President's Address. 
51 
stars. Where both components of a binary star are visible, these changes 
readily admit of direct measurement. In other cases the existence of a 
companion is inferred to account for regular periodic changes of position 
of the visible component, though this companion cannot be seen either on 
account of intrinsic want of light, or on account of its close proximity to 
the primary and the consequent incapacity of our telescope to render the 
two visually distinct. 
These changes are of interest as affording evidence of the validity of 
the Newtonian laws in systems other than the solar system. 
The changes to which I have so far referred are changes which affect 
isolated stars or groups of stars, but which do not occur, at least to a 
sensible extent, in the generality of stars. 
I come now to the changes of position due to the earth's orbital motion 
which, on the other hand, may be expected to influence all stars in 
common. Even where, as in the cases I have already spoken of, their 
influence is obscured by orbital motion within the system, when once this 
orbital motion has been thoroughly examined, its laws deduced, and due 
allowance made for it by computation, we might expect to find the effects 
of the earth's motion still apparent. 
The earth in its orbit round the sun approximately describes a circle 
of 186,000,000 miles in diameter, and its successive positions in space at 
intervals of six months are separated from one another by this extent. 
But experience has shown that recurring changes in the relative positions 
of the stars, as viewed at intervals of six months — that is to say from two 
different points of the universe separated by this vast distance — can only 
be detected in the case of a limited number of stars, and then only by the 
application of the most delicate methods of measurement specially designed 
to bring these changes to light. 
To the Cape Observatory and its former director, Henderson (1832-34), 
belongs the credit of first producing reliable evidence of the existence 
of any fixed star, for which these changes could be unmistakably detected, 
and which therefore was not too remote from the solar system to permit 
of its distance being at least roughly determined in comparison with the 
diameter of the earth's orbit. Henderson's discovery has since been 
fully confirmed by later observers, and other stars likely to yield tangible 
results have now been examined. As illustrative, however, of the evasive- 
ness of the quantities sought, and the excessive labour by which only 
they can be derived, though the problem of stellar distance has always 
been in the forefront of astronomical interest, and has attracted the 
attention of several able observers, the number of stars for which well- 
determined parallaxes have been published up to the present day does 
not exceed some 400. This number is quite insignificant in comparison 
even with the number of stars visible to the naked eye without telescopic 
