INVESTIGATION WITH A ROCK-SALT PRISM. 47 



The position of the transmission minimum at 3.39 /x does not vary, 

 however. The very deep band at 8.25 fj., which has almost disappeared 

 in the second curve, is no doubt due to CMJ, since that substance has 

 a maximum at this point. 



Ethane differs from ethylene in having the 2.31 fi and the 3.28 fi bands 

 of the latter shifted to 2.36/* and 3.39/11, respectively, in the former. 

 Some of the following gases show this same shifting for the region of 

 3 /J.. As noted before, for rock-salt dispersion, all carbohydrates studied 

 show an enormous absorption band, varying in position from 3.1 /a to 

 3.7 fi. The same is true for the 1.69 yu, band, using quartz dispersion. 

 Here it is difficult to establish a definite shifting, as is found for homol- 

 ogous compounds at from 3.08/1 for acetylene (CoHg) to 3.39^1 for 

 ethane (CgHe). 



This study of gases was undertaken to learn the behavior of the CHj 

 group in the molecule, as found in ethane. After studying many com- 

 pounds of CH3, simple and complex, having an absorption band at 

 3.43 fx, w^hich was first announced by Julius^ to be due to CH3, it was 

 rather surprising to find that ethane (HC3 — CH3) has its maximum at 

 3.39 IX. In the simple compounds, like CH3I, this band occurs at 3.4 /u.. 



When we compare ethane (CaHg) with benzene (CeHg) we find the 

 latter has its transmission minimum at 3.28/1, which again shows that 

 the structure as well as molecular weight influences the absorption. 



The cell of ethane was partly exhausted, and one part of it was mixed 

 with two parts of acetylene, and, as a result, the band due to acetylene 

 was obliterated, except a slight break at 3 /i, while the band due to both 

 is at 3.3 fi. Lack of time did not permit examination of other regions. 

 The test was not quite a fair one, considering the one gas as an impurity 

 in the other, yet it serves to show that, for the region of 3 /i, where the 

 dispersion of rock salt is still small, reliance upon the occurrence of an 

 absorption band of an impurity, as a means for detecting the impurity, 

 is not permissible. However, at 4.4 /i, where the dispersion is greater, 

 one can detect the presence of CO2 in CO. (See fig, 21.) In fig. 15 

 curve a is for the gas when purified by liquefaction, and curve b for the 

 same gas after washing in fuming H2SO4. 



Butane. C4H10. 



(Cell, 5,7 cm. ; barom., 75.4 cm. ; temp., 22° ; fig. 15.) 



This gas was made^ from ethyl iodide (C2H3I) by pouring it over an 



amalgam made of sodium and mercury. The presence of the mercury 



retards the action of the metallic sodium upon the ethyl iodide. After 



^Julius, loc. cit. 



*Loury : Jahrsber. in Forstchritte der Chemie, p. 397, i860. 



