364 Past Changes in the Universe. [J uly, 
explanation of the mechanism of gravity that has received 
support by competent judges. 
From the deduction that the stars are moving in straight 
lines in various directions, the analogy (as regards character 
of motion) with a gas becomes natural and inevitable ; for 
dynamical principles are independent of size or relative 
scale, or it is the same as if we were dealing with molecules 
observed to have a proper motion among each other in 
various directions. The analogy of a gas presents itself, 
therefore, rather as a necessary consequence, not as a spe- 
culation.* The degree of aggregation of the separate parts 
of the Universe moving among each other in straight lines 
would therefore appear to be necessarily dependent on the 
mean velocity of proper motion, just as the degree of aggre- 
gation of the components of the molecules of a compound 
gas depends on the mean velocity of their proper motion. 
If the velocity of the compound molecules of the gas be 
increased (which is synonymous with what is called “ raising 
the temperature ”) they begin to separate into their compo- 
nents (or to become “ dissociated ”), and if the molecules 
have a very high complexity the extent to which this sepa- 
ration occurs (i.e., the final degree of aggregation) depends 
on the temperature, or on the velocity of the molecules. If 
the velocity be very high (by a high “temperature”) the 
molecules split up — “dissociate ” — into their ultimate com- 
ponents. By gradually lowering the temperature (velocity) 
the converse process may take place, or the components 
may gradually aggregate together to form massive mole- 
cules. This analogy is applicable also to the larger scale 
parts of the Universe moving in straight lines among each 
other, since dynamical principles are independent of scale. 
Such an excessive velocity of the component parts of the 
Universe (the stars) among each other is quite conceivable, 
at which the whole would break up into single molecules. 
By a lower velocity these molecules collect together under 
* The relative velocity of the stellar masses (meaning by this the relative 
time taken by them to traverse a distance equal to their own diameters) is 
almost indefinitely slow compared with the case of the molecules of gases, 
which traverse a distance equal to many millions of times their own diameters 
in a single second. The absolute velocity of the stellar masses, however (as 
far as observed), is very great compared with that of the molecules of gases, 
and consequently the heat developed at their encounters would be very great. 
On account of the large multiple the distance of the stellar masses is of their 
diameters, they would, no doubt, traverse vast distances before an encounter, 
or their “ mean length of path ” would be very great ; and (considering the 
relative length of time a stellar mass takes to traverse a distance equal to its 
own diameter) an enormous epoch of time would in general elapse between 
the encounters. 
