SCIENCE 



[N. S. Vol. XXIX. No. 752 



count toward the probability of identity, while 

 disagreements do not count against it. I also 

 call attention to the practical agreement in 

 wave-length and in grade of intensity of lines 

 Nos. (4) and (5). A detailed discussion of 

 the probability of identification of each line 

 seems unnecessary here. Admitting the un- 

 certainty in the measurements of the nebular 

 lines, considerable weight can fairly be given 

 to general agreement. I have nothing to say 

 to the spectroscopist who insists on the abso- 

 lute matching of lines in wave-length as the 

 final test of "identification" of all celestial 

 spectra. 



If radioactive outbursts determine both the 

 bright lines and the dark lines, and account 

 for the appearance and disappearance of the 

 former, the celebrated bright lines of Huggins 

 may supply a test of the theory. In the fol- 

 lowing table, the first column contains the 

 estimated wave-lengths of 'bright lines observed 

 in the nebula of Andromeda by Sir William 

 Huggins and Lady Huggins,' the second and 

 third, the corresponding radium emanation 

 lines and their intensities observed by Earn- 

 say, Collie and Cameron, and the fourth and 

 fifth, those observed by Eutherf ord and Eoyds. 



For the sake of simplicity of statement, and 

 60 as to conform to the grade of accuracy of 

 the estimates by the astronomical observers, 

 only the first three figures of the wave-length 

 measurements of the physicists are given. 

 The general agreement is very satisfactory. 



" " An Atlas of Representative Stellar Spectra," 

 p. 125. 



Line (4) is " a group of 4 or 5 lines," and be- 

 comes so, first, because both lines concerned 

 are relatively intense, and secondly, because, 

 as astrophysicists well know, an intense line 

 is at times likely to break up into a group. 

 Line (5) is evidently "suspected" because of 

 decreased sensitiveness of the eye for this 

 region, and because of the low intensity of the 

 three lines concerned. 



A more detailed discussion would, I am 

 sure, place this identification beyond any 

 possibility of doubt, but also carry me farther 

 into the subject than I had planned. It is 

 worth while remembering, however, that ad- 

 mitting radioaction as receiving confirmation 

 in other celestial bodies, it may here, with 

 some weight, predict, as it does, the fact of 

 bright lines, their variability, and their char- 

 acter as radioactive products. The two 

 groups (4) and (5), examined on the basis of 

 the strictest principles of electro-optics would 

 alone prove identity, and their testimony is 

 certainly not weakened by the other lines. 

 The difficulties conquered by the famous 

 chemist, and the masterful physicists in fur- 

 nishing the measured lines of the rare 

 radium emanation are certainly always most 

 gratefully to be acknowledged ; and the refined 

 and conscientious work of Sir William Hug- 

 gins and Lady Huggins in independently ex- 

 amining these evanescent bright lines of the 

 nebula of Andromeda through the weary 

 nights of many successive years, is, I hold, not 

 only emphatically verified by this identifica- 

 tion, but their faith in its high and lasting 

 value adequately commended by the result. 



The spectra of five other spiral nebulte were 

 likewise photogi-aphed by Dr. Fath at the 

 Lick Observatory, and four of these, consist- 

 ing each of but from one to three lines, dis- 

 closed, in each, identity with radium emana- 

 tion in one to two lines. The spiral Nebula 

 N. G. C. 1068 discloses, as stated, no known 

 radio-element of the radium series, unless the 

 bright A 3733 and X 3878 be helium lines. It 

 is, however, practically certain that the cele- 

 brated main nebular lines characterizing the 

 spectrum of this object, and of others like it, 

 are due to radio-elements yet to be identified. 



The spectra of star clusters next claim at- 



