July 26, 1912] 



SCIENCE 



107 



one another, and there are undoubtedly 

 such pairs now known as ' ' optically double 

 stars." But Herschel's mode of attack 

 was bound to fail if the seemingly neigh- 

 boring stars were really so, and were linked 

 together by their mutual gravitation. Al- 

 ready as early as 1707 Michel had sug- 

 gested the existence of such true double 

 stars, but it was Herschel who proved their 

 existence. His first catalogues of double 

 stars, published in 1782, contained 203 

 eases of such doublets, and he already sus- 

 pected a community in their motions ex- 

 plicable only by their real association; but 

 by 1802 he had become certain. In many 

 cases the two components of a binary pair 

 were found to be moving in nearly the same 

 direction and at the same speed, but super- 

 posed on this motion of the system as a 

 whole there was an orbital motion of one 

 star round the other. Herschel even lived 

 long enough to see some of his pairs of 

 stars perform half a revolution about one 

 another. 



After his death Savary took the matter 

 one stage further, and showed that the 

 revolution was governed by the laws of 

 gravity, and thereby confirmed the truth 

 of Herschel's belief. Thus the failure to 

 measure the distance of stars led to the 

 proof that gravity reigns amongst the stars 

 as in the solar system. 



Arago thought that of all Herschel's 

 discoveries this was the one that had the 

 greatest future, and his prophecy has 

 proved singularly correct. Every year 

 adds to the number of double stars, whose 

 orbits are now accurately determinable. 

 These systems are found to be very unlike 

 our own solar system, for the component 

 stars are, in many cases, far larger than 

 the sun and revolve about one another in 

 periods which, in various cases, may be 

 either many years or only a few hours. 



The spectroscope has, moreover, added 



enormously to our knowledge, for the speed 

 of approach or recession of a star from the 

 sun can now be determined as so many 

 kilometers per second. Thus that compo- 

 nent of the motion of a star which was 

 concealed from Herschel is now known 

 with the greater certainty. Moreover, being 

 ignorant of the distance of the stars, he 

 could only express the transverse compo- 

 nent of motion in seconds of arc. 



A wonderful corollary also results from 

 the use of the spectroscope, namely, the 

 existence of many stars known as "spectro- 

 scopic binaries." As seen even with the 

 most powerful telescope such a star is a 

 single point of light, but if the spectral 

 lines are duplicated we know that the 

 source of light is double, and that one com- 

 ponent is approaching us and the other 

 receding from us. In this way the orbits 

 and relative masses of these visually insep- 

 arable stars are determinable. The number 

 of known double stars, including both 

 visual and spectroscopic ones, is already 

 large, and Campbell, of Lick Observatory, 

 has expressed his opinion that one star in \ 

 six is double. Some of them revolve so 

 near to one another and in such a plane 

 that they partially eclipse one another as 

 they revolve, and thus produce a winking 

 light like that of a lighthouse. It would 

 seem that we can now even tell something 

 of the shapes of a pair of stars visually 

 inseparable from one another. But I must 

 not go further into this subject, and will 

 only repeat Arago 's saying, that this dis- 

 covery of Herschel's has "le plus d'ave- 

 nir. ' ' 



It is a figure of speech to refer to the 

 stars as fixed, for a large number of them 

 possess a measurable amount of "proper 

 motion ' ' relatively to their neighbors. The 

 existence of double stars was discovered by 

 the observation of their movements, and 

 thus the study of proper motions is linked 



