28 Report 8.A.A. Advancement of Science. 



conform to the same mode, amplitude and period of variation two 

 cycles running; still certain unmistakable signs which stamp, without 

 controversy, the long period variable are there. 



There is the longer period of variation — hence the name — a period 

 never less than a few months, usually a year or even more. There is 

 the irregular descent to minimum, as if the star were unwilling to 

 descend in the scale of brightness, and from time to time made futile 

 efforts to retain its lost status. 



Various explanations of long period variation have been offered — 

 electrical disturbances, meteoric rings and condensations, tidal action. 

 Perhaps the most reasonable explanation is that offered b}' Prof. 

 Turner, who suggests that long period variation is a phenomenon of 

 the same class as solar spots. One difficulty, however, in accepting 

 this theory is that spots alone seem inadequate as an explanation of a 

 variation of eight magnitudes, the amplitude of many long period 

 variables. What does a change in brightness of eight magnitudes 

 mean? It means that at a maximum the star is 1200 times brighter 

 than it is at a mininuim. 



As in the case of short period variation, we must just be content 

 for a season to carry on our observations in the hope that one day the 

 secret of long period variation will be fully revealed. 



At Lovedale about seventy long period stars are under constant 

 observation, and over 30,000 obser\ations of these stars have been 

 secured and full}^ reduced. 



The two other classes of variable stars, (4) Irregular and (b) 

 Temporary, may be dismissed in a few \\ords. 



To the fourth class, " Irregular Variables," belong an innumerable 

 number of stars whose light has been observed to vary irregularly — 

 that is, their waxing and waning pulsations obey no law, observe no 

 period, at least that we can discover as yet. 



The fifth and last class embraces those remarkable stars which 

 from time to time suddenly blaze forth in the midnight sky, impelling 

 the astronomer to arduous observation, inciting the unskilled beholder 

 to wonder sometimes and always to fear. 



Various theories have been advanced to explain these ''fateful 

 messengers." Collision, explosion, fiiction, are the three theories that 

 hold the field. 



My own mind, with no uncertain definiteness, bearing in mind 

 the vast number of stars in space, their prodigious velocities, their 

 different directions of motion, is drawn to impact as an explanation of 

 temporary stars, notwithstanding the weighty arguments which my 

 own judgment brings to bear against such a view. 



One objection is that, given all the stars we know to exist in the 

 sky, given tlie velocities which on the average we know they possess, 

 granted that they all mo^•e in any concei\able direction, then by the 

 theory of probability there ought to be only one new star every 

 2,000,000 years. 



liut nature sometimes seems to laugh at our theory of probability, 

 our 1/ to the (»* - 1 ) timers of occurrence. 



