670 Prof. Fleming and Mr. Dyke on the Production of 



total electrostatic energy of a molecule at any density p 

 equal to — 27r<r/3(m//)) 1/3 . But according to the law of the 

 inverse fourth power the mutual potential energy of the 

 attracting molecules is — 27re 2 s 2 /3(m/p) per molecule. It 

 appears then that — 2ire' 2 {(m/py 2 / 3 — s 2 }/3(m/p) must be in- 

 ternal electrostatic energy per molecule. The total electro- 

 static energy of a number of molecules is of the same form 

 as that of equal numbers of positive and negative electric 

 charges uniformly mixed as in the ions of an electrolytic solu- 

 tion. Thus an hypothesis advanced years ago by Fessenden 

 in America seems, when suitably interpreted, to be justified 

 by the discovery of Mills made also in America, though both 

 ignored fundamental physical difficulties in obtaining their 

 results. To account for the facts it is proposed to look upon 

 atoms as electrized, just as we speak of a magnet as being- 

 magnetized. But an electrized molecule causes the whole 

 of its domain to be electrized as if for a given chemical sub- 

 stance each molecule possessed an invariable amount of 

 positive electricity and an equal amount of negative, and as 

 if these amounts were uniformly distributed through the. 

 domain of the molecule, whether its state is gaseous, liquid, 

 or solid. This statement applies only to considerations re- 

 specting the total potential energy of molecules. In other 

 words, as regards total energy each molecule behaves as if 

 it had an electric moment proportional to the linear dimensions 

 of its domain, whereas in the matter of mutual energy each 

 molecule has an electric moment of amount es investigated in 

 connexion with the laws of molecular attraction. 

 Melbourne, Feb. 1909. 



LXI. A Note on the Production of Steady Electric Oscillations 

 in Closed Circuits and a Method of Testing Radiotelegraphic 

 Receivers. By J. A. Fleming, M.A., D.Sc. s F.R.S., and 

 G. B. Dyke, B.Sc* 



IN testing radiotelegraphic detectors the difficulty is 

 generally to obtain facilities for working in actual 

 stations and at various distances. Thus, if an inventor 

 desires to know whether an improvement which he has made 

 in oscillation detectors is an advance on anything yet done, 

 he must be able to test this receiver at a station in corre- 

 spondence with others at various and at considerable distances, 

 and even then quantitative measurements are difficult, or 

 impossible, to obtain on account of the continually varying 

 atmospheric conditions which, as is well known, introduce 

 * Communicated by the Physical Society : read March 26, 1909. 



