610 „ REPORT — 1895. 



which are far higher. This confirms the telief that gold is not rightly placed in 

 column 1. 



(3) It is known that the refraction of a salt when dissolved in water is often 

 slightly modified by the proportional amount of the solvent. The author and Mr. 

 Hibbert have recently found that salts of the metallic elements in columns 1 and 2 

 of Mendel(5efF's table show an increased refraction on dilution, those of metals in 

 column 8 a diminished refraction. 



3. A Discussion ' On the Evidence to be Gathered as to the Simple or Cora- 

 pound character of a Gas, from the Constitution of its Spectrum,^ was opened 

 by Professor A. Schuster and Lord Rayleigh, and the following Papers were 

 read : — 



4. The Constiticents of Cleveite Gas. By C. Runge and F. Paschen. 



As the spectrum of the gas contains two sets of lines, each consisting of three 

 ' series,' and no other lines, we may, according to the analogy of other spectra, 

 draw the conclusion that it consists of two, and not more than two, elements. 

 The yellow line D,, belongs to the heavier of the two elements, which therefore 

 should alone be called helium. 



We have separated the two elements to a certain extent by a method of 

 difi"usion, the lighter constituent streaming more easily through a plug of asbestos. 

 It was shown that the lines in the visible and in the ultra-red part of the 

 spectrum ascribed to the heavier constituent are less intense relatively to the other 

 lines the earlier the stream of the gas is cut off. 



The same conclusion that the gas consists of two elements may also be drawn, 

 first from the spectrum of the sun's limb, where the stronger lines of the heavier 

 constituent are always present, while the stronsfer lines of the lighter constituent 

 are only seen once in every four times. On the other hand in the spectrum of 

 Nova AurigiB at its first appearance we have the opposite case, the lines of the 

 lighter constituent being far more prominent. 



On Motions competent to 2^^'0(iuce Groups of Lines which have been 

 observed in Actual Spectra. By G, Johnstone Stoney, M.A., JJ.Sc, 

 F.E.S. 



In most of the spectra that consist of lines very remarkable groups present 

 themselves, in which the lines are seen to be associated into definite series. In such 

 cases, except under special circumstances, we may safely presume that all the lines 

 of a group arise from the motion of a single electron in every molecule of the gas. 



Very striking examples of such groups are present in the absorption spectrum 

 of oxygen and in the bright line spectrum of carbon. The oxygen of the earth's 

 atmosphere produces the great A group of double lines in the solar spectrum, as 

 well as the very similar great B group, and the a group. It also produces a group 

 more refrangible than D, about which we know less. This group is much fainter 

 than the others, and it is only under exceptional circumstances that it can be seen at 

 all in the solar spectrum. Each of the other three groups can be distinguished into 

 two sub-groups, which from their appearance have been called a head and a train. 

 The general features of these three groups are the same, and Mr. Higgs has made 

 a careful geometrical analysis of one of them, the great B group. ^ From his 

 analysis we may infer that the head and the trjvin are due to motions in the mole- 

 cules which are distinct, although related to one another. This conclusion receives 

 further support from the circumstance that in the double lines of ' the head ' it is 

 the violet component of each pair which is the stronger, while in the train it is the 

 red component of each pair which is the stronger. 



In a paper in the ' Scientific Transactions of the Royal Dublin Society ' for 

 1891, p. 563, the present author pointed out that, if we proceed on the probable 



auLuui ^uiuieu uuL LiiUL, ii wt 



' Proo. Roy. Soc, 1893, p, mO. 



