Feb., 1916] Spectra of Some Halogen Compounds 115 



somewhat larger diameter were tried but discarded since the 

 current density of the discharge was not sufficiently great to 

 give an intense spectrum. Too small capillaries could not be 

 used because they readily plugged up with vaporized salt. 

 Rubber stoppers connected the capillary with the remaining 

 parts of the tube. 



The form of apparatus used in the attempts made to investi- 

 gate the absorption spectra of the vapors is discussed later in 

 this paper. 



RESULTS. 



A Geryck two-cylinder pump was connected to the spectrum 

 tube and a low pressure, varying in different experiments from 

 1 cm. to 0.1 mm. mercury pressure, maintained. The dis- 

 charge from the induction coil was then started and without 

 heating the tube the spectrum was examined for lines due to the 

 elements of the compound. In some instances there appeared 

 nothing but the air spectrum,* but in a few cases, as zinc 

 and mercuric iodide, some lines of the elements appeared, 

 indicating that the compound dissociated a little at ordinary 

 temperatures under low pressure. As soon, however, as a 

 little heat was applied the air spectrum began to disappear and a 

 spectrum peculiar to the compound appeared. This remained 

 for some minutes, varying with the compound used, but in 

 general, if further heating was discontinued the air spectrum 

 gradually returned and replaced the spectrum of the com- 

 pound; this process could be repeated until finally the lower 

 half of the electrode would become red hot without any spectrum 

 of the compound being in evidence. But if the heating of the 

 tube was carried on rapidly the air spectrum was quickly 

 replaced by lines characteristic of the elements of the com- 

 pound, which in turn gave way to a continuous region or regions 

 with a few lines peculiar to the members of the compound 

 surviving. After heating for some time (usually fifteen to 

 thirty minutes) the spectrum of the compound and its elements 

 was rather quickly replaced by the air spectrum. These 

 changes were so rapid, photographically considered, that few 

 spectrograms have been obtained. The continuous regions 



*A careful study was made of the spectrum given by the empty tube both 

 when cold and when heated. Under both conditions a large number of bands, 

 diffuse toward the violet, were found in the region 4861 to 4142 t. m. and are due 

 to nitrogen. Some lines and narrow bands were measured in the red-yellow 

 region, but thus far it has been impossible to identify them. 



