502 REPORTS ON THE STATE OF SCIENCE. 



been referred to as a factor in connection with the explosion-wave was 

 not a constant quantity. A bullet, for instance, cannot really travel 

 with a velocity greater than that of sound ; if it' did the air would be 

 shattered as if by an explosion and the bullet stopped. This result was 

 in practice prevented by the compression and consequent heating of the 

 air in front of the bullet, whereby the velocity of sound was momentarily 

 increased immediately in front of the bullet to perhaps three times its 

 ordinary value in accordance with the equation V 2 = fcRT. That very 

 great heating could thus actually occur was shown by the fact that 

 the compression was made use of in the Diesel engine to produce 

 ignition. The rate of explosion must depend a good deal on the amount 

 of exposed surface. Whilst hot surfaces promoted combustion, cool 

 surfaces unfortunately had an opposite effect. This was responsible 

 for the production of vast quantities of soot and smoke, especially in 

 firing steam boilers, and also gave rise to trouble in heating and anneal- 

 ing armour-plates. If a surface could be discovered which would pro- 

 mote combustion even at lower temperatures the discovery would be 

 of very great value. The escape of an electric charge from a contact 

 surface during combustion might very well be purely mechanical, the 

 positively-charged adhering layer being literally scraped away by the 

 force of the flame. 



Professor H. B. Dixon, referring to his recent work, said that 

 Nernst had suggested that a gas fired by compression would be heated 

 uniformly and would therefore detonate as a whole. This was not found 

 to be the case; the gas actually fires in a particular layer, though the 

 explosion begins rather indefinitely, and no well-defined sound-waves 

 are propagated from the point of ignition. The explosion of hydrogen 

 and chlorine by light was of special interest, as it did not occur in the 

 well-dried gas. In the moist gas there was an interval of time before 

 explosion took place, similar to the ' pre-flame period ' in gases fired 

 by adiabatic compression. But when once the wave was started, 

 whether by a spark or a flame or by light, it proceeded independently of 

 moisture, and indeed was actually most rapid in the dry gas. The 

 explosion was then propagated by molecular collision, and on account 

 of its high velocity, probably by collisions between pairs of molecules 

 only — a point on which Sir J. Larmor had laid special stress in his 

 Wilde lecture. In the explosion of hydrogen and oxygen, as in that of 

 hydrogen and chlorine, the action started by a spark was propagated as 

 well in the dried as in the undried gas. So far experiments on the 

 influence of an electric field had given negative results, and the action 

 of Rontgen rays on a mixture of hydrogen and chlorine did not render 

 it more sensitive to light. 



Supporting the suggestion of Sir J. J. Thomson, he thought it not 

 unlikely that an invisible compression-wave might travel just in front 

 of the visible flame, the particles being thereby raised to a high tempera- 

 ture. The brightness of the flame might well be due to the fact that 

 the gas was already very hot before combustion occurred. He proposed 

 to repeat the experiments of exploding gases in a strong magnetic field 

 and photographing the flame. 



Professor Smithells hoped that the report would be carefully read 



