INTENSITY RELATIONS IN THE SPECTRUM OF'HELIUM. 
173 
that the addition of a trace of Hydrogen a.ffects both the Diffuse and the Sharp series 
to a comparable extent, whilst the inverse effect produced, by a larger quantity of 
the lighter gas, affects the Diffuse series to a much greater extent than the Sharp 
series. 
Finally, we may refer again to the qualitative results, which show that the seat of 
maximum emission is widely different for lines of Helium, Hydrogen and Mercury, 
and can be very strikingly seen in spite of the heterogeneous nature of the excitation 
in our tubes. An explanation of the apparent distribution of the elements in celestial 
bodies upon such a basis might be worthy of consideration, but the experimental 
evidence seems hardly sufficient to justify such an extrapolation. 
