With reference to the well-known double (or rather quadruple) star ε Lyræ, near Vega, and supposed faint stars near it, Burnham says, "From time to time various small stars in the vicinity have been mapped, and much time wasted in looking for and speculating about objects which only exist in the imagination of the observer." He believes that many of these faint stars, supposed to have been seen by various observers, are merely "ghosts produced by reflection." The binary star ζ Boötis, which has long been suspected of small and irregular variation of light, showed remarkable spectral changes in the year 1905,[319] somewhat similar to those of a _nova_, or temporary star. It is curious that such changes should occur in a star having an ordinary Sirian type of spectrum! A curious quadruple system has been discovered by Mr. R. T. A. Innes in the southern hemisphere. The star κ Toucani is a binary star with components of magnitudes 5 and 7·7, and a period of revolution of perhaps about 1000 years. Within 6' of this pair is another star (Lacaille 353), which is also a binary, with a period of perhaps 72 years. Both pairs have the same proper motion through space, and evidently form a vast quadruple system; for which Mr. Innes finds a possible period of 300,000 years.[320] It is a curious fact that the performance of a really good refracting telescope actually exceeds what theory would indicate! at least so far as double stars are concerned. For example, the famous double-star observer Dawes found that the distance between the components of a double star which can just be divided, is found by dividing 4"·56 by the aperture of the object-glass in inches. Now theory gives 5"·52 divided by the aperture. "The actual telescope--if a really good one--thus exceeds its theoretical requirements. The difference between theory and practice in this case seems to be due to the fact that in the 'spurious' star disc shown by good telescopes, the illumination at the edges of the star disc is very feeble, so that its full size is not seen except in the case of a very bright star."[321] CHAPTER XVI Variable Stars