22 THE DETERMINATION OF LITHIUM. 



tiveness of the tests they burned chlorates with milk sugar to diffuse 

 the vapors through the air and where the chlorate of a metal was 

 not easily obtained they used some other salt with the chlorate of 

 another metal not so easily detected by the spectroscope. Three 

 milhgrams of sodium chlorate burned with milk sugar in a room 

 of about 60 cubic meters capacity caused a Bunsen burner in the 

 room to show the sodium line in the spectroscope for 10 minutes. 

 Considering that 1 second was enough time to make certain of 

 seeing the line, they figured out that it rec|uired 1/3,000,000 mg of 

 the sodium chlorate, which would be 0.00000006 mg of sodium. 

 In a similar manner they determined that the least amount which 

 they could see of lithium was 0.000002 mg, potassium 0.0003 mg, 

 strontium 0.00003 mg, calcium 0.00002 mg, and barium 0.0006 

 mg. In some cases the products of the combustion of the salts 

 and milk sugar were chstributed throughout the air of the room 

 by swinging an open umbrella. 



Kirchhoff and Bunsen reported the presence of lithium in a large 

 number of substances, in the ash of various plants, in minerals, and 

 in man}^ waters. They also made spectroscopic analyses of several 

 water residues, obtaining the spectra of sochum, potassium, lithium, 

 and calcium in sea water and strontium in the scale from boUers of 

 ocean steamsliips. They also reported the experiments from which 

 they concluded that the dark lines in the solar spectrum were caused 

 by the absoqjtion of light b}" the incandescent vapors of the sub- 

 stances which of themselves would give bright lines where the dark 

 lines occurred. For introducing licjuids into the flame they used 

 a line platinum vnre with a loop at the end, the wire being slightly 

 flattened at the loop. Substances which were not easil}' volatile 

 they decomposed with hydrochloric acid or else ammonium fluorid 

 and sulphuric acid. In the following year ^ they reported the 

 discovery of rubicUum and caesium in the Durkheim mineral water. 

 They could detect by the spectroscope 0.0002 mg of rubidium 

 chlorid and 0.00005 mg of caesium cUorid when using a drop of 

 solution weigliing 4 mg. 



Brewster ^ m some notes on the liistory of spectrum analysis 

 referred to his former publications ^ on the colored flames and spec- 

 tra of substances and noted that in 1842,^ after trying 124 substances 

 in the flame of gas from oil and oxygen, he described the spectra of 

 many substances. 



SiMMLER,^ in some qualitative studies of spectra of minerals and 

 water residues, described the spectrum of the inner cone of the 



1 Pogg. Ann., 1861, US: 337-381. 



2 Compt. Rend., 18G6, 62: 17-18. 



3 Pogg. Ann., 1836, 38: 61-63; PhD. Mag., 1830, £.- 384; Trans. Edin. PhiJ. Soc, 1833, 12: 519; 544; 545. 

 i Rept. Brit. Assoc, 1842, p. 15. 



' Pogg. Ann., 1862, llo: 242-266; 425. 



