The molecules usually consist of _groups_ of atoms. Thus in the vapour of water, each molecule contains one atom of oxygen combined with two of hydrogen, and it is only at immense temperatures that this inner grouping is broken up. Naturally, the more complicated and heavier the molecular group, the more easily it is broken up by outside causes into simpler groupings, and molecules may exist in liquids or solids, which break up before they reach the gaseous form. Of such substances the chemist says that they “cannot be volatilised without decomposition.” In very rare instances does the gaseous molecule consist of a single atom; even those of the most perfect gases, such as hydrogen, oxygen and nitrogen consist of pairs which are not broken up at any known temperature. The pressure of a gas, and its tendency to expand is due simply to the motion and impact of its flying molecules, and it may be noted that at the same temperature and pressure equal volumes of all gases have the same number of molecules, the lighter molecules making up for their want of weight by their greater velocity. The average velocity of a molecule of oxygen (O₂) at freezing point is 461 meters per second or about that of a rifle-bullet. It must not be taken however, that in any given solid, liquid, or gas, all the molecules at any temperature move at a uniform velocity, but that each individual molecule may vary from moment to moment from rest up to a very high speed, while the temperature of the mass only represents the average. Thus it happens that in all liquids, and even in solids, a certain proportion of the molecules at any temperature will have a speed sufficient to enable them to leave the surface, and take the form of vapour, while a certain proportion will fall back and be caught and retained. Thus every liquid, and theoretically every solid, has a “vapour-pressure,” rising with the temperature, and depending on it only, and at the boiling temperature of the liquid equal to that of the atmosphere, or about 15 lb. per square inch, and therefore able to form bubbles in the interior of the liquid. If a little of a liquid is confined in a flask, the flask will become filled with its vapour, and so long as any of the liquid is present, the pressure of the vapour will depend only on the temperature and not at all on the respective quantities of liquid or vapour. Neither will it be affected by the pressure of other vapours or gases present in the flask, the total pressure in which will be the sum of the “partial” pressures of all the gases and vapours present.[52] [52] Cp. p. 421.