90 Prof. Liveing, On the probable presence in the Sun 



There are however some remarkable exceptions to this. They are 

 the rays of helium, and the ultra violet-rays of hydrogen which 

 do not appear as Fraiinhofer lines. He does not hesitate to 

 ascribe 4 rays to helium and 24 others to hydrogen in preference 

 to identifying them with the nearest Fraiinhofer lines. And 

 I think he is right in so doing, notwithstanding the fact that 

 several of these rays have never been observed to be emitted 

 by hydrogen, have not been seen in the spectrum of any star, 

 and are believed to be due to hydrogen only because the wave- 

 lengths found for them agree closely with the values found by 

 calculation for the prolongation of the well known series of 

 hydrogen rays. 



It is a noteworthy fact, that whereas it is very difficult 

 to get from a tube containing a residue of carefully purified 

 hydrogen any rays of this series above H$, A 3889, yet when 

 the tubes contain sensible quantities of helium and neon, they 

 readily give many lines of the series. Up to H„, A 3691*5, these 

 rays give very long arcs on Humphreys' films and are there- 

 fore emitted from the upper layer of the chromosphere. Why 

 the presence of helium and neon along with hydrogen should 

 conduce to the development of this series I cannot guess, but 

 whatever the cause it seems to prevail in the sun. Besides the 

 4 helium rays above mentioned, other 3 helium rays might well 

 be identified with arcs on Humphreys' films; A 3613*78 is much 

 nearer in wave-length to Humphreys' ray A 3613*8 than the 

 scandium ray with which he identifies it; and A. 4143*92 is 

 nearer to X 41440 than the iron ray to which it is ascribed; and 

 Humphreys' ray 38207 gives a very long arc on film IV, and is 

 probably the rather strong helium ray X 3819*75. There are two 

 rays A 3388 and A 3456*5 giving arcs very long, broad and hazy, 

 which Humphreys cannot distinguish by measurement from 

 Fraiinhofer lines ascribed to titanium, but doubts whether they 

 are really due to that metal. We have found that the most volatile 

 atmospheric gases, probably neon, give rays at about A, 3388*8 and 

 X 3456*8; the former is rather a strong ray, the latter a weak one. 

 It is possible the solar rays may be identical with them. 



That the same gases must be in the atmospheres of the sun 

 and planets appears to me certain. The subject has been dis- 

 cussed by many, but some of them seem to have gone off the 

 track in ascribing the diffusion of these atmospheres to those 

 molecules which on the kinetic theory chance to acquire such 

 a velocity as will carry them beyond the attraction of the planet 

 with which they had been associated. Dr Bryan has calculated 

 that if we suppose the earth's atmosphere to have a definite 

 surface there is practically no probability that any appreciable 

 amount, of even so volatile a gas as hydrogen, can escape with 



