ON GASEOUS EXPLOSIONS. 213 



amplitude in the two cases. For there will then be the same number 

 of molecules vibrating in exactly the same way and arranged in the 

 same way, in the two cases. The only difference is in the scale of the 

 arrangement, and this can only affect the matter if the distance 

 between molecules is comparable with the wave-lengths of the radiation 

 emitted, which is not the case. It is only, however, within moderate 

 limits that the molecular vibrations are independent of density. 

 Angstrom found that the absorption of the radiation from a given source 

 in a tube of C0 2 at ordinary temperature and atmospheric pressure was 

 reduced by increasing the length and diminishing the pressure 1 in the 

 same proportion so as to keep the mass of gas constant. Schafer found 

 that on increasing the pressure the absorption bands of this gas were 

 widened, so that the curve connecting intensity of radiation and wave- 

 length did not remain of the same shape. 2 These experiments were 

 made at low temperatures, and at the higher temperatures in which 

 the Committee are more particularly interested there has been but little 

 work. There is no reason to doubt, however, that the character and 

 amount of the radiation from C0 2 and steam at high temperatures will 

 change with the density. 



From the point of view of the molecular theory, such a change 

 might be anticipated from either of two causes. An increase of density 

 implies a proportionate increase in the frequency of molecular col- 

 lisions, and this would result in greater facility of interchange between 

 the translational and atomic types of energy.. It is possible that the 

 equilibrium proportion of the two types might be different in conse- 

 quence. The denser gas may conceivably possess, with a given amount 

 of translational energy, more atomic energy, and therefore radiate more 

 strongly at a given temperature. It is certain that there would be a 

 more rapid attainment of equilibrium in the gas after an explosion, or 

 a rapid expansion. Another possible cause is a direct inter-action 

 between the molecules apart from collisions. Two molecules at a 

 sufficient distance apart will vibrate practically independently, each 

 behaving as though the other was not there, except that there will be a 

 tendency for them to vibrate in the same phase. But if the two are 

 close together they react on each other so that the natural period 

 or periods of the two together will not be the same as those which 

 each would have if it were isolated. 



Such direct measurements as have been made of the radiation after 

 a closed vessel explosion suggest that the flame is more transparent 

 than might be inferred from the experiments on open flames. According 

 to information given to the Committee by Professor Hopkinson, 

 W. T. David has found that the radiation received by a bolometer placed 

 outside a fluorite window in the cover of a cylindrical explosion vessel 

 30 cm. x 30 cm. is greatly increased by highly polishing that portion 

 of the opposite cover which can be ' seen ' by the bolometer. This 

 implies that a thickness of 30 cm. of flame in these circumstances 

 can transmit much of the radiation which it emits. The density of 



1 Ark. for Mat. Astron. och Fysik, Stockholm, vol. iv., No, 30, to. 1. 

 1 Aim, der Physik, vol. xvl. (1905), p. 93. 



