TRANSACTIONS OF SECTION A. 555 



ments of Regnault. It is the authors' intention to continue the experiments, and 

 extend the method to higher temperatures and to higher pressures in order to 

 obtain results that will be applicable to present engineering practice. 



7. Xeiv Varieties of Katliode Rays. Bxj SiLVANUS P. Thompson, F.B.S. 



8. On the Spectra of Oxygen, Sulphur, and Selenium. 

 By C. RuNGE and F. Paschen. 



The spectrum of oxygen when an electric current is passed through a vacuum 

 tube containing that body, and when no spark gap or Leyden jar is interposed in 

 the circuit, closely resembles the spectrum of helium. It consists of six ' series' of 

 lines forming two sets of three each. Each set of lines is very similar to the whole 

 spectrum of any one of the alkali metals. There is therefore no more spectroscopic 

 evidence in favour of the supposition that helium consists of two elements than 

 there is for oxygen. Under similar circumstances sulphur and selenium give out 

 each a spectrum closely resembling one of the sets of three series. But we are not 

 sure whether the other set does not find its analogy also. The three sets in the 

 spectra of oxygen, sulphur, and selenium, which are analogous to one another, all 

 consist of triplets of a very marked character, the difference of wave numbers 

 increasing as we pass from oxygen to sulphur and from sulphur to selenium in 

 roughly approximate proportion to the squares of their atomic weights. The spec- 

 trum of each of the three bodies, as a whole, is situated further to the side of the 

 smaller wave numbers — that is to say, it consists of slower oscillations the greater 

 the atomic weight of the body.* 



9. T1t,e Influence of Pressure on Spectral Lines. 

 By J. Larmor, F.B.S. 



A definite picture of the relations of the tether and matter is obtained by 

 assuming the material molecule to be made up of electrons or intrinsic strain- 

 centres in the tether.'- A system of electrons describing steady orbits round each 

 other, after the manner of the bodies of a solar or stellar system, would represent 

 a molecule ; any disturbance of this steady motion would induce radiation across 

 the tether, which would last until it had reduced the motion again to a state of 

 steadiness. The natural configuration of a molecule would, however, be the 

 unique one of minimum energy corresponding to its intrinsic constant rotational 

 momenta, for the influence of radiation would set towards this configuration, and 

 would not allow much departure from it. 



The wave-lengths of luminous radiation are about 10^ times the linear dimen- 

 sions of the molecules ; thus the intrinsic luminous periods are those of rather 

 slow periodic inequalities (in the sense of physical astronomy) in the orbital 

 motions. This circumstance allows us to roughly appreciate the order of magni- 

 tude of the influence of the surrounding medium on these free periods. On account 

 of their slowness the tethereal oscillations which are governed by the inequalities 

 of the orbits of the electrons are sensible over the space occupied by some thousands 

 of molecules each way, and this number is so great as to tempt us to form an idea 

 of the influence of these imbedded molecules by considering them to form a con- 

 tinuous medium. If now the molecules were vibrating in a homogeneous medium, 

 say, surrounded by simple aether, the free periods would vary inversely as the square 

 root of the elasticity of this ambient medium, provided we could assume that 

 change of the medium did not involve change of type of the steady intramolecular 

 orbits. This latter circumstance, however, will also operate to alter the periods, 



' See AV^iedemann, Annalen, 61, p. 641, 1897. 

 ■ * Cf, Phil. Trans., 1895, pp. 695-743. 



