Recurring Changes in the Universe. 157 



region of the gas, the mean -temperature {per unit of volume) of 

 the gas remaining the same. One known consequence of the 

 kinetic theory is that vast fluctuations of temperature may 

 occur in the gas, from one of the portions of matter (consisting 

 of one molecule or several) that moves as a whole in the mo- 

 tion of translation, to another portion, though the whole 

 (reckoned by units of volume) may be in equilibrium of tem- 

 perature. The same would apply in principle to the case of 

 the universe, consisting of portions of matter in translator^ 

 motion among each other, if we do not limit the scale on which 

 the principles are applicable ; or it would follow that there 

 might be vast differences of temperature from one portion of 

 matter (stellar mass) to another, while the whole (reckoned by 

 units of volume) might be in equilibrium of temperature. In 

 order to have an idea, by inspection, of the mean state of tem- 

 perature of the universe, w r e should require to sw T eep over a 

 unit of volume, containing some hundreds of milions of stellar 

 masses (dark and bright), in the same way as we do (on 

 smaller scale) in examining the state of temperature of a gas ; 

 any appreciable volume of which contains some hundreds of 

 millions of portions of matter in translatory motion (at differ- 

 ent temperatures). 



If it were possible to visualize the individual molecules in 

 a compound gas at normal temperature, molecules would be 

 observed in various parts glowing at a white or red heat (and 

 some in a state of dissociation) ; and if the state of tempera- 

 ture of the whole gas were judged of from these relatively few 

 luminous molecules, the assumption would be that the whole 

 gas was at a white heat. So in the case of the universe, if the 

 state of temperature of the whole were judged of from the 

 perhaps relatively few luminous masses [which of course can 

 alone be visible to us], an entirely false impression might be 

 conveyed of the real state. 



If therefore the above suggestion as to the possible applica- 

 tion of an admitted dynamical principle on a large scale should 

 be found valid, it would follow that the universe may already 

 be at uniform temperature, in the sense that the limits within 

 which there is fluctuation of temperature are indefinitely small 

 compared w r ith the collective universe, but that these limits 

 are, relatively to a planetary system, very great, and amply 

 sufficient to allow continual physical change, adapted to the 

 maintenance of life. 



So in an analogous way as regards the state of aggregation 

 of the matter of the universe : since, by the application of the 

 kinetic theory, this depends on the temperature, it would fol- 

 low that the mean state of aggregation of the matter (per unit 



