NOVEMBEE 21, 1913] 



SCIENCE 



719 



ever, there is a difference. Our represen- 

 tation by the two independent star clouds 

 is one of them. Whether this interpretation 

 is the correct one, is a question of evolu- 

 tion of the system and will have to be con- 

 sidered presently. 



Our conclusion will then be in favor of 

 the two-cloud theory; and so, for the sake 

 of greater clearness, I will provisionally 

 continue to use this representation. In 

 reality what will be advanced will not be 

 changed, or but slightly, if we simply start 

 from the observed facts. 



In the study of the history of the sys- 

 tem, we start from what we know, or think 

 we know, about the evolution of the sepa- 

 rate stars. 



- The stars have been classified by Secchi 

 into four spectral classes. We have at 

 present far more elaborate classifications, 

 but for the present purpose Secchi 's classi- 

 fication will do. The stars of the fourth 

 type are so few in number that we may, for 

 the present, neglect them. Part of the first 

 type has later on been separated from the 

 rest; they show the helium lines in their 

 spectrum and are now generally brought 

 to a separate class, the class of the helium 



We will thus consider the four classes: 

 the helium stars, those of the first, second 

 and third types — helium, first, second and 

 third — in which the bulk of all the stars 

 with known spectrum are contained. 



Now, there is much evidence to show that 

 this classification is a natural one. I mean 

 that this order is really an order of evo- 

 lution; the helium stars being the stars of 

 recent birth; while we get to older and 

 older stars as we pass from the helium stars 

 to the first, from the first to the second, and 

 from the second to the third tjT)e. I will 

 adopt this order of evolution in what fol- 

 lows, although well aware of the fact that 

 all astronomers do not agree with me. I 

 feel justified in this course, not only because 



I think it is the opinion of the great major- 

 ity of our eminent spectroscopists, but also 

 because the very facts which I wish to put 

 before you about star streaming strongly 

 confirm it. 



When we wish to penetrate into the his- 

 tory of the system, it seems natural to in- 

 vestigate the problem of star streaming 

 separately for those four classes of stars in 

 the order of their evolution. There are 

 some difficulties, mainly the consequence 

 of scantiness of material. Still, however, 

 even now it has been possible to carry the 

 investigation through in such a way as to 

 establish a couple of facts, and to give clear 

 indications of others. Of these I will con- 

 sider only the two following, about the 

 reality of which I think there can hardly 

 be left any doubt. 



First, the older the stars, the greater the 

 internal velocity, and 



Second, the older the stars, the richer 

 the second stream, at least in comparison 

 with the first stream. 



I wish to consider some of the inferences 

 to which these facts lead us. And in the 

 first place, these facts at once lead us back 

 to the question just now mentioned, about 

 the order of evolution of the individual 

 stars. For this regularity in the increase 

 both of the internal velocity and of the 

 richness of the second stream exist only if 

 we adopt for the order of evolution either 

 the order, helium, first, second and third, or 

 the exact verse order, third, second, first, 

 helium, and in no other arrangement. 



Therefore, with the same right that we 

 expect that all the properties of the stars 

 will change with age, gradually, and not 

 per saltum, with that same right, I think, 

 we conclude that the order of evolution 

 must be helium, first, second, third, or the 

 exact reverse. That it is not just the re- 

 verse is proved by other facts we can not 

 now consider. 



