02 THE PRINCIPLES OF SCIENCE. 



same exact places, and the stability of the system be esta 

 blished. 



But any such theory of periodic stability is really hypo 

 thetical, and does not take into account a multitude of 

 phenomena resulting in the dissipation of energy, which 

 may be a really secular process incapable of restoration. 

 For our present purposes we really need not attempt to 

 form any opinion on such lofty questions. Any change 

 which does not present the appearance of a periodic 

 character will be empirically regarded as a secular change 

 for the present, so that there will be an abundant supply 

 of non-periodic variations. 



The variations which we produce experimentally will 

 often be non-periodic. When we communicate heat to a 

 gas it increases in bulk or pressure, and as far as we can 

 go the higher the temperature the higher the pressure. 

 Our experiments are of course restricted in temperature 

 both above and below, but there is every reason to believe 

 that the bulk being the same, the pressure would never 

 return to the same point at any two different tempera 

 tures. We may of course repeatedly raise and lower the 

 temperature at regular or irregular intervals entirely at 

 our will, and the pressure of the gas will vary in like 

 manner and exactly at the same intervals, but such an 

 arbitrary series of changes would not constitute Periodic 

 Variation. It would constitute a succession of distinct 

 experiments, which would place beyond reasonable doubt 

 the connexion of cause and effect. 



Whenever a phenomenon recurs at equal or nearly 

 equal intervals, there is, according to the theory of pro 

 bability, considerable evidence of connexion, because if 

 the recurrences were entirely casual it is exceedingly 

 unlikely that they would happen at equal intervals. Thus 

 the mere fact that a brilliant comet had appeared in the 

 years 1301, 1378, 1456, 1531, 1607, and 1682, gave con- 



