ME. E. C. C. BALY ON THE SPECTRA OF NEON, KRYPTON AND XENON. 187 
indeed almost continuous. If the pressure is slowly reduced, the lines begin to 
appear upon a more or less continuous background, and, gradually increasing in 
brightness, they finally become quite sharply defined upon a black baclvground. This 
eflect is especially noticeable in the spectra of krypton and xenon, and in filling 
vacuum tubes with these gases, great care must be taken to reduce the pressure 
sufficiently in order to obtain good definition. When a tube has been completely 
exhausted according to the method described above, the connection to the exhaust 
pump is closed and a quantity of the gas in question, known to be in excess of that 
required, is admitted. The electric current is then made to pass through the gas, 
and the spectrum is examined through a small spectroscope ; the stop-cock connecting 
the vacuum tube to the exhaust pump is then opened and the pressure of the gas 
slowly reduced until the spectrum lines appear perfectly sharp, when the tube is 
sealed off. All the spectrum lines do not become equally well defined at once, and 
the lines olven m the tables as diffused are those which still remain somewhat 
o 
nebulous at the pressures dealt with ; these outstanding hazy lines, however, if the 
pressure is sufficiently reduced, tend to improve in definition, although some of them 
only become sharp when the pressure has been so far reduced as to very materially 
lessen the illumination. Since the absorption of the gases liy the electrodes them¬ 
selves and by the volatilized aluminium appears to be more rapid at low pressures, 
the necessarily low initial pressure renders imperative the frequent filling of the tubes. 
In the tables given below there is a column containing the wave-lengths which 
have already been puldished by Runge for krypton, and by Liveing and Dewar for 
all the gases. PtUNGE was unaware of the existence of xenon, and therefore certain 
lines belonging to the spectrum of tliis gas are to be found in his list of krypton 
lines. Another column headed “Remarks” includes memoranda concerning the 
individual lines, and certain points in connection with these merit some attention. 
In the fii'st place, under the second krypton spectrum, there will be found certain 
lines which are visible in the second or blue argon spectrum ; it is interesting to note 
further tliat these lines disappear from the spectrum of argon after that gas has been 
fractionated by means of liquid air. In view of the discovery of these new gases in 
the atmosphere I made a comparison, with a glass prism spectroscope of considerable 
dispersion, between the lilue spectra of ordinary atmospheric argon and of the same 
gas after fi'actionation by means of liquid air."^ Both the spectra were completely 
measured, but the list of wave-lengths is not worth publishing in its entirety on 
account of the very slightly different values found from those already given by 
Kayser and by Eder and Yalenta ; certain lines however were measured which do 
not appear in these lists, and they are given in Table I. In Table II. are given the 
wave-lengths of the lines which are removed from the spectrum of argon by the 
process of fractionation. Three of these lines at \= 4488T4, 4199'97 and 4047‘38 
* I take this oi^porturity of expressing my thanks to Messrs. W. L. St. J. Alton and A. C. Carter 
for their valuahle help in this investigation. 
