May 1, 1895.] 



KNOWLEDGE 



111 



naked-eye variables to be investigated with a new spectro- 

 graphic apparatus fitted to the thirty-inch Pulkowa 

 refractor in the summer of 189i ; and he obtained, in 

 August and September, a series of thirty-four photographs 



Mag 

 3 7 



4 9 





n IS 24 30 Rf, -/J? -A* S4 GO 6€ 72 /« S4 OO S€ lOZ W8 114 I20 126 132 



Fig. 1.— Light-Curve of 8 CepUei (ScUoufeld). 



of its dispersed light, each plate being impressed besides 

 with spectra for comparison of iron and hydrogen. Their 

 examination showed at once the progress of orbital motion 

 in a period indistinguishable from that of the luminous 

 change. It was, however, betrayed, not by the alternate 

 separation and closing-up of two sets of lines, but by 

 the swinging to and fro of a single set. The star 

 thus circulates, like Algol, roimd a partially or totally 

 dark body. M. Belopolsky has calculated its orbit by 

 a method invented by Dr. Eambaut {Monthb/ Notices, LI., 

 816), and improved by M. Lehmann-Filhes {Astr. Nach., 

 No. 3242). It proves to be so considerably eccentric that 

 the star is three times nearer to the centre of gra-\dty of 

 the system at its furthest retreat fi'om, than at its nearest 

 approach to it ; and the long axis of the ellipse is turned 

 almost directly towards the earth, with the apastron at the 

 hither side. Half this axis (that is, the mean distance), 

 multiplied by the cosine of the inclination to the visual 

 ray, is 800,000 English miles in length. But since the 

 inclination is undetermined, and perhaps undeterminable, 

 this is no more than a minimum estimate. If the orbital 

 plane, for instance, made an angle of 60° with our line of 

 sight, it should be doubled. Nor is any reason apparent 

 why the angle should not be even greater than 60°. 

 Where an eclipse may be supposed to occur, the case is 

 different. Then the inclination must be approximately 

 zero. Half the sum of the measured velocities of Algol, 

 when moving directly towards and away from the earth — 

 in other words, as it passes through the nodes of its orbit 

 — thus represents neither more nor less than its actual 

 velocity of circulation. 



In 8 Cephei, none of the symptoms characteristic of an 

 eclipse are visible. The shape of the light-curve is scarcely 

 compatible with its occurrence. It shows no abrupt 

 hollow, no prolonged high-level of full brightness. The 

 maximum, on the contrary, is no sooner reached than a 

 gradual and indecisive descent towards minimum begins. 

 The former is accordingly much more sharply defined 

 than the latter. So that the conditions imposed by the 

 eclipse- hypothesis, far from being complied with, are 

 reversed. It would indeed be impossible to devise any 

 form of occultation capable of explaining the entire 

 variation of S Cephei ; none even tending to simplify, or 

 usefully to supplement other causes of change. The 

 introduction of geometrical obscurations appears, then, to 

 be superfluous and misleading. 



When they really occur they are unmistakable. Algol - 

 stars have their character writ large enough for the 

 quickest runner to read. Rigid tests can, moreover, be 



applied. Eclipses can only take place when the bodies 

 concerned are moving across the line of sight— that is, at 

 periods of spectroscopic immobility. Prof. Yogel has 

 shown that this requirement is 7iot conformed to at the 



muiima of (3 Lyra, which can- 

 not, therefore, be due to the 

 alternate concealment of one 

 of the star's lustrous compo- 

 nents by the other. And M. 

 Belopolsky's data regarding S 

 Cephei are, as they stand, 

 equally conclusive. In a system 

 like that of Algol, in which a 

 bright star passes behind a 

 dark - one, radial movement 

 must evidently be recessional 

 previous to the eclipse, cease 

 altogether during its progress, 

 and reverse its direction when 

 the eclipse has terminated. 

 These rules, which are of the very essence of the pheno- 

 menon, are strictly obeyed by Algol. But by the Cepheus 

 variable, although likewise composed of a luminous and 

 a non-luminous body, they are disregarded. Its mimina, 

 if eclipses were in any way involved in their production, 

 should virtually coincide with the star's periastron-passage 

 — the periastron in this orbit being 88° from the ascending 

 node, or nearly in our line of sight. In point of fact, they 

 precede periastron by one entire day. Hence, spectroscopic 

 motion, instead of becoming extinct when the lowest phase 

 is attained, proceeds throughout its duration at a relatively 

 high rate, and in an unchanged direction. The star is in 

 course of retreat from the earth no less after than before its 

 minima, which hence demonstrably occur at a part of the 

 orbit where eclipses are impossible. It should be added that 

 the results so far secured at Pulkowa are considered by 

 their author as merely provisional, so that no final 

 judgment can be pronounced on any point connected with 

 them. Still, it is scarcely to be expected that the glaring 

 discrepancy just adverted to will be abolished by any 

 amount of future research. 



The system of 8 Cephei is transported towards the sun 

 at the rate of 8-7 miles a second. Its visible member gives 

 a spectrum of the solar type, but with some faint solar 

 lines widened and strengthened. All previously known 

 spectroscopic binaries, including the whole class of Algol- 

 variables (so far as their light has been examined), are 

 " white stars," and the rule was held to have some 

 significance as regards their history and development ; 

 hence, the discovery of an exception to it deserves note. 

 Delta Cephei is of a full yellow colour, and has an azure 

 companion at 41", the pair making a tinted combination 

 hardly inferior in beauty to /3 Cygni. It is unlikely to be 

 of a merely optical nature. Contrasted hues afford in 

 themselves a presumption of physical relationship ; and a 

 common proper motion, according to Mr. Burnham, 

 assures the reality of the tie. But its amount — less than 

 2" a century — is, perhaps, too minute to justify his 

 inference. 



There is no longer any reason to doubt that all " short- 

 period variables " are really close binaries. Their spectro- 

 graphic study hence acquires exceptional importance, and 

 cannot fail to be eagerly pursued. Among the stars most 

 promising for investigation are v) Aquilfe, ? Geminorum, 

 and R Muscic. The last-named object is circumpolar at 

 the Cape ; and although it very slightly exceeds the limit 

 of naked-eye perception at maximum, its inferiority in 

 brightness may well be compensated by the wonderful 

 rapidity of its changes. They are finished in 21h. 20m., 



