2. These “nitro-aerial particles” are also present in saltpetre or nitre, as can be shown by mixing inflammable substances, such as sulphur and charcoal, with nitre to form gunpowder, filling a tube with the powder, and, after setting it on fire, immediately plunging the open end of the tube under water. The sulphur and charcoal will be as completely consumed as if burned in the open air. Such combustion might, however, be ascribed to a “sulphureous” constituent in saltpetre; by “sulphureous” is to be understood combustible, for those substances capable of burning were imagined to contain a “sulphur” which gave them that property. That nitre does not contain such “sulphur” can be shown by exposing it alone to heat, when no change takes place, except fusion. Besides, nitre is compounded of “spirit of nitre” or nitric acid and pure alkali, neither of which contains a combustible sulphur; hence the particles of fire-air must be present in nitre in no small amount. But it is probable that it is the spirit of nitre which contains such fire-air particles, because, as will be shown later, they are not present in the alkali. One difficulty occurs to Mayow. How is it that so large a quantity of gas as is necessary to support combustion can be contained in a relatively small bulk of saltpetre? He tries whether a solution of saltpetre evolves air-bubbles when placed in a vacuum, and finds that it effervesces less than pure water does. He also prepares saltpetre by mixing nitric acid and alkali in a vacuum; a brisk effervescence occurs, and the dried-up salt is ordinary saltpetre. Hence saltpetre cannot contain elastic air. Mayow consequently draws a distinction between “air” and “air-particles.” The residue left after the “fire-air,” or _spiritus igneo-aerius_, has been removed from ordinary air by breathing or by combustion is proved to be lighter than the fire-air itself; because a mouse dies sooner if kept at the top of air in a confined bell-jar than at the bottom; and a candle goes out sooner. Here the conclusion is right, although the reason given is wrong; for it is the temperature of the respired air which makes it rise, and not the fact that it is specifically lighter than the oxygen. Metallic antimony gains in weight when it is set on fire by a lens, and burns; if this gain in weight, Mayow remarks, is not due to the absorption of nitro-aerial particles and to the fire, it is difficult to say to what it is due. The reason why substances burn so violently in nitre compared with air, is because of the proximity of the fire-air particles; and these are evidently due to the nitric acid, because the residue--the alkali--if mixed with sulphur and inflamed, does not produce ignition.