58 



ASTRONOMICAL PROGRESS IN 1899. 



variable stars. The number now known is a 

 hundred times greater than were known fifty 

 years ago. But the most surprising and least 

 expected is that many stars in several clusters 

 vary, some during short periods and in wide 

 limits. The literature of the subject is extensive. 

 After the fact became known a systematic search 

 by photography was made by Mr. S. I. Bailey 

 at the Harvard Observatory at Arequipa, Peru, 

 which resulted in the discovery of those men- 

 tioned below. The whole number of stars ex- 

 amined in the several clusters was 19,050, of 

 which 509 were variable, or 1 variable to 37. 

 This ratio, however, varies greatly in different 

 clusters. While in Messier there are 13, the 

 splendid naked-eye cluster in Hercules shows but 

 2 variables, or only 1 in 500. Messier 3 showed 

 132 in 900, or 1 in 7. No cause has been as- 

 signed for these extraordinary phenomena that 

 meets with general acceptance.' Prof. E. C. Pick- 

 ering has compiled a table giving the number of 

 variables in 23 clusters, which is published in 

 Popular Astronomy for November, 1898, of which 

 the following are the most remarkable: New 

 General Catalogue 5272 = Messier 3, 132 are va- 

 riables; New General Catalogue 5839 = Omega 

 cluster 125; New General Catalogue 5904=: Mes- 

 sier 5, 85; New General Catalogue 7078 = Mes- 

 sier 15, 51, are variables. The Omega Centauri 

 cluster, just visible without a telescope, surpasses 

 in number of stars in a given area all others 

 known. Its position is right ascension 13& 26m 

 46 s , declination south 46 47'. Although it is much 

 less than some others in extent, yet 8,000 stars 

 have been counted on a photograph plate, but the 

 number actually visible is very much greater. 

 The periods of 106 of the 125 variables in the 

 Centauri cluster have been determined; 98 have 

 periods less than twenty-four hours. The long- 

 est is four hundred and seventy-five days, the 

 shortest six hours and eleven minutes. The 

 largest range in variation is about five magni- 

 tudes, and no star has been included whose light 

 changes do not amount to half a magnitude. In 

 one star whose period is fourteen hours and eight 

 minutes, the rise from minimum to maximum, 

 a change of two magnitudes, takes place in one 

 hour. Eleven of the clusters investigated have 

 11,980 stars and 462 variables, or 1 in 26. 



Double Stars. A large number of the naked- 

 eye stars are found, when examined by the tele- 

 scope, to be double or triple. Those that revolve 

 around each other are called binaries, of which 

 Castor is a familiar example. From 1877 the 

 period of this interesting binary was considered 

 to be one thousand and one years; but this the- 

 ory has been completely upset by the fact that 

 since 1887 the two components have been steadily 

 approaching each other. Prof. Doberck, of Hong- 

 Kong Observatory, has recently computed a new 

 orbit from all available observations. He makes 

 the time of perihelion passage A. D. 1948.86, and 

 the period only 318.23 years. 



The line must be sharply drawn between tele- 

 scopic and spectroscopic doubles. A star may 

 be telescopically double and not be a binary, one 

 star happening to be almost exactly behind the 

 other. Not until orbital motion is detected be- 

 yond all doubt can the star be pronounced a 

 binary. On the other hand, a spectroscopic 

 double is assuredly a binary, for its orbital mo- 

 tion is what determines it to be a binary double. 

 A spectroscopic double never can be seen double 

 by any telescope, so close together are the com- 

 ponents. When the components of such a pair 

 revolve around each other, their orbital plane 

 being coincident with our line of sight, the visible 



star (the other supposed to be dark) must al- 

 ternately approach the Earth and recede from 

 it by the attraction of the dark one. It is this 

 reversal of the motion of light that the spectro- 

 scope takes cognizance of. Several such spec- 

 troscopic doubles have been lately discovered, 

 the latest find being the pole star. This polar 

 pilot has a minute star quite close to it visible 

 in small telescopes, but there is no reason to 

 suppose that they are physically connected. Po- 

 laris (as the pole star is called) is a spectro- 

 scopic binary, lately discovered to be such by 

 Prof. W. W. Campbell, of Lick Observatory, and 

 its duplicity has been fully confirmed by Edwin 

 B. Frost, at Yerkes Observatory, whose observa- 

 tions and measurements tally almost exactly 

 with those of Prof. Campbell. The latter is of 

 the opinion that it is a ternary system instead 

 of a binary that is, the system is composed of 

 one bright naked-eye star and two dark ones, 

 which make a revolution around the bright one 

 in a little more than three days. The spectro- 

 scope has proved that the pole star is moving 

 toward the Earth at the rate of 9 miles a second. 

 Although Polaris is approaching the Earth, yet 

 its velocity is variable. During two days it is 

 approaching our system at the mean rate of 14 

 kilometres a second, and during the next two 

 days but 8 kilometres. This variable motion is 

 ascertained by the displacement of the lines in 

 the bright star's spectrum. When any star is 

 approaching the Earth its spectral lines are 

 moved slightly toward the violet, but if it is re- 

 ceding the same lines are moved toward the red 

 end of the spectrum. The amount of displace- 

 ment gives the velocity. If the velocity of a star 

 is uniform, whether to or from the Earth, this 

 affords positive evidence that the star is not a 

 binary with their orbital planes coincident or 

 nearly so with our line of sight. The spectroscope 

 takes no cognizance of a binary star if its mo- 

 tion is perpendicular to our line of sight. 



The photographic determination of the motion 

 of a star in the line of sight by the spectroscope 

 is one of the marvels of photo-spectroscopy, 

 which, owing to more rapid plates, can now be 

 obtained by an hour's exposure. The plate and 

 the companion spectrum of a terrestrial source 

 are placed side by side, and the measure of the 

 displacement of the two spectra is then made 

 with the microscope and micrometer. M. Des- 

 landres juxtaposed a terrestrial spectrum with 

 that of Alpha Auriga, and found that the dis- 

 placement of the lines toward the red corre- 

 sponded to a velocity of recession of 43 kilome- 

 tres a second. For the dog star (Sirius) the 

 increase of distance was 18 kilometres a second. 

 Gamma Pegasi was found to be approaching the 

 Earth at the rate of 3 kilometres a second. Beta 

 Auriga showed a doubling of the lines, first de- 

 tected by Prof. Pickering (proving that the star 

 is a spectroscopic double, which no te^scope is 

 able to divide), which indicates velocities of 100 

 kilometres a second, or a little more than 62 

 miles. 



In the progress of the work of a spectro- 

 graphic determination of stellar motion to and 

 from our system several other instances of rapid 

 motion were detected. From four plates of the 

 spectrum of Eta Cephei a mean result was ob- 

 tained showing a velocity of approach of nearly 

 87 kilometres a second, and four plates of the 

 brighter component of Zeta Hcrculis give a ve- 

 locity, also of approach, of 70.3 kilometres a 

 second. Belopolski's result for the latter star is 

 70 kilometres. These results, however, are some- 

 what reduced when corrected for the motion of 



