i.] THE BIRTH OF SPECTRUM ANALYSIS. 5 



candlebeam is, as I have said before, a very tangled skein, and 

 our knowledge comes from the unravelling of it, which at once 

 lays bare what we may term its structure. 



This unravelling was first described, so far as I know, by 

 Kepler, who used a three-sided prism, that is, a piece of glass 

 shaped as in the diagram, Fig. 1. In his Dioptrics he shows that 

 if a beam of sunlight be allowed to fall on such a prism in a 

 certain way it passes out as a coloured beam the colours of the 

 rainbow being thus artificially reproduced. 



As Newton is generally supposed to have discovered this 

 effect, I give in a note Kepler's own words, from his work 

 entitled Dioptrice, sive demonstratio eorum quce visui et visibilibus 

 propter conspicilla, hoc est, vitra sen crystallos pellucidos, accidunt. 1 



A. period of ^0 years now elapsed, during which nothing 

 more, that we know of, was done in this matter ; till Newton 

 took up and extended the researches of Kepler, and by 

 reason and experiment greatly enlarged our knowledge. 

 Indeed, from his labours, the birth of spectrum analysis may 

 be said to date. 



Newton made a series of experiments which have since 



1 " xvi. Colores iridis jucundissimi oriuntur cum refractio est tanta ; idque 

 tarn si oculi transpiciant quam si Scl transluceat. 



" XVTI. Sole prisma irradiante tria genera radiorum resultant. Sincerus, 

 vitri colore, et iridis coloribus. 



" Sit enim F Sol. Is radiet in D. Hie quasi dividitur radii Solaris densitas, 

 quse minima sui parte repercutitur in D 7, et anguli, ADI, equali ipsi B D F 

 quo illabitur. Sincerum igitur radium, sed tenuem per D I vibrat in /. Sincerus 

 est, quia in vitro tiuctus non est, cujus corpus non ingreditur. 



" Potior autem pars de densitate ipsius F D penetrat D et refringitur in J) E. 

 In E vero rursum dividitur, ratione densitatis. Potior enim pars transit E, et 

 propter geminam magnam refractionem colores iridis jaculatur in Gf. 



"Residuum ipsius D E tentie admodum repercutitur a superficie A (7 in E M : 

 quod si D E paulo obliquius in A E incidit obliquius igitur in E M, refringitur 

 quam hie. Nam si minuas D E A erit et minuendus MEG, ex lege repercussus. 

 Et sic denique E M in B C rectus incidet, itaque nihil in M refnngetur. Cum 

 autem F D hoc pacto bis pertransieiit corpus vitri, quippe semel in D JZ, iterum 

 in E M, exiens recta per J/, radium vitri colore jaculatur in K rectius tamen 

 e rcgione ip.sius A. Nam docemur ex optieis, radios lucidcs tingi in mediis 

 coloratis." 



