iv "BUNSENIANA" 81 



Lecture given before the Chemical Society. It is of 

 special interest as it is the first communication of the 

 discovery of spectrum analysis by Bunsen and Kirch- 

 hoff. It also gives an idea of the way in which 

 he and I worked together in our photo-chemical 

 researches. 



HEIDELBERG, 



November i^th, 1869. 



My best thanks, dear Roscoe, for your last letter. Your 

 induction experiments appear to me to be very important 

 and interesting. From our former experiments I consider 

 it certain that considerable induction takes place with our 

 rotating disc, and from further experiments which I have 

 made since then I have obtained the same results by exposing 

 the paper on its reverse side ; whilst according to my last 

 experiment the duration of the insolation as compared with 

 that when the light is cut off lasts frequently from -g^j- to 

 20 seconds. If we assume that the induction in the case 

 of the chloride of silver paper passes off very quickly, this 

 would not be noticed in your experiments, whilst with the 

 rotating disc it would become visible. A new apparatus 

 about which I wrote to you will do all that we hoped. 

 Unfortunately the weather is not favourable for me to make 

 experiments ; but I hope at Christmas or Easter to have 

 obtained results for which we have so long striven. At the 

 moment I am occupied by an investigation with Kirchhoff 

 which does not allow us to sleep. Kirchhoff has made a 

 totally unexpected discovery, inasmuch as he has found out 

 the cause for the dark lines in the solar spectrum and can 

 produce these lines artificially intensified both in the solar 

 spectrum and in the continuous spectrum of a flame, their 

 position being identical with that of Frauenhofer's lines. 

 Hence the path is opened for the determination of the 

 chemical composition of the sun and the fixed stars with the 

 same certainty that we can detect chloride of strontium, 

 &c., by our ordinary reagents. By this method the chemical 

 elements occurring upon the earth may also be detected and 

 separated with the same degree of accuracy as upon the sun ; 

 for example : In 20 grams of sea-water the presence of 

 lithium can be shown. In the detection of many elements 

 this method is to be preferred to all previously known 

 processes. A mixture contains Li Ka Na Ba Sr Ca ; all 

 that is necessary is to bring a milligram of this mixture into 



G 



