of Edinburgh, Session 1870 - 71 . 
457 
titled to any merit in the matter. As a paper by Mr Talbot, on the 
early history of the subject, is to be read this evening, I shall content 
myself for the present with the remark, that, of the two objects of 
spectrum analysis above named, Talbot and Herschel were unques- 
tionably foremost in the enuntiation of the first ; Brewster, Angstrom , 
and especially Stokes and Balfour Stewart, in that of the second. 
Why some of their statements were incomplete or inexact, and 
what was required to complete or to correct them, will be more 
usefully stated after I have given some preliminary explanations. 
Spectrum. — Newton’s fundamental experiment. 
Reason of separation of colours. 
Reason of impurity. 
How to obtain a pure spectrum. 
Object of trying to do so. 
Effect of Additional Prisms. 
Note that the source of light in all these experiments has been carbon 
heated to incandescence by resistance to a powerful current of voltaic 
electricity. 
I. Incandescent solids and liquids give generally a continuous spectrum. 
Its highest radiation, and the amount of radiation of each wave 
length, depend on the temperature. 
Hence the necessity of using the highest temperature we can 
obtain. 
Illustrate by different lengths of platinum wire heated by current. 
II. Gaseous bodies, incandescent, give generally a (limited) number of 
perfectly definite wave lengths (though under certain circumstances 
of pressure, &c., they give a continuous spectrum). The number 
depends for each substance on its temperature and pressure, and their 
appearance is characteristic of the substance. For, under the same 
physical circumstances, we have always the same effect — as, indeed, 
must be assumed to be the case, if we think physics can be studied 
at all. This remark was virtually made by Carnot, and is all that 
was wanting in Talbot’s earliest paper to make it the complete 
statement of this first part of the subject. 
Illustrate by the spectra of the incandescent vapours of 
Thallium, 
Lithium, 
Magnesium, 
Sodium. 
Illustrate the conductivity of the vapour of the latter by the 
increased breadth of the spectrum when it is present ; also by its 
effect in improving the spectra of other substances when a weak 
battery is used. 
