DOUBLE SHUFFLE. An expedient by which John A. Macdonald and his colleagues of the Canadian Macdonald-Cartier administra tion, when they succeeded the short-lived Brown Dorion ministry in 1858, evaded the necessity of vacating their seats and going before their constituencies for re-election. The Brown Dorion ministry had only lasted three days; and the statute governing the case stated that if a minister should resign and within one month• accept another office in the ministry, he should not vacate his seat. Accordingly Mac donald and his colleagues changed portfolios, took the oath in their new offices, and immedi ately thereafter resumed their former places in the ministry.
In many such cases it is afterward found, how ever, that the two stars are not members of one system but that one of them is vastly farther away from us than the other, their seeming close proximity being only apparent and not real. Such pairs are called Qoptical° pairs. As it is not until measures have been accumu lated for at least several years that a distinction can be made in any case between an optical pair and a true binary system, all discoveries are tentatively catalogued as double stars. About 20,000 such pairs are now known. The first systematic study of these objects was made by Sir William Herschel, who, between the years 1779 and 1784, made extensive catalogues of them, because he thought it likely that they were only optically doable. If that were the case, the earth's orbital motion about the sun might be great enough to make an appreciable variation in the distance or direction of one star from the other, if the system were viewed at different times in the year. After recording the distances between many of these stars, as well as the direction of the line joining the two components, he went over them again to see if any changes could be noted; and in so doing (as his son, Sir John Herschel, has re corded) °his attention was altogether diverted from the original object of the inquiry by phe nomena of a very unexpected character, which at once engrossed his whole attention." In stead of finding an annual and alternate in crease and decrease of the distance, and a corre sponding periodic variation in their direction, he saw that in many cases there was a contin uous change in both, advancing steadily in one direction. If this were due to the fact that the solar system and the two components of the double star were moving independently through space, we should expect the motion of either of the components, relatively to the other one, to be sensibly rectilinear. After the lapse of 25 years Herschel became satisfied that this is not always the case, for in some instances he found good evidence of actual orbital mo tion, and in 1802 he announced that "there exist sidereal systems, composed of two stars revolving about each other in regular orbits," in the same manner that the earth revolves about the son, or the moon about the earth.
About the year 1819 William Struve took tip the systematic study of these stellar systems, and made a long series of most excellent meas ures of them covering a period of many years. Subsequently the work was taken up by other skilled observers, and we now have a vast mass of data of this sort, from which fairly good orbits of some 60 of the systems have been obtained. A beginner's book on the sub ject is Crossley, Gledhill and Wilson's book of Double Stars,' which contains methods for computing the orbits of double stars, as well as advice concerning the making of obser vations. The standard and authoritative work is S. W. Burnham's Catalogue of Double Stars,' which contains the description of every pair discovered, and references to all measures ever published, and to all proper mo tions and orbits computed, up to the year 1906. In reality, both of the components of a double star are in motion, each revolving about the common centre of gravity of the two; but it is known that in a system composed of two bodies revolving about each other no error is com mitted, so far as the relative motion of the two is concerned, by regarding one of them as fixed, and assuming that the other revolves about it. It is customary, therefore, to regard the brighter component of such a star as fixed. The diagram shows the orbit of the pair known as F. 70 Ophiuchi, in which one star is of about the fourth magnitude, while the other is of about the sixth. The large black spot at the intersection of the two straight lines represents the larger star, the straight lines themselves representing, respectively, an hour circle and a declination-circle passing through it. The ellipse represents the apparent orbit of the component star, as determined by Burnham from all the available data up to the year 1905. The dots that appear irregularly along the ellipse, being sometimes within it and sometimes without, represent the observed positions of the com panion star. According to the computations here cited, the period of revolution of 70 Ophiuchi is 87.75 years. In addition to the stars that are visibly double when viewed through the telescope, numerous stars are now known which have been proved to be double by means of spectroscopic observations. (See Doreunt's PRINCIPLE). The periods of revolu tion of some of these are exceedingly short. Thus, one of the components of the telescopic ally double star Kappa Pepsi is itself a spec troscopic binary with a period of only six days. At least one of the spectroscopic double stars (Epsilon Hydra), is also a visual double in the telescope; the orbit computed from the di rect measures agrees remarkably well with that found from the spectroscopic observations, thus affording a proof (if one were needed), that the displacement of the spectral lines in these cases is due to a true duplicity and is not the effect of some other, unknown, cause.