250 REPORT—1905. 
that it still remains the basis of all refined investigation dealing with the 
variation of (3 Lyre and stars of the same peculiar type. But progress 
for full thirty years more ended here. What conditions of motion or 
position, what chemical or physical forces, produced the definite changes in 
brightness described so fully by Goodricke and Argelander remained an 
unbroken secret. 
In 1891 Professor Pickering | from spectroscopic observations made at 
Harvard found /3 Lyre to be composed of two stars revolving round each 
other in a period synchronous with the alternations in brightness of the 
dual system. This discovery was subsequently abundantly confirmed by 
Lockyer,? Belopolsky,* Vogel, and Sidgreaves,® It proved incontestably 
that the variation of /3 Lyre was connected in no indefinite way with the 
movements round one another of the component stars, The immediate 
cause of variation might be eclipse, or tidal ebb and flow, or cycles of heat 
and light waves due to changes in surface pressure. It might be all these 
and much more, but the primal cause of the star’s variation was its 
duplicity ; it varied in brightness because it was a binary system, and 
this was a discovery of no mean value. It meant an advance from a 
region of uncertainty and conjecture to one of certainty and pregnant 
facts. 
The discoveries of Pickering and other spectroscopists were further 
enhanced in value by the detection of new variable stars, both in the 
northern and southern hemispheres, whose light-changes were almost an 
exact replica of those of 3 Lyre. 
Up till 1892 (3 Lyre stood alone, sui generis, a class by itself. Its 
variation, therefore, was regarded as anomalous, and probably due to 
anomalous causes. When, however, observational research revealed the 
fact that 3 Lyre, instead of being the sole representative of a very dis- 
tinct type of variation, was the prototype of an ever-increasing class of 
variable stars, the problem of its variation became forthwith an investi- 
gation of far-reaching importance and scope. 
In 1897 Myers® pushed the investigation to its legitimate issue by 
determining the orbital elements of (3 Ly yre from an examination of m2 
light-curve. 
This was followed by a similar investigation of the variation of the 
northern star U Pegasi.’ 
The writer had meantime been following out the same line of research 
at Lovedale, and a preliminary consideration of the light-fluctuations of 
certain southern stars, especially that of the close binary system V Puppis,® 
indicated the need for more refined methods of observation if the results 
sought were to have any pretensions to accuracy. 
The results sought were the form, figure, size, density of those stars 
that, from their analogy to } Lyre, were binary systems. In addition to 
the determination of these more easily ascertained orbital elements there 
arose out of, and indeed formed part of the investigation, the question of 
1 Astronomische Nachrichten, No. 3051. 
2 Astronomy and Astrophysics, vol. xiii. p. 575. 
8 Mélanges Mathématiques et Astronomiques, tome Vii. p. 423. 
4 Sitzungsberichte, Berlin, February 8, 1894. 
5 Monthly Notices R.A.8., vol. lxiv. p. 168. 
® Astrophysical Journal, vol. vii. p, 1 
7 [bid., vol. viii. p. 163. 
8 Jbid., vol. xili, p. 177. 
