SOUTH AFRICA MEDAL. XXXV 



—? 



stars, and. in fact, none were known to the great astronomers of 

 the early ages; in short, it was not until 1596 that Fabricius dis- 

 covered the variable star since known as Mira Ceti. and nearly 

 another two centuries elapsed (1782) before Algol was dis- 

 covered to be variable. It is perhaps remarkable that these two 

 variable -tars are at the extremes of the varieties or species into 

 which variables can be separated. The star Mira Ceti varies 

 slowly from about the second magnitude to the eighth in a period 

 of about 332 days — in other words, its brilliancy changes eight 

 hundred-fold in less than a year, and life as we know it on our 

 planet would be impossible if our Sun varied its emission in this 

 ratio. In spite of numerous investigations, powerfully aided by 

 spectroscopic means, the causes of such variations, and more 

 especiallv of its periodicity, remain hidden. At the other end of 

 the scale are the eclipsing or Algol variables. The diminution of 

 light in these cases is short ; thus Algol shines with an invariable 

 light for two days thirteen and a half hours, at the end of which 

 its light diminishes for three and a half hours ; it remains dimmed 

 for fifteen minutes, and in another three and a half hours it is 

 again shining with its usual light. The obvious explanation that 

 the temporary diminution is caused by an eclipsing body has 

 been amply confirmed. With such a result one might have thought 

 the matter was ended, but it is just here that the sagacity of our 

 medallist came in. By means of his careful studies of the rates 

 of diminution and regain of light, he was able to greatly extend 

 our knowledge of Algol systems. Here I cannot do better than 

 quote some of his words concerning the variable star V Puppis 

 which appeared in the Astrophysical Journal of April, 1901 : 



" ' If, however, the two stars of this eclipse variable revolve in close contiguity 

 there must be distortion to such an extent as to modify slightly the form of the 

 light curve. But whatever be the amount of distortion, unless actual contact takes 

 place, there will be a stationary period. 



" ' On the other hand, if the stars are near enough for their mutual attrac- 

 tions to form a nexus between .them, then there will be no stationary period, but 

 the light curve at maximum will be rounded, the sharpness of the curve depending 

 on the oblatene;<s of the stars. 



" ' A simple examination of the light curve of V Puppis indicates that there 

 is no stationary period at either maximum and accordingly we must infer that the 

 two component stars revolve around one another in actual contact. 



" ' In this case there must be considerable distortion in the form of the two 

 stars, especially at the point where the two bodies meet. 



" ' It is not possible, although an attempt has been made, to determine the 

 amount of this distortion ; the conditions of the problem are too complex, the 

 nature of the action of the forces to be considered too indefinite, and the data at 

 our disposal too meagre to enable us to come to a satisfactory conclusion. ' 



" Pear-shaped rotating bodies and rotating fluid masses/ 

 where the pear-shape is elongating to the point of fission have 

 formed the subject of many mathematical investigations by such 

 mathematicians as Poincare, G. H. Darwin, Jeans and others. 

 It is pleasing to know that our medallist has proved the actual 

 existence of such bodies in our universe. 



" Dr. Roberts's first researches on the density of Algol-type 

 stars appear in the Astrophysical Journal for December, 1899, 



