Ejfect of Pressure upon Ave' Speetra. 571 



but afttT a lono- nin the ^vintlows were clouded by a coatino- 

 of tinely divided carbon. Durino- the time of exposure the 

 arc Avas carefully observed by means of a direct-vision spec- 

 troscope, and Avas found to show very distinct alteration in 

 appearance according to the conditions under which it was 

 burnino-. AVhen the arc was short and steady the difference 

 from the spectrum of the normal carbon arc was not very 

 markeil, whereas when the carbons were drawn apart and 

 the arc was on the point of breaking, a spectrum showing 

 more sharply the fluted carbon bands flashed out vividly for 

 a short tune. Several photographs were taken with the 

 object of studying this eftiect, tbe observer screening off the 

 light bv means of a photogra]ihic shutter except when this 

 jihenomenon was observed. The photographs taken at about 

 40 atmospheres pressure show a marked reversal of the five 

 beads of the cyanogen band beginning at 3883. It may be 

 noted that the analysis of the gas indicated that at the time 

 these photographs were taken only a very small percentage 

 of the oxygen remained uncombined. With a view to in- 

 creasing the percentage of cyanogen compounds in the 

 atmosphere, hydrogen was admitted step by step to the fur- 

 nace charged with compressed air, care being naturally 

 tnken to allow time for combination to take place, so that 

 the mixture should always be below the explosive limit. 



AVith such a mixture the reversal was also photographed 

 easily at 16 atmospheres pressure. The presence of hydro- 

 cyanic acid could be readily detected in the residual gases. 



The accompanying table (p. 572) gives details of the 

 conditions under which the experiments were carried out. 



The reversal of the cyanogen bands has been previously 

 observed by Liveing and Dewar"^, but only under special 

 conditions, and when cyanide of titanium or borate of am- 

 monium were introduced into the furnace. As above stated, 

 the photographs taken while the arc was burning normally, 

 in most cases, do not show the reversal of the cyanogen band. 

 A point, however, worthy of notice is that the calcium lines 

 occur in nearly all the photographs, and are generally sharply 

 reversed ; the blue line 4226 appears much more brilliantly 

 than the H and K lines. The appearance of the calcium 

 lines in those spectra which show the reversed cyanogen 

 band is still more curious since here the calcium blue line 

 4226 is very strongly reversed, whereas there is no indication 

 of the presence of the H and K lines. 



* Liveing- and Dewar, Proc. U. S. vol. xxxiii. p. 3 (1881). 



