76 INVISIBLE LINES IN THE SUN'S SPECTRUM. [MEMOIR III. 



those on the sensitive surfaces as to position and magni- 

 tude with considerable accuracy. 



In order to identify these lines I have made use of 

 the map of the spectrum published by Professor Powell 

 in the Report of the British Association for 1839. With 

 the apparatus, as above described, they are exceedingly 

 distinct ; no difficulty arises in the identification of the 

 more prominent ones. The spectrum with which I have 

 worked occupies upon the screen a space of nearly four 

 inches and a quarter in length from the red to the violet, 

 or, more correctly speaking, from the ray marked in that 

 map A to the one marked ~k. In stating, however, that 

 no difficulty arises in identifying these lines, I ought to 

 add that I am referring to that particular map. In 

 the figure annexed to Sir John Herschel's " Treatise on 

 Light," in the "Encyclopaedia Metropolitans," the rays 

 marked Gr seem to differ from that in the report. But 

 Professor Powell's map being drawn from his personal 

 observations, with reference to these very difficulties, and 

 as it agrees with my own observations and measures, I 

 have employed it, and therefore take the letters he gives. 

 (I may add that in all the earlier of these memoirs his 

 nomenclature of the fixed lines is used. It differs a lit- 

 tle from that now employed by spectroscopists.) 



It will be understood that the whole spectrum and all 

 its lines cannot be obtained at one impression. The dif- 

 ficulty is that the different regions of the spectrum act 

 with different power in producing the proper effect. 

 Thus, if on common yellow iodide of silver the attempt 

 were made to obtain all the lines at one trial, it would 

 be found that the blue region would have passed to a 

 state of high solarization, and that all its fine lines were 

 extinguished by being overdone long before any well- 

 marked action could be traced in the less refrangible ex- 

 tremity. It is necessary, therefore, to examine the differ- 



