ADDRESS. 31 
selves able to form ideas as to the stages by which the present conditions 
have been reached. 
Up to afew years ago there was no evidence that the law of gravi- 
tation extended to the stars, and even now there is nothing to prove the 
transmission of gravity from star to star. But in the neighbourhood of 
many stars the existence of gravity is now as clearly demonstrated as 
within the solar system itself. The telescope has disclosed the double 
character of a large number of stars, and the relative motions of the pairs 
of companions have been observed with the same assiduity as that of the 
planets. When the relative orbit of a pair of binary or double stars is 
examined, it is found that the motion conforms exactly to those laws of 
Kepler which prove that the planets circle round the sun under the action 
of solar gravitation. The success of the hypothesis of stellar gravitation 
has been so complete that astronomers have not hesitated to explain the 
anomalous motion of a seemingly single star by the existence of a dark 
companion ; and it is interesting to know that the more powerful telescopes 
of recent times have disclosed, in at least two cases, a faintly luminous 
companion in the position which had been assigned to it by theory. 
By an extension of the same argument, certain variations in the 
spectra of a considerable number of stars have been pronounced to prove 
them each to be really double, although in general the pair may be so 
distant that they will probably always remain single to our sight. Lastly, 
the variability in the light of other apparently single stars has proved 
them to be really double. A pair of stars may partially or wholly cover 
one another as they revolve in their orbit, and the light of the seemingly 
single star will then be eclipsed, just as a lighthouse winks when the light 
is periodically hidden by a revolving shutter. Exact measurements of the 
character of the variability in the light have rendered it possible not only 
to determine the nature of the orbit described, but even to discover the 
figures and densities of the two components which are fused together by 
the enormous distance of our point of view. This is a branch of astronomy 
to which much careful observation and skilful analysis has been devoted ; 
and I am glad to mention that Alexander Roberts, one of the most 
eminent of the astronomers who have considered the nature of variable 
stars, is a resident in South Africa. 
I must not, however, allow you to suppose that the theory of eclipses 
will serve to explain the variability of all stars, for there are undoubtedly 
others whose periodicity must be explained by something in their internal 
constitution. 
The periods of double stars are extremely various, and naturally those 
of short period have been the first noted ; in times to come others with 
longer and longer periods will certainly be discovered. A leading cha- 
racteristic of all these double stars is that the two companions do not 
differ enormously in mass from one another. In this respect these 
systems present a strongly marked contrast with that of the sun, attended 
as it is by relatively insignificant planets. 
