The determination of the angles of position and distance of double stars forms a very important and extensive branch of work in connection with sidereal astronomy. In cases where double stars preserve stationary places relatively to each other, there is clearly no need for frequent re-observation. But in those numerous instances where the two components form a binary system it is desirable to obtain as many measures as possible, so as either to verify the calculated orbit or to furnish the materials for an orbit if one has not been computed before. Dr. Doberck, whose name is well known in these researches, mentioned, in 1882, that ample data for purposes of computation had not been secured for more than thirty or forty binaries out of between five and six hundred such systems that were probably known to exist. Sir W. Herschel, in 1803, estimated the period of revolution of α Geminorum as 342 yrs. 2 mths. and of γ Virginis as 1200 yrs. Orbits[51] do not appear, however, to have been computed until 1827, when Savery of Paris showed that the companion of ξ Ursæ Majoris was revolving in an ellipse with a period of 58-1/4 years. The accomplished Encke also turned his attention to this work, and adopted a more elaborate method; and many others have pursued the subject with very interesting and valuable results. On pp. 302-305 is a selected list of some of the most noteworthy double and binary stars, arranged according to the distance between the components.
In compiling the above list, I have used some of the latest measures available, as most of these doubles are binary systems, and therefore in motion, so that a few years effect a perceptible difference in the angles of position and distance of the components. Some of the pairs are closing up, others are opening, and thus it happens that a binary star, divided with great difficulty to-day, may become an easy object some years hence, and _vice versâ_. In fact, as telescopic tests they are constantly varying.
Before leaving this part of the subject it may be interesting to refer individually to a few brilliant examples of double stars.
α _Canis Majoris_ (_Sirius_). A red star according to ancient records, but it is now intensely white. In 1844 Bessel inferred from certain little irregularities in the proper motion of this star that it consisted of a binary system with a period of about half a century[52]. Peters confirmed this idea in 1851, and it was observationally verified eleven years afterwards. On Jan. 31, 1862, Alvan Clark, jun., while testing a new 18½-inch refractor, discovered a very faint companion 10″ distant. Measures in the few subsequent years proved that the position-angle was decreasing, while the distance showed a slight extension. In 1872 it was about 11″·50, but since then the two stars have been approaching each other, and Mr. Burnham’s measures in April 1890 gave the distance as only 4″·19. It is now, therefore, a very difficult object, and only visible in large instruments. In England it is never easy, owing to its southern position, and it has been little observed, but it is satisfactory to note that the large refractors at Washington, Princeton, and Chicago, U.S.A., have been often employed on this object in recent years. Mann gives a period of 51·22 years for this interesting binary, and places the time of periastron-passage as 1890·55. This differs from Gore’s orbit, quoted in the table.