INVESTIGATION WITH A ROCK-SALT PRISM. 53 



SO intense after the gas had stood a while in the mercury gasometer. 

 Whether this is due to decomposition into HgS and H has not been 

 determined, on account of lack of time. Certain samples tarnished the 

 mercury, while others did not, after standing in the gasometer for some 

 time. 



Sulphur has a band at 7.9 /x, while H^S has a band at 7.8 /t. Julius 

 found the emission band of a sulphur flame at 7.85 /* (H2S?) instead 

 of at 7.4 fi, which is the first large maximum in SOg. The second maxi- 

 mum of SO2 occurs at 8.65 (i. It does not seem possible that in the 

 sulphur flame he observed the mean of these two bands of SOg. 



Ammonia. NHs. 

 (Cell, 6.3 cm., barom., 74.8 cm. ; temp., 22°. 5 ; fig. 25.) 



Made by heating NH4CI and solid KOH and drying over freshly 

 heated CaO. Upon finding several bands in common with those of 

 certain carbohydrates, and that NH2CH3 or NH2C2H5 might be present 

 in the NH4CI, the latter was purified according to the method of Stas,^ 

 by boiling with HNO3 for half a day. The gas was liquefied in order 

 to remove the supposed water band at 2.95 fi. A sample was placed in 

 a combustion pipette containing freshly heated CaO over mercury, for 

 five days, when the band at 2.95 (i was found to be as intense as on pre- 

 vious determinations. Moreover, the absorption band of water was 

 found at the same time at 3.0 fi, which shows that the 2.95 fi band is 

 characteristic of ammonia. It will be noticed elsewhere that this band 

 is to be found in compounds containing amido (NH2) groups, as well 

 as in certain ones containing nitrogen. 



Ammonia is one of the most interesting compounds studied, because 

 of the numerous deep, narrow bands from 9/x to 13 /u,. At 5.7 ju. and 

 7.3 fj. there is evidence of existing bands, but the narrow dispersion of 

 rock salt prevents their being resolved. This is the only compound 

 studied having such a series of maxima, which are, in addition, so regu- 

 larly distributed that it reminds one of the ammonia and the hydrogen 

 spectrum in the optical region. Whether the " law of constant differ- 

 ence " of the wave-numbers is true, or whether the coincidence in the 

 values of certain wave-numbers is simply accidental, is difficiult to decide 

 with the few examples at hand. We have the following examples : 



^Stas: Fresenius Zeit. f. Anal. Chemie, 6, p. 423. 



