SIR WILLIAM HUGGINS, K.C.B., O.M. 185 



of comparison. The stars and the sun could be directly and 

 independently compared with the standards, and thus the 

 similarity or dissimilarity of their spectra could be certainly 

 established. In the preface to the second volume Sir William 

 Huggins says : 



" Looking back with the knowledge of the more efficient 

 and perfectly adapted instruments and methods of work which 

 have been gradually introduced during the last forty years, 

 no one can be more conscious than I am of the inevitable short- 

 comings of my pioneer instruments and methods of work, which 

 had to be created under circumstances of no little difficulty. 

 These shortcomings prevented the attainment of accurate re- 

 sults in some single cases, but time has shown that they did 

 not affect the fundamental general correctness of my early 



V. 



Whilst Huggins and Miller were carrying on their work 

 of detailed examination of a few typical spectra, Secchi devoted 

 himself to passing in review a vast number of stars, searching 

 for information as to the general types of their spectra. He 

 adopted the observational method of Fraunhofer and classified 

 first of all 600 spectra, and later on 4,000 spectra of different 

 stars. He found that for the purposes of a general classification 

 the spectra of all the stars examined might be referred to one 

 or other of four different types. The later work of Pickering 

 and of Vogel and others serves to show that there are stars 

 which exhibit spectra intermediate in character between Secchi's 

 types. But such intermediate spectra are relatively few in 

 number. It would accordingly appear that here are evidences 

 of a process of evolution in which the typical spectra of Secchi 

 indicate stages of stability of some kind. The relatively scarce 

 intermediate spectra are those in which, if anywhere, we may 

 least unreasonably expect to be able to detect change ; but it 

 is hardly necessary to add that the time-intervals in such a 

 process of evolution are far too long for us to have discovered 

 any sign of actual change in the few years in which observa- 

 tions have been made. 



Huggins's detailed work and his series of photographs of 

 stellar spectra, which form the splendid feature of his Atlas 

 of Representative Spectra, have contributed enormously to the 

 advance of the subject. This was the work that he devoted 

 himself to from the year 1875, when he succeeded with splendid 



