Researches in Solar Chemistry. 1(53 



of only a few years ago. It is now possible to record every 

 change which goes on on the sun down to a region so small 

 that one hardly likes to challenge belief by mentioning it. 

 Changes over regions embracing under one second of angular 

 magnitude in the centre of the sun's disk can now be faith- 

 fully recorded and watched from hour to hour. 



One of the advantages which has come from the introduc- 

 tion of the new apparatus has been the possibility of making 

 maps, on a very large scale, of the solar lines and of the me- 

 tallic lines which have to be compared with them. Thanks to 

 the great generosity of Mr. Rutherfurd, who is making the most 

 magnificent refraction-gratings which have ever been seen, 

 and who is spreading them broadcast among all workers in 

 science, one has now easy means of obtaining with inexpen- 

 sive apparatus a spectrum of the sun, and of mapping it on 

 such a scale that the full magnification of the fine line of 

 light which is allowed to come through the slit will form a 

 spectrum the half of a furlong long : an entire spectrum on 

 this scale, when complete (as I hope it some day will be, 

 though certainly not in our time) from the ultra-violet, already 

 mapped by Mascart and Cornu, to the ultra-red, which has 

 quite recently for the first time been brought under our ken 

 by Captain Abney, will be 315 feet long. This is a consider- 

 able scale to apply to the investigation of these problems ; but 

 recent work has shown that, gigantic as the scale is, it is really 

 not beyond what is required for honest patient work. I have 

 already had an opportunity of bringing before the Physical 

 Society several of the methods in use for comparing the spectra 

 of the various elementary bodies with that of the sun. It is not, 

 therefore, necessary now to refer to them. There are, however, 

 others of recent application which are of very considerable 

 importance. 



When, instead of inquiring into the coincidence of the me- 

 tallic lines, we wish to determine the coincidence of the lines 

 due to various gases, the method hitherto employed has been 

 to enclose the gases in Geissler tubes, to reduce their pressure, 

 and in that way to fine down the lines. The importance of 

 this apparently small matter can be very well demonstrated by 

 an experiment easily arranged in an electric lamp, which sodium 

 enables us to perform without any great difficulty. The point 

 of this experiment is that, if we vary the density of any vapour, 

 we vary sometimes to a very considerable extent the thickness 

 and intensity of the lines. I am about to throw the spectrum 

 on a small screen which I have behind the lamp ; and I hope 

 I shall succeed in rendering the phenomena visible. I want 

 you to observe the variation in the thickness of the reversed 



M2 



