16 



PROFESSOR A. FOWLER, F.R.S., ON 



and because it has the advantage of being nearer to us, so that 

 it can be analysed more completely than the stars. 



When we reflect that the stars which can be observed or 

 photographed with the large telescopes now in use are to be 

 counted by hundreds of millions, it will be evident that the field 

 of investigation is practically unlimited. It is neither possible 

 nor necessary, however, to observe the spectra of all the stars. 

 The two or three hundred thousand which have actually been 

 observed to the present time may surely be taken as representa- 

 tive examples. 



Our present knowledge of stellar spectra has been made 

 possible through the application of photographic methods. By 

 the use of prismatic cameras with prisms of small angle, it is now 

 possible to record the spectra of hundreds of stars with a single 

 exposure of the photographic plate, and we are indebted to the 

 Harvard College Observatory for a descriptive catalogue of the 

 spectra of many thousands of stars obtained in this manner. 



Such photographs only reveal the more general features of 

 the spectra, but detailed studies of individual stars have been 

 made with more powerful instruments by Huggins, Lockyer, and 

 many other astronomers in various parts of the world. From 

 the specimens exhibited it will be realized that a marvellous 

 amount of detailed information is obtainable by this spectroscopic 

 analysis of even the feeble light of a star, if we are skilled 

 enough to interpret its message. 



All stars are alike in the sense that they are highly-heated 

 self-luminous bodies, but the spectroscope shows that they are 

 not all alike in the character of the light which they emit. As 

 already remarked, many of them are indistinguishable from the 

 sun. 



What we have already learned about the sun is therefore 

 applicable to all stars of this class, and the presence of helium 

 and " coroniurn in such stars, for example, may reasonably be 

 inferred, though there are no direct indications of their 

 presence. 



It was early found that, although there are many stars which 

 are unlike the sun, the number of distinct varieties is by no means 

 crreat. The first systematic investigation of a large number of 

 stellar spectra was made in 1864 by the Italian astronomer 

 Father Secchi, who found that the great majority of the stars 

 could be classed in one or other of four types, numbered from I 

 to IV, which were associated with well-marked differences of 

 colour. This classification still serves to mark the broad features 

 of the different classes of spectra. 



