August, 1903. 



KNOWLEDGE. 



177 



enhanced lines, but the view that they are generally 

 associated with high temperature at least seems to liave 

 as much in its favour as any other which has been put 

 forward. 



It lias long been recognised that the Siriau type of 

 spectrum is closely related to the solar type, this connec- 

 tion Ijeing very clearly shown when intermediate types are 

 studied. Such a connecting link is afforded by stars like 

 Procyon, in which the giant hydi-ogen lines of the Sirian 

 type are considerably toned down, while the metallic lines, 

 notably those of calcium and iron, are correspondingly 

 intensified. The perfect sequence from the Sirian, through 

 the Procyonian to the Arcturian type, is clearly shown by 

 the three spectra in Fig. 9. 



Such photographic comparisons at once suggest that 

 some variation in the condition of a Siriau star, without 

 change of tlie nuiterials involved, might modify the spec- 

 trum so that it became of the Procyonian type, while a 

 further change in the same direction might well produce 

 the Arcturian type of spectrum. The idea of a stellar 

 evolution is, in fact, here presented in its greatest simplicity. 



and Sirius respectively, and on this supposition stars 

 of the first and second types may be ai-ranged as 

 follows : — 



Increasing 

 Temperature. 



Decreasing 

 Temperature. 



Proto-Metullic Stars < ' -j-ypg j jj 

 Metallic Stars ... Type 11. 



Cygnian Sirian 



Polarinn Procyonian 



Aldebarian Arcturian 



On this view an Aldebarian star passes through the 

 Polariau and Cyguian stages to a maximum of temperature 

 represented by one of the groups of Orion stars, and then, 

 with decreasing temperature, through stages represented 

 in succession by the Sirian, Procyonian, and Arcturian 

 types. 



If the evolutionary idea be accepted, it follows that 

 the chemical constituents of the Polarian, Cyguian, Sirian 

 and Procyonian stars are identical with those of the 

 Arcturian, and therefore with those of the sun. The 



C* Al Ca Ft m* f, s« Ft SKtA 



Fia. 9.— Photographic Spectra of P Arietis (1), Procyon (2), and Arcturus (3). 



the substances actually represented in the three spectra 

 being identical so far as they go. 



There are good groun<ls for concluding that this change 

 of spectrum from type to type is brought about by change 

 of temperature. Apart from the indications afforded by 

 the relative intensities of the proto-metallic and metallic 

 lines, this conclusion is suggested by the difference in the 

 strength of the violet and ultra-violet radiations in the 

 three types of spectra. Experiments teach that as the 

 temperature of an incandescent body, such as a platinum 

 wire, is increased, the violet and ultra-violet rays are 

 added in greater proportion than those less refrangible. 

 Hence the greater intensity of the violet and ultra-violet 

 radiations of Sirius as compared with Procyon, and of 

 Procyon as compared with Arcturus, points to a rc<luction 

 of temperature in the passage from the Sirian to the 

 Arcturian type in the evolutionary process. 



The consideration of the temperatures of stars, as 

 indicated by the relative strengths of their violet and 

 ultra-violet radiations, leads to another very important 

 result, one which seems to indicate that the evolutionary 

 processes are, i)erhaps, not quite so simple as might appear 

 from the example already given. Sir Norman Lockyer 

 finds that stars of about the same temperature may usually 

 be divided into two groups, in one of which the hydrogen 

 lines are much less strong than in the other, while the 

 additional lines are of greater intensity. He concludes 

 that the first series of stars are still increasing in tempera- 

 ture while the others are decreasing. Thus, Polaris and 

 a Cygui nearly correspond in temperature with Procyon 



argument for the distribution of solar and terrestrial 

 matter throughout the universe, therefore, becomes 

 stronger the more the spectra of the stars are investigated. 

 Some of the principal stars of Type I. and I. -II. are as 

 follow : — ■ 



JUPITER AND HIS MARKINGS. 



By W. F. Dknnino, f.u.a.s. 



The surface phenomena of Jupiter show no present indica- 

 tiou of declining in interest. Nor is it likely that the study 



