SPECTROSCOPIC DETERMINATION OF LITHIUM. 23 



Biinsen burner Ilaiiio, whicli may interfere with the observation of 

 the spectra desired, but wliich does not necessarily appear. 



MiTSCHERLicH ' in spectrum observations fed the sohition to the 

 flame by means of a wick of phxtinum wires fastened into the bot- 

 tom of a tube containing the solution. In later work ^ he used 

 several other defaces, such as mixing vapor of the substances with 

 a combustible gas and burning the gas with oxygen. He volatil- 

 ized the solid substance or solution by heat or an electric spark, 

 wliile a stream of hydrogen passing through the vessel took up the 

 vapor and, when burned with ox3^gen, gave the spectrum. 



Brassack^ made some experiments on the sensitiveness of the 

 spectroscopic test with the spark. He measured the loss in weight 

 of the points in one hour and calculated the loss per second which 

 he considered the minimum amount of the metal which would give a 

 spectrum. 



Cappel * made a study of the sensitiveness of the spectroscopic 

 tests for various elements. He used a spark from an induction, coil 

 and placed on one of the points a drop of solution weighing 1 mg. 

 By this means he could detect 0.000000025 mg of litliium or one- 

 seventieth the amount which Kirschhoff and Bunsen could detect by 

 the gas flame. He found it necessary to use quantities of rubidium, 

 ca?sium, and potassium six or seven times as great as the least 

 quantities shown in the Bunsen flame. 



Church ^ used the spectroscope to test for lithium in studying 

 the connection between difi"erent bodies of gi'ound water. Wliere 

 a quantity of lithium salt was put in one well, lithium could be 

 detected in the water from a near-by well in a few hours. It is 

 probable that this method would still be better than the applica- 

 tion of common salt followed by a study of the chlorin content of 

 the water in the second well, though it is not so satisfactory as the 

 use of a suitable dj-e. 



Janssen,® in the transactions of the British Association for 1869, 

 referred to some experiments which he made with the sodium spec- 

 trum, using a luminous flame between the Bunsen burner and the 

 slit of liis spectroscope. When the sodium salt was brought into 

 the flame the bright line of sodium could be seen in the continuous 

 spectrum from the luminous flame. He expanded this note some- 

 what and published it '^ because he had suggested to Champion that 

 a quantitative method might be developed from this idea, or else 



1 Pogg. Ann., lS(j2, IW: 499-507. 



2 Pogg. Ann., 1864, 121: 459-488. 



9 Zts. anal. Chem., 1865, i: 87 (a6s^). 



« Pogg. Ann., 1870, 139: 628-639. 



6 Chem. News, 1870, 23: 322. 



« Trans. Brit. Association for Advancement of Science, 1809, p. 68. 



'Compt. rend., 1870, 71: 626, 629. 



