534 



SCIENCE. 



[N. S. Vol. II. No. 43. 



uranium minerals cleveite, brdggerite and 

 Connecticut uraninite. Eamsay has since 

 detected it in small amounts in several 

 other minerals and in the gases occluded 

 or combined in certain meteoric irons. The 

 uranium minerals give the best yield, but 

 it is also found in smaller quantities in 

 samarskite, orangite and mouazite, and in 

 traces in yttrotantalite, hjelmite, ferguson- 

 ite, tantalite, polj^crase and xenotime. All 

 of these, including the uranium minerals 

 above mentioned, are of comparatively rare 

 occurrence and are found in small quanti- 

 ties only. They are all of complex com- 

 position and each contains several so- 

 called ' rare earths.' It is a noteworthj' 

 fact that all minerals thus far found to con- 

 tain helium consist in part of one or more 

 of the elements uranium, thorium and yt- 

 trium, and it would seem that it is in some 

 way associated with these, especially with 

 uranium. As to the nature of this associ- 

 ation nothing whatever is known, and at- 

 tempts to cause them to recombine with 

 helium have thus far failed. 



Helium has also been found in the gases 

 escaping from certain mineral springs. 

 Kayser, of Bonn, has detected it in the gas 

 from the Wildbad spring in the Black 

 Forest. After sparking with oxj'gen to re- 

 move nitrogen, which is the chief constit- 

 uent, 340 c. c. of the gas left a residue of 9 

 c. c, which gave a brilliant helium spectrum. 

 Bouchard has found it associated with nitro- 

 gen and sometimes also with argon, in the 

 gas obtained from several sulphur springs 

 in the Pyrenees. As it is thus constantly 

 escaping; we might well expect to detect it 

 in traces in the atmosphere, and Kaj'ser 

 claims to have observed faint helium lines 

 in the spectrum of the purest argon obtained 

 from the atmosphere of Bonn. In fact, it 

 seems to be a very widely distributed sub- 

 stance. 



Cleve and Langlet have obtained helium 

 with a density as low as 2.02, or about 



double that of hydrogen; it is, therefore, 

 next to hydrogen, the lightest gas known. 

 This figure corresponds to a molecular 

 weight of 4.04, and since the molecules of 

 the gas appear to consist of single atoms, 

 like those of argon, 4.04 represents the ap- 

 proximate atomic weight also. The im- 

 portance of this observation lies in the fact 

 that, although our present classification of 

 the elements would lead us to infer the ex- 

 istence of several elements with atomic 

 weights between those of hydrogen (at. wt. 

 1) and lithium (at. wt. 7), none of these 

 are known. Helium may, therefore, well 

 be one of these missing substances. 



Careful spectroscopic studies by Crookes, 

 Lockyer, Runge and Paschen, and others, 

 have shown, however, that what we now 

 call helium is not a single substance, but a 

 mixture of two or more hitherto unknown 

 gases. The composition of the mixture 

 varies both qualitatively and quantitatively 

 according to the source from which it is ob- 

 tained. Crookes has published a very elab- 

 orate study of the spectra of samples of 

 helium from different varieties of uraninite, 

 from which it appears that even these are 

 not absolutely identical. Sixteen bright 

 lines are mentioned as common to all these 

 samples of helium, including the famous 

 D3 line, first seen by Lockyer and Frank- 

 land in the solar spectrum. The gas from 

 Connecticut uraninite seems to be the most 

 complex and shows fourteen strong lines 

 which are absent from the spectra of other 

 varieties and presumably belong to another 

 element. Twenty-seven lines are mentioned 

 which seem to coincide with lines of the 

 solar spectrum. 



Runge and Paschen, who have studied 

 the spectrum of the gas from cleveite, in- 

 cluding the invisible ultra-red portion, have 

 reached the conclusion that even this gas 

 consists of two substances, one of which, 

 the true helium, giving the D3 line, is 

 denser than the other (unnamed) constitu- 



