ON OUR KNOWLEDGE OF SPECTRUM ANALYSIS. 295 



celestial bodies. He mentions tbat this line appears vciy strong nnder 

 conditions nnder which the other lead lines are weak. The line is given 

 as a strong line by Lecoq de Boisbandran, who used feeble sparks, and 

 these facts wonld suggest that it is a low-temperature line of lead. But 

 Thalen has already pointed out that the line is only seen in the neigh- 

 bourhood of the electrodes, and it figures as a short line on Lockyer's 

 map. Also Profs. Liveing and Dewar, who have reversed the long lines 

 of lead, do not notice the reversal of this line. 



Ti7i (6100). Salet ' notices that when a hydrogen flame contains a 

 compound of tin, an orange line appears which is apparently coincident 

 with the orange line of Uthium. This line does not figure on any of the 

 maps of the tin spectrum. 



Zinc. Lockyer ^ found that zinc, volatilized in an iron tube, showed 

 by absorption a green line. Liveing and Dewar ^ do not mention having 

 seen this line reversed, but it is very likely the line, 5184, given by Lecoq 

 de Boisbaudran, and proved by him not to be due to an impurity. 



Sodium and Potassium. In the absorption spectra of sodium and 

 potassium, lines appear in the green which were noticed by Roscoe and 

 Schuster,* but referred by them to known metallic lines of these bodies ; 

 but Profs. Liveing and Dewar ^ have pointed out that they are not coin- 

 cident with any known metallic lines. They have determined the wave- 

 lengths for sodium to be 5510, and for potassium, 5730. 



This is the place ^o notice some remarkable phenomena mentioned by 

 Mr. Lockyer.^ He passed a spark through an exhausted tube, below the 

 lower pole, a piece of sodium was heated, and under the experimental 

 conditions mentioned in the paper, the vapour above the sodium arranged 

 itself in layers of different colours. The layer adjoining the sodium was 

 green, and showed the green and red sodium lines without the yellow 

 lines, while the layer above was yellow, and only showed the yellow 

 lines, without the green and red. Mr. Lockyer also could obtain the 

 yellowish-green lines of potassium without the red. In a subsequent 

 paper, Mr. Lockyer ^ has opened out the question whether the spectrum 

 of an element as it separates out of different combinations containing 

 different numbers of atoms of the element in question, is identical or not. 

 This is an important point, but does not come within the range of this 

 report. 



The question has been raised how far the presence of a molecule of a 

 different kind may affect the spectrum of an element. The wide range 

 within which Mr. Lockyer's law of long and short lines, and the elimi- 

 nation of impurities affected thereby, is true, shows that in a great many 

 cases at any rate, the admixture of another atom only alters the spectrum 

 in so far as it gives a greater prominence to the long lines ; but Mr. Lockyer*^ 

 himself had brought forward examples in which the law of long and 

 short lines does not hold. He has obtained a spectrum of iron in which 

 the longest manganese lines were absent, while some shorter ones were 

 strongly represented. The question is connected with the preceding one, 

 and its discussion is not possible at present, as the material is still too 

 scanty. 



Photometric measurements of the relative intensity of different lines 



' Ann. Chim. Phys. xxviii. p. 69 (1873). * p.,.gc_ jioy_ Soc. xxvii. p. 132 (1878). 



2 Proc. Roy. Soc. xxii. p. 372 (1874). " Ibid. xxix. p. 266 (1879). 



' Spectres Lumincux, texte p. 138 (1874). ' Ibid. xxx. p. 31 (1880). 



* Proc. Roy. Soc. xxii. p. 363 (1874). « ibj^j xxviii. p. 157 (1879). 



