in relation to Pressure, Volume, and Temperature. 395 



become so much the more considerable as the molecular state 

 of the gas corresponds less to these conditions." 



The deviations here mentioned must make themselves mani- 

 fest, so far as the pressure is considered at a given tempera- 

 ture and a given volume, in this being either greater or less 

 than it should be, according to Mariotte and Gay-Lussac's 

 law, taking as the starting-point the highly rarefied state of 

 the gas. The space above-mentioned (under 1) actually filled 

 by the molecules conditions an augmentation of the pressure, 

 since by it, at a given volume of the gas, the free space for the 

 motion of the molecules is lessened, and accordingly the num- 

 ber of the collisions is increased. The molecular forces men- 

 tioned under 3 effect, when attraction prevails, a diminution 

 of the pressure. The circumstance mentioned under 2, namely 

 the time that passes during a collision, has a more complicated 

 effect, since, when two molecules rush against one another, 

 first an acceleration, and only after this the retardation and 

 reversal of the motion take place ; hence, in considering the 

 total effect, that of this circumstance can be joined partly to 

 that of the first, partly to that of the last circumstance. 



Of the two opposite effects of augmentation and diminution 

 of pressure, the one or the other may preponderate, according 

 to circumstances. The above-mentioned experiments of Reg- 

 nault and batterer show in hydrogen a universal preponde- 

 rance of the increase of pressure, while they make known that 

 in the other gases diminution of pressure predominates when 

 the densities are less, and increase of pressure when the den- 

 sities are greater. 



The behaviour of a substance becomes still more complicated 

 w r hen on condensation it does not still remain gaseous, but 

 changes its aggregate-state by becoming liquid. On the con- 

 nexion of this process with those previously discussed some 

 very fine experiments have recently been made by Andrews*. 

 Submitting carbonic acid at various temperatures to powerful 

 compressions, and observing the increases of pressure that took 

 place, he found that there is an essential difference in its be- 

 haviour, according as the temperature is above or below 31°. 

 Above 31°, only the above-mentioned deviations from Mariotte 

 and Gay-Lussac's law are shown, while below 31° at a certain 

 pressure condensation takes place. 



Andrews represents the relations between pressure and 

 volume observed by him by curves which have pressure and 

 volume as abscissae and ordinates. The curves referring to 

 temperatures above 31° exhibit a continuous course ; while 

 those referring to temperatures below 31° are broken, as with 

 diminishing volume the pressure at first increases, but from a 

 * Andrews, Phil. Trans. 1869, p. 575. 

 2F2 



