650 Professor A. Fowler [March 31, 



That is, if Rydberg were right, some of the stars ought to show 

 a line in the blue, about wave-length 4G88, and it should be very 

 bright because it would be the first member of a Principal series. 

 The other lines were not to be expected in the stars because they lie 

 beyond the region for which our atmosphere is transparent. A Very 

 prominent line near 4688 — actually at 4686 — was in fact ah'eady 

 well known in some of the special class of AVolf-Rayet stars, where it 

 occurs as a bright line, in company with fainter bright lines of the 

 Pickering series. It occurs also as a bright line in ^ Puppis, and in 

 many of the gaseous nebulae. A very interesting example of its 

 occurrence is in the Ring Xebula in Lyra, where Max Wolf's 

 photographs show it to be restricted to the apparently dark central 

 space, which must accordingly shine with a feeble blue light. 



Rydberg's view as to the constitution of the hydrogen spectrum 

 thus appeared to be amply confirmed by astropliysical spectra. It is 

 true that the new lines had not been obtained in the laboratory, but 

 the numerical evidence of their hydrogen origin seemed strong 

 enough to outweigh this objection. There was thus for many years 

 a pretty general belief that the line 468G and the Pickering lines 

 were due to hydrogen in a particular state which was only realized 

 under the extraordinary conditions which might conceivably prevail 

 in some of the celestial bodies. Sir Xorman Lockyer regarded them 

 as representing one of the last stages of chemical simplification 

 brought about by the action of high temperatures in the hotter stars, 

 and he gave the name " proto-hydrogen " to the supposed simplified 

 form of hydrogen which produced them. The alternative name 

 " cosmic hydrogen " was, I believe, due to Miss Gierke. 



The next step in the development of the story arose out of 

 observations of total eclipses of the sun. A high-level chromospheric 

 line in the vicinity of 4686 was first recorded in some photographs 

 which I took for Sir Xorman Lockyer in West Africa in 1893 ; but 

 Rydberg's calculations had not then been made, and its special 

 significance was not apparent. This line was still better shown in an- 

 other series of photographs which I took under Sir Xorman Lockyer's 

 direction in India in 1898 ; its wave-length was then given as 4686 • 2, 

 and it was noted that it might be identical with Rydberg's calculated 

 line 468>!. It so happened that in order to facilitate the reduction 

 of the eclipse spectra, a number of laboratory comparisons were photo- 

 graphed with the same instrument, and among these were some 

 spectra of helium tubes. Strangely enough, one of these showed a 

 line not previously seen in helium, and it was apparently coincident 

 with the high-level chromospheric line to which I have referred. The 

 wave-length of this line was given by Lockyer and Baxandall a few 

 years later as 4685-97. The tube which gave the mysterious line 

 appears to have got broken, and it was not known' whether the 

 line was due to an unknown impurity, or to some special condition 

 of discharge which had accidentally been dropped upon. 



