Researches in Solar Chemistry. 171 



spaces or flutings. Perhaps it will be convenient that I should 

 throw one of these spectra on the screen, and point out exactly 

 the difference to which I refer. I will first call attention to a 

 line spectrum. Those lines are due to the vibrations of mole- 

 cules of calcium and aluminium. The flutings which I now 

 throw on the screen are perfectly different in appearance ; 

 in this case they have been produced by the vibrations of 

 carbon at exactly the same temperature at which we get the 

 line spectrum from aluminium and calcium. 



Now, while we got these thirty-three metals to give us line 

 spectra coincident with Fraunhofer lines, the only evidence 

 (very doubtful evidence) of the existence of the metalloids in 

 the sun at all, depended on the fact that, in the case of iodine 

 and chlorine, some of the channelled spaces observed in their 

 spectra at a very low temperature were imagined to be traced 

 among the Fraunhofer lines in the spectrum of the sun. It 

 is four years ago since evidence was gathered of a more con- 

 clusive kind in the case of carbon. The kind of evidence will 

 be sufficiently indicated by throwing a comparison of the solar 

 and carbon spectra on the screen. Below we have the bright 

 flutings due to carbon-vapour ; and above the solar spectrum 

 this photograph includes a part in the ultra violet. When this 

 negative is placed under a magnifying-glass, we find that 

 most of the very delicate lines constituting the fluting in the 

 bright portion have their exact equivalents among the Fraun- 

 hofer lines. This is the best-established piece of evidence, so 

 far as I know, which seems to indicate that we have truly some 

 of the metalloids present in the atmosphere of the sun by 

 the coincidence of their spectra with the Fraunhofer lines. 

 Further, carbon at all events exists under such conditions that 

 its molecular structure is very much more complex than that 

 of the metals in the reversing layer; and therefore it is pro- 

 bably withdrawn from the excessive heat of the lower region oc- 

 cupied by the reversing layer, which is competent, as we know 

 from from other considerations, to drive even carbon and silicon 

 into the line-stage, supposing carbon and silicon to be there. 



This branch of the work to which I have just referred, a 

 branch which enables us to say that such a temperature must 

 exist in such and such a region of the solar atmosphere, de- 

 pends, in the main, upon questions raised by the differences 

 between the spectra of certain bodies in the sun and in our 

 laboratories. If, for instance, one wishes to observe the coin- 

 cidence between, let us say, iron and the sun, iron is placed 

 in the electric lamp ; its spectrum is photographed : side by 

 side with it we have the spectrum of the sun also photo- 

 graphed ; and, as a rule (I say as a rule ; but this is not abso- 



