52 PROCEEDINGS OF THE AMERICAN ACADEMY. 



In addition to its remarkable dispersion, the nitroso, as I shall call 

 it for short, exhibits a most beautiful violet surface color, which can 

 best be exhibited by employing a small cell heated by steam, such as I 

 shall describe later on. 



The substance possesses in addition several other interesting features. 

 It has, for example, in addition to its sharp and narrow band of metallic 

 absorption in the violet, a weaker pair of bands near the end of the ultra- 

 violet which flatten out the dispersion curve, but do not bend it into 

 oppositely directed branches as the stronger band does. Moreover, the 

 substance can be vaporized without decomposition, which makes it pos- 

 sible to compare its optical properties in the three states, — solid, liquid, 

 and gaseous. 



I shall take up in order the dispersion in the visible spectrum, the 

 ultra-violet dispersion, the reflecting power in differeut parts of the 

 spectrum, the principal azimuths and incidences for differeut wave- 

 lengths, the angles of maximum polarization, and the changes in the 

 position of the absorption band which accompany a change of state. 

 The various results will finally be discussed in their bearing on the 

 electro-magnetic theory of dispersion and absorption. 



Dispersion, in the Visible Spectrum. 



As the refractive index of the nitroso changes very rapidly with the 

 temperature, it was necessary in making the determinations of the dis- 

 persion to keep the temperature of the melted substance constant. The 

 point chosen was the solidifying point, as it simplified the experimental 

 conditions. The prism was constructed of a pair of interferometer plates 

 accurately plane-parallel, and was mounted on the table of a spectrom- 

 eter in a small clamp-frame made especially for it. A current of hot 

 air was directed against the prism by means of a bent glass tube, under 

 one end of which a small gas flame was burning. The slit of the spec- 

 trometer was illuminated with approximately monochromatic light fur- 

 nished by a monochromatic illuminator built by Fuess. This extremely 

 useful instrument is not as well known as it deserves to be, and a few 

 words regarding it may not be out of place. It is essentially a small 

 spectroscope with collimator and telescope at right angles. The two 

 prisms, which are enclosed in the body of the instrument, are turned by 

 mean3 of a micrometer screw, from the reading of which the wave- 

 length can at once be determined from the calibration curve of the 

 instrument. The eye-piece cau be removed and a draw tube carrying 



