Sronsy—Of the Kinetic Theory of Gas. 369: 
nitrogen, and hydrobromic acid the motions associated with their 
two Ba degrees of freedom are able to rouse a certain amount of 
activity in adjoining Bd events, and that thus a part of their 
energy gets to be exposed to linkage. 
Hitherto we have regarded molecules as acted on in two ways 
only—by dynamical interactions between or within the molecules, 
and by the effect of electro-magnetic waves on such of the electrons 
as are undisguised charges of electricity. But there are other 
ways in which molecules may be acted on, of which the most 
conspicuous intervenes energetically on those critical occasions 
when chemical reaction takes place. Here it is the electrons that 
are primarily concerned, as is manifest from Faraday’s law of 
electrolysis. In cases of friction also, or of disruption of a crystal, 
it is manifest that some of the electrons are started into activity. 
In fact it may be presumed that the intermolecular bonds within 
a crystal are fundamentally of the same kind as the interatomic 
bonds within a molecule, and that in both it is interaction between 
electrons that is primarily called into play. It should also be 
noted that the number of electrons within an atom may be_ 
greater than its place in Mendeléefi’s Table would seem to suggest, 
as is, for example, evidenced by the chemical behaviour of potassium. 
In chemical reactions, if the product is gaseous, and if the 
electrons which are set swinging are associated with Ba events, 
the most obvious effect is a sudden increase of temperature and 
pressure ; if associated with Bb, events the most obvious effect is. 
a flash of light; and if associated with Bd, events both effects 
will be conspicuous : and this, in accordance with what we learn 
through the phosphoroscope, would seem to be what most fre- 
quently happens. 
It seems probable that it is when excited by chemical reactions 
that electrons produce their most conspicuous luminous effects, 
whether in flames or in so-called incandescent spectra. It should 
be remembered that as electrons are for the most part associated 
with 6) motions within the molecule, it may happen that they but 
slowly influence the temperature of the gas as indicated by the 
thermometer, and that accordingly the luminous effects may be 
greatly in excess of what a mere incandescent body at the same 
moderate temperature could produce. Hence the phrase incan- 
descent spectrum is not always appropriate, since the supposition 
