10 REPORT — 1887. 



the spectroscopic examination of such a body must indicate the existence 

 of these common ingredients by the appearance in the spark spectra of 

 these elements of identical bright lines. Coincidences of this kind have 

 indeed been observed, but on careful examination these have been shown 

 to be due either to the presence of some one of the other elements as an 

 impurity or to insufficient observational power. This absence of coinci- 

 dent lines admits, however, of two explanations — either that the elements 

 are nob decomposed at the temperature of the electric spark, or, what 

 appears to me a much more improbable supposition, each one of the 

 numbers of bright lines exhibited by every element indicates the existence 

 of a separate constituent, no two of this enormous number being identical. 

 Terrestrial analysis having thus failed to furnish favourable evidence, 

 we are compelled to see if any information is forthcoming from the 

 chemistry of the sun and stars. And here I would remark that it is not 

 my purpose now to dilate on the wonders which this branch of modern 

 science has revealed. It is sufficient to remind you that chemists thus 

 have the means placed at their disposal of ascertaining with certainty the 

 presence of elements well known on this earth in fixed stars so far dis- 

 tant that we are now receiving the light which emanated from them 

 perhaps even thousands of years ago. 



Since Bunsen and Kirchhoff 's original discovery in 1859, the labours 

 of many men of science of all countries have largely increased our know- 

 ledge of the chemical constitution of the sun and stars, and to no one 

 does science owe more in this direction than to Lockyer and Huggins in 

 this country, and to Young in the New England beyond the seas. 

 Lockyer has of late years devoted his attention chiefly to the varying 

 nature of the bright lines seen under different conditions of time and 

 place on the solar surface, and from these observations he has drawn 

 the inference that the matching observed by KirchhoflP between, for 

 instance, the iron lines as seen in our laboratories and those visible in 

 the sun, has fallen to the ground. He further explains this want of 

 uniformity by the fact that at the higher transcendental temperatures of 

 the sun the substance which we know here as iron is resolved into separate 

 components. Other experimentalists, however, while accepting Lockyer's 

 facts as to the variations in the solar spectrum, do not admit his conclu- 

 sions, and would rather explain the phenomena by the well-known difier- 

 ences which occur in the spectra of all the elements when their molecules 

 are subject to change of temperature or change of position. 



Further, arguments in favour of this idea of the evolution of the 

 elements have been adduced from the phenomena presented by the 

 spectra of the fixed stars. It is well known that some of these shine with a 

 white, others with a red, and others again with a blue light ; and the 

 spectroscope, especially under the hands of Huggins, has shown that the 

 chemical constitution of these stars is different. The white stars, of 

 which Sirius may be taken as a type, exhibit a much less complicated 

 spectrum than the orange and the red stars ; the spectra of the latter 



