108 PROCEEDINGS OF THE ACADEMY OF [Feb., 



micron, and have been attributed to the presence of the cerium group 

 of elements, but a comparison of the spectra will demonstrate that 

 this cannot be the case, as not a single band occupies a similar position. 

 Another writer refers to them as "the characteristic spectrum of 

 zircon," which is likewise incorrect, as a large proportion of the varieties 

 of zircon show no trace of them. Neither are they due to the coloring 

 matter which gives various tints to this mineral, for except in the case 

 of the red variety, known as hyacinth, these colors are discharged by 

 heating without affecting the spectrum. Having examined more 

 than fifty varieties and colors of zircon, natural crystals and cut gems, 

 many of them both before and after bleaching by heat, can generalize 

 somewhat as follows: In no case have I seen a trace of absorption 

 bands in the spectra of the red hyacinth variety or in the colorless 

 microscopic crystals which can be picked out of magnetic and gold- 

 bearing sands from many localities. They are generally likewise 

 absent in crystals of brown and gray tints, while in the transparent 

 gem varieties from India and Ceylon, those of the characteristic 

 pinkish-brown and yellow tints sometimes shoAv a complete spectrum, 

 sometimes only a few bands and occasionally none whatever. When 

 bleached by heating, the spectrum is unchanged. The strongest and 

 most distinct absorption bands are shown by those stones more or 

 less green in color which become very pale blue after heating. A 

 similar strong spectrum is shown by the microscopic bluish zircons 

 occurring in Vesuvian lava, but no trace of it in the opaque white 

 varieties from the Laacher See and the Azores. It would therefore 

 seem that the spectrum sometimes shown by this mineral is due to the 

 same cause, whatever it may be, that produces a faint blue tint, 

 whether this is originally visible or veiled by other coloring matter 

 which can be bleached by heat. As at least a few of the bands are 

 almost always visible in the spectra of the varieties of zircon employed 

 as gems, except the red, they afford a means of identifying the stone 

 when set. If loose, its high specific gravity is sufficient for the purpose. 

 Red garnets likewise have a readily distinguished absorption spec- 

 trum, including three broad and one narrow band, centering about 

 wave length .618, .578, .520 and .500 micron. This is unquestionably 

 due to the color, possibly manganese, as it is not shown by garnets of 

 other tints than red, but is remarkably constant in all the varieties 

 used as gems. Examination of over a hundred from all parts of the 

 world, including spessartite, almandite and pyrope, as well as varieties, 

 such as the beautiful "rhodolite" from North Carolina, which include 

 both the almandite and pyrope molecules, showed no important ex- 



