286 



KNOWLEDGE. 



[December 1, 1896. 



lines were weak. On the other hand, the spectra of 

 nebulir, rich in lines of origin unknown until idontified 

 with those of the clcveite gases, appear to give all the 

 lines of helium within the region thoroughly explored, 

 but only some of its companion element. 



The striking peculiarity of the ])., line, that it is not 

 ordinarily seen dark in the spectrum of the solar disc, is 

 shared by the other lines whose origin has been revealed 

 to us by Prof, llamsay's discovery. We might reason- 



ably expect, therefore, that those stars where the photo- 

 sphere alone reveals itself, but not the chromosphere, 

 would fail to show us any indication of these gases. And, 

 accordingly, the Sirian and solar stars, Secchi's first 

 and second types, as a rule, show no indications of the 

 spectra we are considering, though, of course, hydrogen 

 is strongly marked. The two types of fluted spectra, 

 those like a. Ilerculis and the red star type, give 

 little or no indication of hydrogen ordinarily, and hence 

 we should scarcely look to them for the two new gases. 

 But directly we deal with the bright line stars — stars, that 

 is to say, where the chromosphere is able to make its 

 presence felt — then helium and parhelium at once begin to 

 show themselves, p Lyne, y Cassiopeiae, P Cygni, are 

 rich in both gases. In /3 Lyrse, indeed, the hydrogen 

 lines are outnumbered by those of helium, and parhelium 

 is strongly represented. The Wolf-Rayet stars, on the 

 contrary, are not prolific in the lines of either of the new 

 elements, four only having yet been identified, all 

 apparently belonging to helium. 



One subdivision of Secchi's first type shows the helium 

 lines dark on a bright ground. These are the stars in the 

 constellation of Orion, the spectra of which have long been 



known to be, to so great an extent, negatives of the 

 nebular spectrum ; indeed, the lino X 4172, one of the chief 

 helium lines, has been known for some time as, par 

 excellence, the Orion line, from its prominence in these 

 spectra. And a careful scrutiny of some hundred and 

 fifty of the brighter stars has yielded to Profs. Vogel and 

 Scheiner about thirty examples of helium spectra outside 

 the constellation of Orion ; Spica, Algol, /3 Tauri, ij Urs83 

 Majoris, y Pegasi being among the number.* Parhelium 

 is also represented in these stars, but 

 less fully. 



To sum up, the spectrum of the new 

 gas proves divisible into two parts, each 

 analogous to the complete spectrum of 

 a distinct element. The behaviour of 

 the two spectra in the laboratory, in 

 the sun, and the other celestial bodies, 

 strongly suggests that they belong to 

 two distinct elements — elements evi- 

 dently only less light than hydrogen, 

 and having a very similar distribution 

 in nature. 



But there is still something to he 

 said on the other side. The two gases 

 have not yet been separated, and the 

 various sources from which they have 

 been prepared have given so nearly the 

 same density for the derived gas — the 

 gas from samarskite giving a density 

 of 2-118, and that from broggerite of 

 2-181 — as to show that, if they are 

 really two distinct elements, there must 

 be little difference in density between 

 them, and their intermixture in nature 

 must be peculiarly thorough. 



There are other points to note. It 

 has been found that by merely varying 

 the pressure it is possible to diminish 

 the brightness of the entire helium 

 series of lines as compared with the 

 entire series of parhelium. Green 

 vacuum tubes — that is, tubes wherein 

 the parhelium giant A 5016 is pre- 

 dominant — can be prepared in this 

 way, as well as yellow tubes where 

 the helium line Dj reigns supreme. 

 More curious still. Prof. Ramsay has found that by 

 allowing the gas to diffuse slowly into a vacuum, it was 

 possible to obtain from it a portion of density as low as 

 1-874 and another as high as 2-133, but in this case the 

 spectrum of the lightest portion was identical with that of 

 the densest. This result appears so puzzling that Prof. 

 Ramsay raises the question as to whether we have the 

 right to assume that the molecules of a gas are homo- 

 geneous, and suggests the possibility that he may have 

 separated between the lighter and heavier molecules of 

 a single element.! 



We may not, therefore, consider that the existence of 

 the two distinct elements, helium and parhelium, is as 

 yet fully proved. But we certainly may take it as very 

 probable, and in this case Profs. Runge and Paschen give 

 a hint as to the probable density of helium. It will be 

 remembered that D3 proved to be a close double, and, so 

 far as can be ascertained, the entire helium spectrum is 

 one of close doublets, whilst that of parhelium is of single 



* Vogel, "On the Occui-reiice in Stellar Spectra of the Lines of 

 Cluveite Oras." As/rop/ii/sical Juuniaf, Vol. II., \o. .5, p. .'333. 



t Ramsay and CoUie, -' The Homogeneitv of Helium and Argon." 

 Froc. Hoyal Society, Vol. LX., No 362, p. "205. 



