46 INFRA-RED ABSORPTION SPECTRA. 



The first large band is shifted to 3.28 /u,. The large transmission 

 minima at 6.98 fi and 10.5 /x are of interest in connection with the study 

 of CH, groups, and with piperidine. 



For the region beyond 9 fi the pressure had to be reduced to 20 cm. 



Ethane. C2H6. 

 (Cell, 5.7 cm. ; barom., 74.5 cm.; temp., 23°; fig. 15.) 



This was the most difficult gas to be prepared. Unlike ethylene, the 

 chemicals used are more complex, and the reaction can not be con- 

 trolled, so that — besides ethane — ethylene, methane, etc., are formed. 



The ethane gas was generated according to the method first described 

 by Frankland,^ who found a " volume purity of 97.88 per cent, using 

 liquefaction in addition. Equal parts of CoHjI (Kahlbaum) and abso- 

 lute C2H5OH were poured over a zinc-copper couple, when the gases 

 began to evolve immediately. No heat was applied, since that causes 

 greater impurity. The gas passed through bromine water in a U-tube, 

 with liquid bromine at the bottom, to remove the ethylene, then through 

 KOH to remove the Br, and then through a tube of P0O5 to remove the 

 water. Since ethane boils at — 93° and ethylene at — 105°, fractional 

 liquefaction and distillation is an uncertain method of purification. In 

 this case the bromine water offers the most serviceable means of sepa- 

 ration. Methane will not liquefy till cooled to — 160°, and hence will 

 pass through the bulb without being condensed. The same is true of 

 the free hydrogen. 



After fractional liquefaction (about 6 cc.) and fractional distillation, 

 the first and second fractions of the distillate were again washed in 

 fuming sulphuric acid and KOH, and dried in PaOg, it being thought 

 preferable to run the risk of letting in air in doing so, rather than be 

 uncertain about the presence of ethylene, carbon dioxide, etc. 



The analyses were not satisfactory. The second fraction was wasted 

 in two attempts at combustion with oxygen, there being a violent explo- 

 sion each time. The explosion-pipette method is less accurate on 

 account of the small quantity (5.4 cc.) of gas that can be used. How- 

 ever, three separate analyses of the method gave a fairly concordant 

 purity of 96 per cent. This is the first gas noticed in which the absorp- 

 tion bands decrease in intensity or disappear entirely, which is an excel- 

 lent clue to the impurities present. 



The change wrought in the transmission curves is most conspicuous 

 at 10.5 /A (fig. 15). That transmission minimum has entirely disap- 

 peared, and the bands at 12 /x, which were obscured by it, now appear 

 in their proper intensities. 



^Frankland : Jour. Chem. Soc. London, 47, p. 236, li 



