SECTIONAL TIL\NSACTIONS.— A. 305 



where [i. is the spin moment of the electron, Hq the magnetic field necessary 

 to destroy the superconductivity, h is Planck's constant, and Vq McLennan's 

 destructive frequency ; AWq, Hq and Vq are all extrapolated to the 

 absolute zero of temperature. In other words, the magnetic energy and 

 the vibrational energy required to break up the superconductive structure 

 are each approximately equal to the energy of the structure itself. 



What, in particular, I would like to ask Prof, de Haas is this : Does 

 he consider it likely that this anomaly in the Thomson effect is really so 

 intimately associated with superconductivity ? Or is the disparity between 

 the temperature ranges in which the two phenomena manifest themselves 

 too great for this to be possible ? 



(In his reply Prof, de Haas was understood to say that he considered 

 the temperature disparity to be too great for the effects to be so intimately 

 connected (or connected at all, I am not certain), and that he thought the 

 numerical agreement was an accident. He added, however, that Keesom 

 had recently found a real change in the ordinary specific heat at the super- 

 conductive critical temperature.) 



Sir R. T. Glazebrook, K.C.B.,F.R.S.,andDr. Ezer Griffiths, F.R.S.— 

 Electric and magnetic units. The Paris Conferetice of July 1932. 



Dr. J. M. Holm.— The initiation of gaseous explosions by small flames . 



Experiments are described in which a commonly accepted belief is 

 disproved — i.e. that the limiting diameter for propagation along a tube filled 

 with an explosive mixture is determined primarily by the thermal con- 

 ductivity of the tube material. The variation of the limiting diameter for 

 propagation along tubes, and for ignition through circular apertures in thin 

 plates , is represented by a series of graphs for explosive mixtures of hydrogen , 

 methane and ethyl ether with air. An experimental equation has been 

 found to fit the curves. A description is given of an almost spherical flame 

 which may burn in certain mixtures for several seconds without producing 

 general ignition. The shape of the various types of flame formed by an 

 explosive mixture burning at a circular orifice is shown b}^ several photo- 

 graphs. 



A theory of the failure of flames to travel along small tubes containing an 

 explosive gaseous mixture has been developed on the assumption that the 

 extinction of the flame is caused chiefly by the cooling effect of the unburnt 

 gas in contact with its external surface and an approximate formula, which 

 gives reasonable agreement with experimental values, deduced for the 

 limiting diameter. 



Afternoon. 



Visit to the Physics Department of the University of Leeds, where the 

 following demonstrations were arranged : 



Mr. J. EwLES. — Cathodo- luminescence. 



Mr. A. W. Foster. — Sotne thermo-electric measurements. 



Mr. F. A. Long. — Electromagnet protection. 



Dr. J. E. Roberts. — Electron impacts in gases at low pressures. 



Dr. J. E. Shirodker. — A new method of measuring the temperature 



of flames by the use of (x-particles. 

 Mr. F. W. Spiers. — The crystalline nature of tin amalgams. 

 Dr. J. E. Taylor. — Photographic action of electrons. 



