is contained in Part I of the present volume, and is intended to take the place of the previous catalogue published by Miss Cannon. The value of the work is greatly increased by the remarks and tables following the Catalogue. Part II will contain additional material extracted from the bibliography, including a study of all the published maxima and minima of variable stars of long period.
21 December 1906.
2. Extract of Introduction by Annie Jump Cannon.
Since 1903, the number of known variable stars has increased greatly, owing mainly to the discovery of large numbers of these objects in nebulous regions. 240 of these have been found by Wolf at Heidelberg, and 2,110 by Miss H S Leavitt at Harvard. The greater portion of these stars are so faint that they can be observed visually only with large telescopes, and cannot even be found except with the aid of marked photographs. However, to render the Catalogue as complete as possible, it seemed best to include these objects, with the exception of 1,791 found in the Magellanic Clouds. The present catalogue contains 1,957 stars, counting those in globular clusters. Adding to this, the 1,791 found in the Magellanic Clouds, we have 3,748 known variable stars. 2,909 of these have been found at Harvard. 514 of the latter have been found by Professor Bailey in globular clusters, and 221 by Mrs Fleming, mainly by the presence of bright hydrogen lines in stars of the third type, as a result of the examination of stellar spectra forming a part of the Henry Draper Memorial. The material for the preparation of this Catalogue has been obtained from the Card Catalogue containing references to all published observations and investigations of variable stars, supplemented by unpublished photo and visual observations made here. Sequences of comparison stars are selected from the photographs and measured by Mrs Fleming for all variable stars discovered here, except those in clusters or special regions. Measures of the variables are made by Miss S E Breslin, comparing them by Argelander's method with two of a sequence whose magnitude has been determined. A graphical representation of the observations is then made. The plates upon which these stars appear, extend in general from 1889 to 1906, and thus the observations enable us to derive a more accurate value of the period and a better knowledge of the light curve of recently discovered variables than is possible from visual observations which extend over only a short period of time. Since 1904, the list of variables of long period observed visually at this Observatory has been increased to 309 stars. Monthly observations of these objects have been made by Mr Leon Campbell with the 5-inch or 24-inch telescopes, and by the writer with the 6-inch telescope. Observations have also been sent to this Observatory from the Halsted, Leander McCormick, Vassar, Mt. Holyoke, and Whiteside Observatories, and from Mr J H Eadie and Senor Pereira.
This large amount of unpublished material has been of use in obtaining the observed range of variation of a number of stars, and in testing the given periods or in deducing new ones. Perhaps the most difficult task of the compiler of a Catalogue of Variable Stars is the selection of the best period. Different investigators give different values, and that which is accepted today, may be rejected tomorrow, in the light of additional facts. Much time has been spent on this part of the work, and it is believed that some of the formulas given here are more accurate than any yet published. For a large number of the stars discovered here, new elements have been derived, by a method of Least Squares, from all the available dates of maxima obtained from photographic measures, or from both photographic and visual observations. Circular 81 contains the results obtained by this method for 7 stars. For purposes of comparison, a trial of Professor Turner's method, as detailed in Memoirs Royal Astronomical Society, Volume 55, was also made in the case of several stars. The period of S Lacertae obtained by the method of Least Squares, and published in Circular 81, is 237.5 days. It is of interest to note that a computation of the period of this star was made by Professor Turner, by his method, using all the visual observations made here from 1894 to 1904. His result is sensibly the same, 237.6 days, or, omitting one discordant observation, 237.5 days. Somewhat larger differences were found for three other stars whose periods were deduced by both methods. When the elements derived here are given in Table I, the facts upon which they are based will be found in the remarks following the table.