330 Spectroscopic Examination, of Silicates. 



H2 O 1-48 = 99*70. But this mineral, containing over 1 

 per cent, of potassa, gave no potassium band. The ortho- 

 clases and oligochises which I have thus far examined 

 undergo decomposition hy this process and reveal potassium. 



The hexagonal biotite from Lake Laach gives a faint 

 potassium line, while an analysis (Dana's j\Iin., p. 305) 

 shows that it contains 8-GO per cent, of potassa. The ex- 

 amination of various altered pyroxenes indicated no potassa, 

 so that if thermal waters containing potassa in solution were 

 in some instances the agents of decomposition, they have 

 left no traces which this process can detect. 



In conclusion, it may perhaps be fairly said, that while the 

 calcic chloride process, for a spectroscopic examination, is a 

 most convenient and valuable one for detecting the presence 

 of substances, other than calcium and sodium, which afford 

 spectra, yet it does not equally well prove their absence, 

 since they may be present inl)odies which resist decomposi- 

 tion in this manner. 



In this case, however, a preliminary spectroscopic exam- 

 ination would be of value, as aft'ording an indication of the 

 amount of reliance to be j^laced upon a quantitative deter- 

 mination of the alkalies, by calcic chloride alone. 



Finally, my thanks are due to Mr. F. E. Hilgard, of the 

 first class of the Institute, for his painstaking examination of 

 a large number of minerals by this process. 



