186 F. W. Very— A Cosmic Cycle. 



centrated at the center and periphery of a body of stellar 

 dimensions shonld change sign without a literal turning inside 

 out of the body. This argument assumes that there are actions 

 going on at solar centers which generate electricity, and that 

 the processes producing the original electrification of stars, 

 formed out of the same ingredients in nearly identical pro- 

 portion, must always give an external electric charge of the 

 same sign. Since there are no exceptions to the order of 

 brightness of the hydrogen lines in the same class, this argu- 

 ment seems warranted. 



If the variability of the white stars found in clusters 

 is an indication of youth, that of the red stars, on the contrary, 

 must be attributed to some change connected with old age. In 

 the younger stars there is tidal action due to close proximity of 

 newly formed and as yet barely sundered masses. Periods of 

 a few days, or even of a few hours, indicate a superabundant 

 vigor. 



The variability of the red stars has for its characteristic 

 feature that the length of the period increases with the redness 

 of the star, the progression from periods of 50 to 500 days 

 being attended by a variation of color from orange to fall red.* 

 Like the geysers, which in youth have a frequency of a few 

 hours, but in old age spout in a period of as many days, this 

 change points to a waning eruptive force. 



Professor Halef finds a close similarity between the 5th and 

 6th divisions when limited regions in the green or the violet 

 regions are compared. This applies both to the number and 

 relative intensity of the lines in these parts. Other regions of 

 the spectrum are entirely different, but here again there is a 

 close resemblance between the spectra of stars in the 5th and 

 4th divisions. Hence there can be no doubt that the 5th is a 

 transition type between the 4th and 6th divisions, and that it 

 is a truly stellar form, and not intermediate between a meteoric 

 swarm and a star. If bright lines reappear in the spectra of 

 the 6th division, as indicated by Professor Hale's observations, 

 reminding us in this respect of the earliest stars, the circum- 

 stance cannot over weigh the immense discrepancy between the 

 energy of the violet rays in the first and the last types. 



There is one other fact which, if it stood alone, I am free to 

 admit, would shake my confidence in the position assigned to 

 the stars of the 6th division. Mr. J. A. Parkhurst finds that 

 stars having spectra of the 6th division congregate in and near 



* Law formulated by Lockyer, "Meteoritic Hypothesis," chap. 43, from the 

 "Catalogue of variable stars," by S. C. Chandler, Astronomical Journal, vol. viii T 

 p. 81, 1888, and from Duner's -'Les Etoiles a Spectres de la III. Classe," Stock- 

 holm, 1884. 



f George E. Hale, Astrophysical Journal, vol. ix, p. 2*73, 1899. 



