NO. 5 THE MICROSPECTROSCOPE IN MINERALOGY WHERRY J 



dium content. Many artificial salts of the former metal, as well 

 as the chlorite minerals colored violet by it, show spectrum A, so it 

 may be considered proved that one factor in the color of magnesium 

 (and manganese) garnets is the element chromium. (Calcium gar- 

 nets", which are colored green by this element, show an entirely 

 different spectrum.) Spectrum B, it must be admitted, has never 

 been observed in minerals or artificial compounds of vanadium, but 

 no other silicates containing vanadium as a red compound have been 

 available for study (the green roscoelite showing no bands) and as. 

 the mode of combination has great influence on the character of the 

 spectra shown by a given element, it may be regarded as probable 

 that vanadium is a second factor in the color of garnets. The total 

 manganese shows no connection with the spectrum, and the presence 

 of more or less ferrous iron in all garnets precludes the possibility of 

 the existence of any manganic compound. 



TABLES 



The results of the examination of about 200 minerals with the 

 microspectroscope are here presented in tabular form. Only a third 

 of them exhibit distinctive spectra, but as the absence of bands may 

 also have diagnostic value in some cases, it has seemed best to list 

 all those tried. The wave lengths of bands which are especially 

 characteristic of the various minerals are given in bold face type, 

 and of those which are faint and difficult to see in parentheses. The 

 limits of visibility (recorded as " To 700, 440 on," etc.) vary rather 

 widely with the thickness of mineral through which the light passes, 

 but are added for the sake of completeness. 



To increase the practical usefulness of the tables a determinative 

 table, or analytical key, is added after the lists of mineral spectra. 

 It is based on general character of spectrum, number of bands and 

 mineral colors, and covers all minerals showing bands of sufficient 

 intensity for diagnostic purposes. 



Finally, as this method may also prove useful for demonstrating 

 the presence or absence of certain chemical elements, a table of the 

 elements showing spectra, with their forms, and the limits to the 

 amounts present, is also given. It should be noted here that the ele- 

 ments causing the colors and absorption bands of some minerals are 

 as yet unknown; thus, the band at or near wave length 605 in the 

 rare earth minerals with the yttrium group in excess cannot be 

 ascribed to any known element ; and in the other tables interrogation 

 points ( ?) in the " coloring elements " columns show the lack of 

 information in many cases. 



