262 University Geological Survey of Kansas. 
solve in the liquid methane. When the desired amount of gas 
had passed into A the connection with the gas-holder was 
closed, and after standing for several minutes the stop-cock 
leading into B was opened for a moment, that between B and C 
being closed. The gas was held in B for five minutes to allow 
the absorption of everything except the helium. Stop-cock L 
being closed, the gas was passed into C and allowed to stay 
there another five minutes, and finally admitted to the Pliicker 
tubes and pump. Here it was examined with the spectroscope, 
removed by means of the pump, and collected in J. When the 
pressure in the system became low, stop-cock K was closed and 
more gas admitted into B from A, allowed to stand, passed into 
C, and finally, after standing there the required time, the con- 
nection to the pump was again opened. This entire series of 
operations was repeated until no more helium could be ob- 
tained. It has been repeatedly proven by our experience that 
practically all the helium may be removed by this process. To- 
wards the end of the operation, when the pressure in the entire 
system was low, the liquid methane solution in A could be 
made to boil even when surrounded by liquid air, and this in- 
sured the removal of all the helium from the bulb. It might 
be mentioned, in passing, that the freezing-point of the meth- 
ane was lowered by the nitrogen, etc., so that only part of it 
solidified in A. 
With care all the helium could be transferred to the charcoal 
bulb without permitting so much nitrogen to pass over as to 
saturate the charcoal. The purity of the gas could be insured 
by watching the spectrum, and if there was any suspicion of 
nitrogen the apparatus was prepared as for a new analysis and 
the gas run through again. The operation was tedious, but not 
more so than most gravimetric determinations. 
By essentially the same method, but using larger charcoal 
bulbs, considerable quantities of helium have been obtained, 
and a still larger apparatus is being constructed. 
That the gas obtained in these experiments is helium is 
proven by the fact that ail of the visible lines of helium cata- 
logued have been observed, the wave-lengths agreeing, within 
the limits of error, with the accepted values. This is also es- 
tablished by direct comparison of the spectra with those given 
by known specimens of helium. 
The spectrometer in use for the examination of the gases is 
a new Pulfrich auto-collimating instrument made by Zeiss. It 
